. ,2.
r m?
E-T‘J.

.2 1 7
. “E;
mag.

:
'5 .c U
. .fihfixrh

rt .1.‘
a 5915 .3

.fififiwu

 

 

 

3.“: .x ‘ L.
. (.x.:x.s.
.. .

a. L
. .1‘
i . .5, .Wmmwimm
: . . . .
t . _ ,w .4. ms .
.a,‘
. IE: .
u As

13.. .

div. .
9'}:
r!

 

 

. V ‘ V . . #:54?‘
. ‘ ‘

Wuwmmr WHfi.

g

.7... 9 an“ «5.5

a.
. a ”in
w%

f~

4.4“. h.

: . ”rarémn?»
Em. fig

.1 .

$1
..

\ ru
“our... V
2

5,!

K.

34]..»
1 n... 3

2mm

3“ . I
:35 Ag...
xi: ,. . .1
s: tn...“ a?!) 4
. ~ .. but“ tin“:
2 i s .n
; m..§.....n:...a... 3‘
as (
311m}. ..

imwvuhi.
; _ V9 .

L:
3.

. whim. ‘
Vufivafih‘n
r: c .

.14.: is.
I). t \1 \
“wanton...

 

 

THFSrs

25H

This is to certify that the

dissertation entitled

NEW SYNTHETIC ROUTES TO GLYCOSIDASE INHIBITORS,
BIOLOGICALLY ACTIVE DISACCHARIDES AND GLYCOSYL DONORS

presented by

Gabriela Pistia-Brueggeman

has been accepted towards fulfillment
of the requirements for

Ph.D. Chemistry

degree in

 

 

L ;~ 91: .0. -
~60: profesior

Date December 7, 2000

 

MSU is an Affirmative Action/Equal Opportunity Institution 0- 12771

 

 

l LIBRARYW'
Michigan State
University ,5

 

 

 

this checkout from your record.

PLACE lN RETURN BOX to remove
on or before date due.

 

To AVOID FINES return
MAY BE RECALLED with earlier due date if requested.
DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c/CIRCDateDUOp65-DJS

 

NEW SYNTHETIC ROUTES TO GLYCOSIDASE INHIBITORS,
BIOLOGICALLY ACTIVE DISACCHARIDES AND GLYCOSYL DONORS

By

Gabriela Pistia-Brueggeman

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Chemistry

2000

ABSTRACT

NEW SYNTHETIC ROUTES TO GLYCOSIDASE INHIBITORS,
BIOLOGICALLY ACTIVE DISACCHARIDES AND GLYCOSYL DONORS

By

Gabriela Pistia-Brueggeman

Complex carbohydrates and their conjugates in biological systems are either structural or
informational molecules. On the cell surface many carbohydrates of glycoconjugates are
involved in various types of biochemical recognition processes, including growth,
development, immune responses, infection, cell adhesion, metastasis, and numerous
signal transduction events. This dissertation is focusing on the synthesis and screening of
sugar analogs with high therapeutic potential, preparation of the disaccharide structural
component of bacterial cell wall lipid A and finding new glycosyl donors, to facilitate

oligosaccharide synthesis.

The first part of this dissertation (chapters 2 & 3) describes new synthesis and screening
procedures for glycosidase inhibitors, especially 1,5-iminoalditols and aldono 5-lactones.
These two classes of compounds have high inhibitory activity and they constitute viable
candidates for cancer, AIDS and diabetes treatment. A new, general method for the

preparation of such compounds with the D-gluco and D—galacto configuration starting

 

 

 

 

from
due

mctl.
to 4 t
and
glucul

screen

Chdpi;

Odedt:

The 1%:
headgn,
“pom,
of Gran»
re“lions,-

Stine-tun:

 

from B-glycosides is presented. The procedure stands out among the preexisting methods
due to its generality, straightforwardness, high yield and stereoselectivity. This
methodology has been used in the one pot synthesis of a library of compounds belonging
to 4 classes of substances: methyl glycosides, aldonic acid lactams, aldonic acid lactones
and 1,5-dideoxy-l,S-iminosugars. These have been screened for activity against
glucosidases. The approach provides a facile and rapid route towards the synthesis and

screening of a multitude of potential glycosidase inhibitors.

Chapter 4 discusses an improvement and new application of the chromium trioxide
oxidation of glycosides, while chapter 5 describes a novel protecting/activating group for
N-acetylglucosamine. o-Nitrophenyl 2-acetamido-2-deoxy-B-D-glucopyranoside can be
activated with Zan to afford, in the presence of an alcohol, high yields of N-acetamide

B-D-glucosides.

The last chapter presents various attempts towards the total synthesis of the disaccharide
headgroup of lipid A from Rhizobium T rifolii. Lipid A is a component of
lipopolysaccharides (LPS), which are a unique class of glycolipids found on the surface
of Gram-negative bacteria. LPS are capable of eliciting a wide array of biological
responses when they interact with cellular systems of animals. Determination of their

structure can help in the elucidation of their physiologic activity.

To my mother

Without her all this would not have happened

iv

ACKNOWLEDGMENTS

Entering graduate school has been an important decision in my life, and now, at the verge
of concluding this ‘chapter’, I can say that I cherished every moment. I want to believe
that I’m emerging as a more knowledgeable, wiser, and stronger person. However, this
would not have been possible without the help of some wonderful people.

Most of my gratitude goes to my advisor Rawle I. Hollingsworth. With his help I
discovered the beautiful and complex world of carbohydrate chemistry and a whole new
philosophy of life. He was always there for me with help, advice and encouragement
from the first day I joined our lab to the last step of my graduate work. To him I owe the
mastering of all techniques that I learned, the expanded vision of chemistry and the new
perspective on life. Thank you Rawle, you have been a valuable mentor and at the same
time a cherished friend.

I want to address my appreciation to the members of my committee, Dr. Gregory Baker,
Dr. Marcos Dantus and Dr. Katharine Hunt. I want to thank Dr. Baker and Dr. Hunt for
taking their time to revise this thesis and Dr. Dantus for giving me valuable advice
regarding resume writing, job hunting and coping with my newborn.

I want to thank the member of Hollingsworth’s lab for being there for me in good times
and in stressful times, for their help and support and for being such good friends.
Together with Rawle, they made coming to school not only worth wile but also a
pleasure. Thank you Carol, Ben, Jim, Atima, Hussen, Lakshmi, Guangfei, Rob, Jeongrim,

JJ, Ing, Guijun, lie and Luc.

 

 

 

 

 

 

Th3:

flier

 

 

l arr.
my P
|
the 117‘
uncor

“hem

lamr

 

and lb
did no
precio'
has bn

T0 all

Thanks are also going to Long Lee and Kermit Johnson from the NMR facility, for their
friendly help.

I am concluding my acknowledgments by thanking the most important people in my life:
my parents, my son Justin and my husband Michael. There are not enough words to cover
the gratitude that I have towards my parents, and especially my mother. She gave me her
unconditional love and support, guided me along my life journey and was there for me
whenever I needed her. I only hope that I can give back the love and help that I received.
I am thanking my husband Michael for the peaceful life and support that he has offered
and the cheerfulness that he has brought into our relationship. Because of this, my work
did not seem as hard as it sometimes was. And at least but not at last I want to thank my
precious little son Justin for the loving smiles and delightful 0003 and for the hope that he
has brought to our lives.

To all of you THANK YOU.

vi

 

TABLE OF CONTENTS

LIST OF TABLES .............................................................................. ix
LIST OF SCHEMES ............................................................................ x
LIST OF FIGURES ............................................................................. xii
Chapter 1. Introduction for glycosidase inhibitors ....................................... 1
1.1. Classification of glycosidase inhibitors ......................................... 3
1.2. Activity and importance of glycosidase inhibitors ............................ 7
1.3. General mechanism of glycoside hydrolysis ................................... 17
1.4. Mechanism of action of glycosidase inhibitors ................................ 19
1 .4. l . Aldonolactones ........................................................... 19
1.4.2. 5-Amino-5-deoxylactams ............................................... 21
1.4.3. Cyclic iminosugars ...................................................... 23
1.5. References ........................................................................... 29
Chapter 2. A General Approach to the Synthesis of Imino-Dideoxy and
-Trideoxy Alditols from b-D-Glycosides ..................................... 33
2.1. Introduction ......................................................................... 35
2.2. Results and discussions ............................................................ 50
2.3. Conclusions ......................................................................... 55
2.4. Experimental section ............................................................... 55
2.5. References ........................................................................... 62
Chapter 3. A Preparation and Screening Strategy for Glycosidase Inhibitors. . 66
3.1. Introduction ......................................................................... 68
3.2. Results and discussions ............................................................ 71
3.3. Conclusions ......................................................................... 74
3.4. Experimental section ............................................................... 76
3.5. References ........................................................................... 83
Chapter 4. Chromium Trioxide Oxidation of Glycosides. A Comparative Study. 85
4.1. Introduction ......................................................................... 87
4.2. Results and discussions ............................................................ 92
4.3. Conclusions ......................................................................... 97
4.4. Experimental section ............................................................... 99
4.5. References ........................................................................... 103

vii

Chap!

/
&

Chapte

oqnfi‘

 

 

 

 

 

 

 

Chapter 5.

5.1.
5.2.
5.3.
5.4.
5.5.

Chapter 6.

6.1.
6.2.
6.3.
6.4.
6.5.

Development of a New Glycosylation Procedure for 2-Amino-2-

Deoxysugars ........................................................................ l 04
Introduction ......................................................................... 1 06
Results and discussions ............................................................ 1 12
Conclusion .......................................................................... 1 16
Experimental section .............................................................. l 17
References .......................................................................... 120
Approaches Towards the Total Synthesis of the Head-Group of

Rhizobium T rifolii ................................................................. 122
Introduction ......................................................................... 124
Results and discussions ............................................................ 127
Conclusions ......................................................................... 140
Experimental section ............................................................... 140
References ........................................................................... l 53

viii

 

 

Table 3.-

Table 3.

Table 5.

 

Table 3.1.1.

Table 3.2.1.

Table 5.1.1.

LIST OF TABLES

Inhibitory action of some glucosidases ........................................ 70

Structure, characterization and inhibitory action of some synthesized

glucosidases ........................................................................ 75
Conditions, advantages and limitations for the most common glycsylation
procedures ........................................................................... 111

ix

 

 

 

 

Scheme

Scheme 1..

Scheme 1.-

Scheme
Scheme
Scheme
Scheme
Scheme
Scheme
Scheme
Scheme
Scheme
SchL’me
Scheme
Schem e
Scheme

Schme

Scheme :

SChEllle

 

 

Scheme

Scheme 1.4.1.1.

Scheme 1.4.1.2.

Scheme
Scheme
Scheme
Scheme
Scheme
Scheme
Scheme
Scheme
Scheme
Scheme
Scheme
Scheme
Scheme
Scheme

Scheme

Scheme

1.3.

1.4.2.

2.1.1.

2.1.2.

2.1.3.

2.2.1.

2.2.2.

2.2.3.

3.2.1.

3.2.2.

LIST OF SCHEMES

General mechanism of glycosidase inhibitors ............................. 18

Interconversion between 1,5-gluconolactone and 1,4-gluconolactone. 19

Geometrical similarity between aldono-1,5-lacones and the
pyranosyl oxocarbenium ion and difference to the conformation of
aldopyranosides ................................................................ 20
Configuration of aldono-1,5-lactams ....................................... 22
Synthesis of di-DNJ by enzymatic aldol condensation ................... 39
Synthesis of di-DNJ by asymmetric Diels-Alder reaction ............... 41
Synthesis of di-DNJ by double reductive amination ..................... 42
. Synthesis of DNJ from S-pyroglutamic acid .............................. 44
. Synthesis of DNJ by intramolecular aminomercuration .................. 45
. Synthesis of DNJ from L-sorbose ........................................... 47
. Chemo-microbiological synthesis of DNJ ................................. 48
. Chemo-enzymatic synthesis of DNJ ........................................ 49
Synthesis of 1,5,6-trideoxy-l,5-imino glucitol ............................ 52
Synthesis of 1,5,6—trideoxy-l,5-imino galactitol .......................... 53
Synthesis of gluconolactam and DNJ ....................................... 54
. Preparation of 1,5-gluconolactam from nojirimycin ..................... 69
Synthesis of d-gluconolactam, dideoxynojirimycin and DNJ ........... 72

Synthesis of partially methylated S-mnino-S-deoxy-gluconic acid
lctams, 1,5-dideoxy-l ,5-imino-glucitols and gluconic acid lactones... 73

. Chromium trioxide oxidation of methyl 2,3,4,6-tetra-O-acetyl-

gucoside ........................................................................ 87

 

 

 

Schem-
Schem

Schem

Schemt

Scheme

Scheme

Scheme

Scheme
Scheme
SCl'lt’rne

Scheme 1
Scheme 6
Scheme 6
SchCine 6
Schem 6

Schme 6.

Schfihe 6 -

 

Schelne 6

 

Scheme
Scheme

Scheme

Scheme

Scheme

Scheme

Scheme

Scheme
Scheme
Scheme

Scheme

Scheme

Scheme

Scheme

Scheme

Scheme

Scheme

Scheme

4.1.2.

4.2.1.

4.2.2.

4.2.3.

4.2.4.

4.2.5.

5.2.1.

5.2.2.

5.2.3.

6.2.1.

6.2.2.

6.2.3.

6.2.4.

6.2.5.

6.2.6.

6.2.7.

6.2.8.

Oxidation of methyl 6-deoxy-S-C-methy1-B-D-xylo-hexopyranoisde.. 9O

. Propose mechanism for the oxidation of glucopyranosides by CrO3... 91

Chromium trioxide oxidation of methyl 2,3,4,6-tetra-O-acetyl-
glactoside ...................................................................... 92

Oxdation of l,2,4,6-tetra-O-acetyl-3-O-benzyl-B-D-glucopyranose... 93

Oxidation of benzyl 2-acetamido-2-deoxy-3,4,6-tri-O-acetyl-or-D-
gucopyranoside ............................................................... 94

Classical route for the synthesis of benzoyl glycosides .................. 95

Competition between benzyl oxidation and ring opening in the
oxidation of benzyl B-D-glucopyranosides ................................. 96

Synthesis of nitrophenyl-2-amino-2deoxy-g1ucoside ..................... 1 l3
Transglycosylation between o-nitrophenylglycosidess and alcohols... 114
Mechanism of transglycosylation reaction ................................. l 15

First attempt at the synthesis of the disaccharide head-group of
lipid A ........................................................................... 128

Unsuccessful glycosylation between acetobromogalactose and
o-nitrophenyl glucosamine ................................................... 129

Synthesis of b-disaccharide 8 and unsuccessful anomerization
towards 9 ........................................................................ 130

Synthesis of glycosyl donor 11 and attempted glycosylation
reactions ........................................................................ 132

Synthesis of lactyl 2-acetamido-2-deoxy—3,4-di-O-acetyl-B-D-

glucopyranoside ethyl ester 13 .............................................. 133

Synthesis of donor 18 and acceptor sugar 19 .............................. 135

Second attempt at the synthesis of the disaccharide head-group of

lipid A ........................................................................... 136

Oxidation of methyl or-D—galactoside with TEMPO ..................... 137
xi

Scheme 6.2.9. Third attempt at the synthesis of the disaccharide head-group of

lipid A ......................................................................... 138
Scheme 6.2.10. Synthesis of lactyl 2-acetamido-2-deoxy-3,4-di-O-acetyl-B-D-
glucopyranoside methyl ester 13 ........................................... 139
xii

 

 

Fi

ry-
E,
m

Figure
Figure
Figure
Figure
Figure 1
Figure 1
Figure 2
Figure 3.
Figure 4_;
l:lgure 5.1
Figure 6.]

Figure 6.:

LIST OF FIGURES

Figure 1.1.1. Distinction between competitive and noncompetitive inhibitors. . . . . . . . 3

Figure 1.1.2. Representatives of the most common classes of glycosidase inhibitors. .. 5

Figure 1.2.1. Three major classes of glycosidase inhibitors ................................. 7

Figure 1.2.2. Aza-sugars possessing antidiabetic activity .................................... 9

Figure 1.2.3. Compounds possessing cancer metastasis inhibitory activity ............... 12
Figure 1.2.4. Compound possessing antiviral activity ........................................ 15
Figure 1.4.3. Assumed ground-state binding of DNJ to B-glucosidase ..................... 25
Figure 2.1.1. Iminosugars with therapeutic importance ...................................... 36
Figure 3.1.1. Biologically active 5—lactarns ................................................... 69
Figure 4.2. 1. Oxidation of benzyl-B-D-glucopyranoside .................................... 98
Figure 5.1.1. General scheme of glycosylations ............................................... 109
Figure 6.1.1. Structure of Rhizobium T rifolii lipid A ......................................... 126

Figure 6.2.1. Assigned structure for the lipid A head-group of Rhizobium T rifolii ...... 127

xiii

 

Chapter 1

Introduction for Glycosidase Inhibitors

 

 

disorder
multiple
inhibitim
rei'iewec
require 5;

and the a.

 

 

Abstract
Glycosidases modify glycoconjugates by hydrolyzing glycosidic linkages, a process
which is essential for normal cell growth, regulation, and development. These
glycoenzymes are interesting targets for inhibition, as they are involved in metabolic
disorders and other diseases. Glycosidase inhibitors have the potential to produce
multiple beneficial therapeutic effects including reduction of the blood glucose level,
inhibition of tumor metastasis and stopping viral replication, all of which will be
reviewed in this introductory chapter. Among the numerous inhibitors, three classes
require special attention: iminosugars, lactams and lactones. Their mechanism of action

and the advantages over the other inhibitors will be discussed.

 

 

 

1.1. (
A var
denia
Classii
dissoci
the en.
charact
inhibitir

bUt not '

HEW 1.1,

1.1. Classification of glycosidase inhibitors

A variety of glycosidase inhibitors are known so far, most of them being carbohydrate
derivatives or their structural analogs. According to their way of action, they can be
classified into reversible and irreversible inhibitors [1]. An irreversible inhibitor
dissociates very slowly from its target enzyme because it becomes very tightly bound to
the enzyme, either covalently or non-covalently. In contrast, reversible inhibition is
characterized by a rapid dissociation of the enzyme-inhibitor complex. In competitive
inhibition, the enzyme can bind the substrate (forming an ES complex) or inhibitor (E1)

but not both (ESI) (Figure 1.1.1).

Competitive
Substr\at: inhibitor Substra\t:
Non-
competitive
inhibitor

 

ZJ/

   

 

(a) (b) (c)

Figure 1.1.1. Distinction between competitive and noncompetitive inhibitors:
(a) enzyme-substrate complex; (b) a competitive inhibitor prevents the
substrate from binding; (c) a noncompetitive inhibitor does not prevent the

substrate from binding.

Many competitive inhibitors resemble the substrate and bind to the active site of the
enzyme. The substrate is therefore prevented from binding to the same active site. A
competitive inhibitor diminishes the rate of catalysis by reducing the proportion of
enzyme molecules bound to a substrate. In noncompetitive inhibition, which is also
reversible, the inhibitor and substrate can bind simultaneously to an enzyme molecule,
but their binding sites do not overlap. A noncompetitive inhibitor acts by decreasing the
turnover number of an enzyme rather than by diminishing the proportion of enzyme
molecule that are bound to the substrate. Noncompetitive inhibition, in contrast with
competitive inhibition, cannot be overcome by increasing the substrate concentration.
Figure 1.1.2 presents the most common classes of glycosidase inhibitors. Reversible
inhibitors are: aldonolactones, 5-amino-S-deoxyaldonolactams, glycosylamines, cyclic
iminosugars like nojirimycin and deoxynojirimycin (DNJ), indolizine alkaloids like
castanospermine and swainsonine, polyhydroxypyrrolidines, and aminocyclitols. There
are also sugar-related compounds, called pseudosubstrates, that are chemically
transformed by glycosidases, ofien forming long-lived intermediates and thereby acting
as reversible inhibitors. Some of these enzyme-bound intermediates are cleaved so slowly
that they constitute a transition towards irreversible inhibitors. Some of these
pseudosubstrates are: D-glycals, heptenitols, glycosyl fluorides, and 2-deoxy-2-fluoro-
glycosides. The last to be mentioned are the irreversible inhibitors, which react
specifically at the active site and inactivate the enzyme, thereby providing important
information about the mechanism of the enzyme-catalyzed reaction. Some examples are:
conduritol epoxides (1,2-anhydroinositols), aziridines, glycosylmethyltriazenes, and

glycosyl isothiocyanates [2]. This dissertation will focus on three of the most potent

 

 

 

glycosic'

Before t

 

will be d

H(

Y '\
EIULD

HO/\
HO\

Ewe I]

glycosidase inhibitors which are 1,5-5—lactams, -lactones and 5-imino-1,5-dideoxysugars.

Before focusing on their mechanism of action, their inhibitory activity and importance

will be discussed.

OH
H O
H
OH O
gluconolactone
OH

NH

H
H
nojirimycin

OH OH
NH
HO 0
[m0 “mm
0“ on

S-amino-S-deoxy

gluconolactam B-D-glucosylmnine

OH OH
NH
H H a.
H H
OH OH

deoxynoynmycm castanospermine

Figure 1.1.2 Representatives of the most common classes of glycosidase inhibitors.

 

 

HO’

5\

Ho/S
Ht

HO
H

.,
“‘deox \.- fi

H
no N‘"
HO

swainsonine

OH

r1232;

glucal

F OH

2-deoxy-2-fluoro-g1ucoside

OH

HO O
H

OH

glucosyl methyl aryltriazene

Figure 1.1.2 (cont’d).

HO
NH

OH
OH

polyhydroxy-
pyrrolidine

OH
O

OH CH2

1 -deoxy-D-gluco-hept- 1 -enitol

conduritol epoxide

’N§N,Ar H

OH

HO NH
H 2
OH
aminocyclitol
OH
HO O
H
on F

glucosyl fluoride

NH

OH

aziridine

OH
O

OH

glucosyl isothiocyanate

1.2. A!
Chem
tamer
biolog
genera
meau

md5+

.0;

L5

 

1.2. Activity and importance of glycosidase inhibitors

Glycosidase inhibitors have important therapeutic effects in the treatment of diabetes,
cancer and AIDS. This beneficial activity derives from their interference with important
biological processes that are involved in these diseases. This subsection will describe the
general way glycosidases are involved in these diseases, however, it will focus only on

the activity of three classes of glycosidase inhibitors, which are: 1,5-5—1actams, -lactones

and 5-imino-l,5-dideoxysugars (Figure 1.2.1).

0 NH NH
HM R 54
O HO O HO

1 ,5-5-lactone 1 ,5-8-lactam 1,5-iminosugar

Figure 1.2.1 Three major classes of glycosidase inhibitors.

Diabetes

Carbohydrates are a main component of human food, 80-90% consisting of starch and
sucrose. In general more than 250 g of di- and polysaccharides must be enzymatically
split in the intestinal tract before they can be utilized by the organism. It has been
suggested that a pharmacological interference with the intestinal carbohydrate digestion
by suitable a-glucosidase inhibitors should be a feasible way to regulate and retard
carbohydrate digestion, control the rate of absorption of monosaccharides and by this

way influence the intermediary metabolism of the carbohydrates [3]. Glycosidases

catalyze the hydrolysis of complex saccharides and convert non-absorbable
carbohydrates into absorbable sugars. The rapid action of these enzymes leads to acute
undesirable elevations in blood glucose in diabetes. Potent inhibitors of these enzymes
prevent harmful hyperglycemic excursions of this type. It is desirable, however, that the
inhibition of these hydrolytic enzymes be limited to those present in the intestines.
Otherwise, inhibition of systemic glycohydrolases or glucose transport can lead to
difficulty in the utilization of intracellular carbohydrates as an energy-source and thus
cause metabolic problems. It has been demonstrated that iminosugars are active as
inhibitors of carbohydrate digestive enzymes and can be used in the treatment of diabetes.
More specifically, they can be used to prevent the development of hyperglycemia. Rather
than achieving this effect by promoting the metabolism of glucose present in the blood,
some 1,5-iminosugars like homonojirimycin glycosides 1 (where R is a glycosyl radical)
can act by preventing the initial formation of glucose in the body thereby holding down
the quantity of glucose that could appear in the blood [4] (Figure 1.2.2). An example is
the alpha-glucosidase inhibitor MDL-25637 2 produced by Aventis Pharrna and which is
in phase I clinical trials [5]. N-Hydroxyethyl deoxynojirimycin 3, commercialized by
Bayer under the name of Miglitol, was launched in 1998 as a substitute for the
antidiabetic Acarbose [6]. Since it has fewer gastrointestinal side effects, it is preferred.
Miglitol is used to treat both type 1 and type 2 diabetes. It inhibits the action of alpha
glucosidase and delays glucose absorption from the intestine, thereby preventing high
blood sugar levels. Camiglibose 4 is an antidiabetic and antiobesity agent. This long-
lasting alpha-glucosidase inhibitor is in preclinical trials [7]. Also in preclinical trials is

the alpha-glucosidase inhibitor N-methyl deoxynojirimycin 5, also named MGR-14 [8].

 

"""OH

 

"""0H

 

Figure 1.2.2 Aza-sugars possessing antidiabetic activity.

 

 

I. 0
"" \CH3

"""0H

CH3
0.]

 

 

[IS aCle’
considerr

been dlSt

Cancer
Cancer h
main thl
Chemotm
However.
I'mlxtssib:
decrease

dlSCOVEn

Acuteqrar

 

from reCO
that C0D tr

transfonm

Its actions are hypoglycemic and antithrombotic. Finally, Emiglitate 6 was also
considered as an antidiabetic. However, the clinical development of this compound has

been discontinued due to low efficacy and poor tolerability [9].

Cancer

Cancer has been for many years one of the most fatal deseases. After its detection, the
main therapeutic endeavors for its treatment were surgery, radiotherapy and
chemotherapy. Surgical operation and radiotherapy may eliminate primary tumors.
However, the success of these two procedures is many times hampered by the
impossibility of early prognosis of the disease. The chances for an efficient therapy
decrease once metastasis has started. This is why an intensified effort is applied for

discovery of drugs that can stop or reverse metastasis.

Acute-transforming retroviruses carry in their genomes specific oncogenes that arose
from recombination events between nontransforming viruses and normal cellular genes
that control growth and/or differentiation. In most of these retroviruses a single
transforming protein is synthesized from the viral oncogen, and this protein product is
responsible for initiation and maintenance of the transformed, cancerous state. About 30
acute-transforming retrovirus isolates have been reported so far, and these have been
found to harbor about 20 distinct oncogens. The frns, erbB, sis, and neu oncogenes are
known to have glycosylated expression products and the fins and erbB glycoproteins are
ultimately expressed in the plasma membrane. In general, the insertion of the cellular

gene within the viral genetic framework resulted in two critical alterations that permit a

10

 

normal
the vin
merprm
cellular
and qua

acute-tra

It has be
haVing
PTOCL‘SSir
Phenom?
and X47
PTOCessjn
leading h
be efffi‘ct:

DIOdUCtS.

 

 

normal cellular gene to function in an uncontrolled manner within infected cells. First,
the virus has provided a strong promoter for the cellular gene, enabling the
overproduction of the acquired cellular gene, and second, in almost all cases the acquired
cellular gene has been modified either through truncation or mutation. These quantitative
and qualitative changes are believed to be important in neoplastic transformation by

acute-transforming viruses.

It has been shown [10] that the cellular expression of acute-transforming retroviruses
having glycosylated expression products is interrupted by the administration of
processing glycosidase inhibitors, leading to remission of transformed cells to the normal
phenotype. In particular, administration of the glucosidase 1 inhibitors castanospermine 7
and N-methyl-l-deoxynojirimycin 8 (Figure 1.2.3) interrupts the glycosylational
processing and plasma membrane expression of the v-fms transforming glycoprotein,
leading to cancer remission. These and similar processing glucosidase inhibitors should
be effective in controlling cancers that are mediated by oncogenes with glycosylated
products. These inhibitors are also effective in suppressing the proliferation of cells that
normally express glycoprotein growth factor receptors on their surfaces. In particular,
castanospermine inhibits the proliferation of monocytes and macrophages that bear
glycosylated c-fins expression products on their plasma membranes, thereby affording a
measure of selective immunosuppresion for therapeutic effect. Also, it is expected that
the effectiveness of the existing various therapies could be further enhanced by inhibiting
metastasis of cancer cells. However, there are few substances showing inhibition of

metastasis as their main activity and none of them has been used in the clinical field.

11

 

 

Tsuruol
posses :
sodium

melanor

glucuror

 

 

 

Tsuruoka has found that alkylated and arylated 5-amino-5-deoxy gluconolactams 9
posses an excellent cancer metastasis inhibiting activity [11]. D-Glucaric acid 5—lactam
sodium salt 10 is an antimetastatic agent that has been found to inhibit invasion of B 16
melanoma cells. It is used in preclinical trials [12]. Its potassium analog 11 is a beta-

glucuronidase inhibitor with antinephrotoxic activity [13].

"""on

   

 

10 11

Figure 1.2.3 Compounds possessing cancer metastasis inhibitory activity.

12

AIDS

Acquired immunodeficiency syndrome (AIDS) is one of the most fatal disorders of our
century, for which no completely successful chemotherapy has been developed so far. As
a consequence, a great effort is being made to develop drugs and vaccines to combat
AIDS. The AIDS virus, first identified in 1983, has the capacity to reproduce within cells
of the immune system and thereby lead to a profound destruction of T2 T-cells (or CD4+
cells). The causative agent of AIDS is a retrovirus of the Lentiviridae family [14,15].
This enveloped single-stranded RNA virus is called human immunodeficiency virus
(HIV) [16,17] and two genetically distinct subtypes, HIV-1 and HIV-2, have been
characterized [18-20] of which the former has been found to be prevalent in causing the
disease. The HIV-1 infection, which targets monocytes expressing surface CD4
receptors, eventually produces profound defects in cell-mediated immunity [21]. Over
time, infection leads to severe depletion of CD4+ T-lymphocytes (T-cells), resulting in
opportunistic infections, neurologic and neoplastic diseases, and ultimately death.
Besides T-cells, other cells expressing DC4 on their surface may also harbor HIV-1
infection. These include macrophages, monocytes, and lymphoid cells [22].
Theoretically, an anti-HIV agent may exert its activity by inhibiting a variety of steps in
the life cycle of the virus. However, medicinal chemists have focused their attention
predominantly on the following stages: viral binding to target cells, virus cell fusion,
virus uncoating, reverse transcription of genomic RNA, viral integration, gene
expression, cleavage event, and virion maturation. By hitting any of these stages, the viral
replication can be terminated [23]. The replicative cycle of HIV-1 presents several viable

targets that could be exploited for the development of anti-HIV chemotherapy. Ideally, an

13

anti-HIV agent should arrest the virulence and further infection of healthy cells without
displaying toxicity toward normal cellular physiology. To achieve this goal, attention has
been focused on several intervention strategies and various kinds of inhibitors. Most of
the compounds approved so far by the FDA in the United States for the treatment of HIV

infections are reverse transcriptase inhibitors and protease inhibitors.

More recently, certain glucosidase inhibitors have been tested for activity against the
AIDS virus. Because the envelope glycoproteins of HIV are heavily glycosylated,
compounds that interfere with co- and posttranslational processing of glycoprotein gp120
and the transmembrane glycoprotein gp 41 may prevent viral entry into a cell.
Glucosidase cleaves off glucose units from the oligosaccharide chain and thus helps the
maturation of infectious virion. The inhibition of this enzyme, therefore, will lead to the
inhibition of virion maturation. Polyhydroxylated compounds (Figure 1.2.4) have
demonstrated inhibitory potential in preclinical evluation [24] and have been suggested as
potential anti-AIDS drugs [25-28]. Castanosperrnine 12, which is an alkaloid isolated
from the seeds of Australian chestnut tree, has been found to interfere with normal
glycosylation of HIV virions, thereby altering the envelope glycoprotein and preventing
entry of HIV into target cells. It also is studied as antimetastatic, treatment for multiple
sclerosis and for prevention of organ transplant rejection [29]. However, since
castanospermine is in limited supply due to its natural source and its chemical synthesis is
laborious, it is extremely expensive and therefore not a realistic candidate for a drug

urgently required on large scale. Additionally, castanospermine has demonstrated

14

16

Figure 1.2.4 Compounds possessing antiviral activity.

15

 

17

cytotoxicity at dose levels of 0.7 mg/mL [30]. A mono-butanoyl derivative 13 of
castanospermine named Celgosivir is also a potent anti—HIV agent and is tested in phase
11 clinical trials [31]. Alkaloid 14 called Kifimensine, produced by the actinomycete
Kitasatosporia kifunense is an immunomodulator used in preclinical trials [32]. A large
number of deoxynojirimycin analogs have been tested for anti-HIV activity. The
pentabutanoyl derivative 15 of DNJ is a potent inhibitor in phase II preclinical trials [33].
The N-nonyl-DNJ 16, which in 1992 was in preclinical trials, proved to be not only a
potent anti-HIV agent but also active against hepatitis B and hepatitis C [34]. N-methyl
deoxynojirimycin was also disclosed as having activity against HIV ostensibly based on
its glucosidase I inhibitory activity. However it was subsequently shown by Fleet [35]
that not all glucosidase 1 inhibitors are effective inhibitors of HIV. Therefore, some other
mechanism may be responsible for HIV inhibitory activity. The N-butyl derivative of
deoxynojirimycin has been found to have enhanced inhibitory activity against the HIV
virus at not-toxic concentrations compared to that exhibited by the corresponding N-
methyl and N-ethyl derivatives. N-Butyl-DNJ uniquely reduces the virus titer by over
five logaritms at non-cytotoxic concentrations whereas the N-methyl- and N-ethyl-DNJ
derivatives cause only a two to four log-order of reduction in the yield of infectious HIV.
As such, the N-butyl derivative (BuDNJ) has significant potential use for the treatment of
AIDS, and phase II clinical trials of this agent have been started. Also, long-term culture
of HIV-infected cells in the presence of BuDNJ gradually decreased the number of
infected cells to a point where infectious HIV was no longer detectable [36]. These data
suggest that even with proviral DNA integration, HIV -infected cells probably undergo a

lytic cycle and that long-term culture of infected cells in the presence of a drug such as

16

BuDNJ could break the cycle of HIV replication and infection. This gives hope that
compounds of this type might reduce or even eliminate HIV infection in vivo in patients
with latent infection or overt disease. Also, combinations of anti-HIV therapies which
include agents that attack the HIV replicative cycle at multiple sites offer several
advantages and synergistic combinations of AZT and BuDNJ have been found to give
excellent results [37]. Recently it has been discovered that N-butyl DNJ is also a
glycosphingolipid formation inhibitor and it is tested in the treatment for
glycosphingolipid storage disorders such as Gaucher’s disease and Fabry’s disease [38].
Based on all this information we can conclude that there is indeed a real interest in the
use of glycosidase inhibitors in the treatment of diabetes, cancer and AIDS. The
discovery of simple and efficient routes for the synthesis of such compounds would

increase their availability and make them viable drug candidates.

1.3. General mechanism of glycoside hydrolysis

Hen-egg lysozyme is still the only enzyme for which detailed mechanistic information is
available from X-ray structure analysis with respect to active-site structure and functional
groups involved in catalysis [39]. For all other glycosidases, especially the vast number
of exo-hydrolases, we depend on the interpretation of kinetic and inhibition studies with
reversible and irreversible inhibitors. Published data on the reversible inhibition by basic
sugar analogs, and active site-directed inactivation by conduritol epoxides, show that the
majority of “retaining” a— and B-glycosidases have an active site that features a proton-
donating group close to the glycosidic oxygen atom, and a carboxylate group that can

stabilize the partial positive charge developing on the anomeric carbon atom and that may

17

HO
HO

ROH

O
HO
.L.
I
A
a)
110%
o\\c,()e
|

 

 

 

N
r... r
O

HO

Scheme 1.3 General mechanism of glycosidase inhibitors.

18

possibly form a covalent glycosyl-enzyme intermediate that is subsequently hydrolyzed
(Scheme 1.3). It is proposed that substrate distortion towards the transition state (change
from 4C1 chair towards a half-chair conformation) constitutes a major contribution to
catalysis. However, the enzyme also adjusts its conformation to the requirements of the
ligand, rather than forcing a drastic conformational change upon the latter. It is also
assumed that access of solvent water to the active-site region of the enzyme-inhibitor
complex is largely restricted. This is important for the close alignment of the substrate

with respect to the catalytic groups that is required for effective catalysis.

1.4. Mechanism of action of glycosidase inhibitors

1 .4.1. Aldonolactones

The strong inhibition of glycosidases by aldonolactones was first mentioned in 1940 by
Japanese scientists who studied B-D-glucosidases from Aspergillus [40] and almonds

[41]. These studies were extended to other glycosidases by Conchie and Levvy [42]. The

OH
OH HO
H O 0
H ‘_—_> 0
OH O H 1‘

1,5-gluconolactone 1,4-gluconolactone

Scheme 1.4.1.1 Interconversion between 1,5—gluconolactone and 1,4-gluconolactone.

19

authors showed that the aldonic acids themselves are non-inhibitory, and that 1,5-lactones
are better inhibitors than their 1,4-isomers. Also, the 1,4-lactones are not stable and easily
convert to their 1,5-isomers (Scheme 1.4.1.1). The fact that the aldonolactones might
exert their powerful inhibition by virtue of their structural similarity with a glycosyl
oxocarbonium ion intermediate or a related transition state was first pointed out by
Leaback [43]. Both the lactone and the oxocarbonium ion have a trigonal, planar
configuration at C-1 and adopt a half-chair conformation which is in marked contrast
with the tetrahedral C-l configuration and 'C4 conformation of aldopyranoside substrates

and aldoses (Scheme 1.4.1.2). In addition to these geometrical factors, there have to be

110% Homo/R

R

Pyranosyl oxocarbenium ion B-Aldopyranoside
trigonal planar configuration 4C1 conformation

Aldono- l ,5-lactone
trigonal planar configuration

Scheme 1.4.1.2 Geometrical similarity between aldono-1,5-lactones and the pyranosyl

oxocarbenium ion and the differences in the conformation of aldopyranosides.

20

considered electrostatic interactions arising from the large dipole moment of the lactone
which would enhance lactone binding if there is a negatively charged group in close
proximity to C-1 of the bound inhibitor. There is a general agreement that the aldono-1,5-
lactones are better inhibitors for glycosidases than are the 1,4-isomers [44] which are
probably no better inhibitors than aldoses or polyols of comparable structure. It was
found that the (Jr-specific enzymes were inhibited by D-gluconolactone at least 100-fold
less potently than the B-specific ones. The latter are inhibited by aldono-1,5-lactones
several hundred to many thousand-fold better than by the corresponding aldoses. It has
been demonstrated that the inhibition is competitive; that is, the inhibitor competes with
the substrate for the free enzyme. Generally K; for lactones are in the micro- and even

nanomolar range.

1.4.2. 5-Amino—5—deoxylactams

The problems regarding ring size and stability that are encountered with aldonolactones
disappear when the ring—oxygen atom is replaced by an NH group. The resulting 5-
amino-S-deoxyglyconolactams constitute a new group of inhibitors, closely related to the
lactones. Like their oxygen analogs, they have a trigonal, planar configuration at C-1; the
dipole moment of the carbonyl group is expected to be even larger than that of the lactone
carbonyl, due to the larger contribution of the dipolar resonance structure (Scheme 1.4.2).
Differences in their interaction with glycosidases may arise if substrates and lactones are
bound with the ring-oxygen atom in a closely fitting cleft of the active site. The NH
group may then cause steric repulsion, or a hydrogen-bond donor for the ring-oxygen

atom may fail to interact properly with the NH group because of its amide resonance. The

21

 

 

 

strong
minor
comp:
enzyn
grate
orient.
Paulin
com]
the In
inhibi
Value:
From
glycc
res‘prj

[46.]

SChcml

strong inhibition observed with the lactams tested so far shows that these effects are of
minor importance. In cases where the inhibition constants for lactams and lactones can be
compared, they have the same order of magnitude. The weak inhibition of or-specific
enzymes by lactams has also been observed. The reason for this may be seen in the
greater structural similarity of lactams (orientation of the C=O dipole) with the
orientation of the anomeric oxygen atom of B-glycosides than of (Jr-glycosides. If
Pauling’s hypothesis [45] is followed about an evolution of enzyme-active sites towards
complementarity to the transition state, this would indicate a considerable resemblance of
the transition-state structure with that of the substrate. 5-Amino-5-deoxya1donolactams
inhibit B-D-glycosidases 100- to > 10,000-fold better than by the parent aldoses, with K,
values from 200 uM to < 0.1 M. The inhibition is competitive like the one for lactones.
From structural consideration it may be assumed that aldonolactams are bound by the
glycon binding-site of the enzyme, interacting with the same functional groups that are
responsible for substrate binding and, possibly, catalysis. A molecular modeling study

[46] performed by C.-H. Wong’s group indicated that the 2-OH group, the 6-OH group, a

4-
WC <-—> 0&0.
H H

Aldono- 1 ,5-lactam
trigonal planar configuration

Scheme 1.4.2 Configuration of aldono-1,5-lactams.

22

 

 

in?

to 11

gluc

3&1:

positive charge and a half-chair conformation are important for a good inhibitor binding

to the enzyme. Each of these factors contributes to the overall binding.

1.4.3. Cyclic iminosugars

In 1970, it was reported that nojirimycin, identified as 5-amino-5-deoxy-D-
glucopyranose, is a powerful inhibitor of B-D-glucosidases [47]. It had been discovered
by virtue of its antibiotic activity in culture filtrates of certain strains of Streptomyces
[48]. Its 1-deoxyderivative, a strong inhibitor of or-D-glucosidases, is produced by certain
strains of Bacillus [49]. Subsequent studies with 5-amino-5-deoxyhexopyranoses and 1,5-
dideoxy-l,5-imino-hexitols related to D-mannose [50], D-galactose [51] and 2-

acetamido-2-deoxy-D-glucose [52] showed that the strong inhibition of glycosidases by

on on
N
H H
OH O” OH

nojirimycin deoxynojirimycin

these compounds is a general phenomenon probably based on common mechanistic
features. These sugar analogs with nitrogen in the ring have much greater stability in
aqueous solution. No spontaneous decomposition reactions are known for the 1,5-
iminohexitols. The half-life at 25°C and pH 5 of the D-manno analog of nojirimycin is
larger than 100 h [50]. The large inhibitory potential of these basic sugar derivatives (in

the micro and nanomolar range) and their stability, have made them valuable tools not

23

only for mechanistic studies but also for the investigation of problems in cell biology
where glycosidases are involved. Irrespective of the individual variations, it can be stated
that sugar derivatives having an amino group in the ring inhibit glycosidases from several
hundredfold to more than ten thousand fold better than their oxygen analogs. This means
that the electrostatic interactions of protonated inhibitor and negatively charged group of
the active site are not impaired by the different position of the positive center in the N-
cyclic inhibitors. The formation of an ion pair consisting of the protonated inhibitor and
an anionic group at the active site can take place in two ways: (i) the enzyme binds the
neutral form of the inhibitor, which then gets protonated, probably by the same group
which protonates the glycosidic oxygen atom during substrate hydrolysis; (ii) the enzyme
binds the inhibitor cation, which then forms an ion pair with a carboxylate of the active
site (Figure 1.4.3). Both modes show a similar increase of inhibitory strength with pH,
because, with increasing pH, an increasing proportion of the inhibitor will be present in
the unprotonated form and ionization of an active-site carboxylic group will increase the
proportion of the carboxylate responsible for the tight binding of cationic glycosyl
derivatives. As there is a considerable overlap of pKa values for protonated inhibitors
(pKa 5.3 to 6.1 for glucosylamines [53], pKa 5.1 to 5.6 for S-amino-S-
deoxyhexopyranoses [46,49,50], and pKa 6.3 to 7.2 for l,S-didieoxy-l,5-iminohexitols)
[49,50,51] and pKa for carboxylic groups (pKa 4 to 6), it is not possible to discriminate
between the two modes of action by pH studies with basic inhibitors alone. A
straightforward discrimination between the two pathways is possible with enzymes that
are inhibited by basic glycosyl derivatives as well as by permanent cationic ones. This is

illustrated by B-D-glucosidases A3 form Asp. Wentii, where K, values for B-D-

24

 

 

glucosylpn
dcpendcnn
of the pro
ermine bi
carboxylic
carboxylat
acidic go
[33] and i
are bound
such isos
Eli‘cosylh
xname

PTOblt’m:

FlEUre 1,.

glucosylpyridinium ion, B-D-glucosylamine, and nojirimycin have the same pH
dependence, provided that the concentration of the last two is based on the concentration
of the protonated form [54]. The most plausible explanation in this case is that the
enzyme binds the inhibitor cation and the pH-dependence reflects the ionization of a
carboxylic group having pKa 5.6 which provides strong binding only when it is present as
carboxylate. Binding of the basic form of the inhibitor and subsequent protonation by an
acidic group of the active site probably take place with B-D-glucosidase from almonds
[53] and B-D-galactosidase from E. coli [55], because here the B-glycosylpyridinium ions
are bound no better than their uncharged analogs, the B-glycosylbenzenes. In cases where
such isosteric pairs of cationic and neutral inhibitors as glycosylpyridinium ion and
glycosylbenzene are not available, a permanently cationic derivative can be prepared by
N-permethylation of a basic sugar analog. Interpretation of inhibition results present no

problems if the quaternary ammonium ion is bound with affinity similar to that of the

B
i e
OH x02
e
H
OH ..... 9

Figure 1.4.3 Assumed ground-state binding of deoxynojirimycin to B-glucosidase.

25

parent to
mdhm
dwuNl
nnnd
not ohn
nhmin
he unn

menu

 

N-all‘yla:

 

“11h glUr

Who]

A 51111 e
111 [he 1‘1
“118 We
Cushlej
dmhn
Readyag
enlime
Inhibnm
ConcCnlr
to four ‘

parent compound. This was observed with an a-D-glucosidase of the endoplasmic
reticulum involved in glycoprotein biosynthesis (glucosidase I) which is inhibited by 1-
deoxy-N,N-dimethylnojirimycin, with K; 0.4 uM, whereas the monomethyl derivative
inhibited with K. 0.07 11M and l-deoxynojirimycin [56] with K. luM. Similar results
were obtained with lysosomal B-D-glucosidase from calf spleen and human placenta,
which were inhibited by l-deoxy-N,N-dimethylnojirimycin about 4-fold better than by
the unmethylated compound. These results also showed that enzyme-substrate
interactions can not be very close to the ring-oxygen atom of the substrate; if they were,
N-alkylation would have weakened the inhibition by steric interference, as it did [57]

with glucosidase 11, another or-D-glucosidase involved in glycoprotein biosynthesis and

with or-D-galactosidase from coffee beans [51].

A still enigmatic feature of glycosidase inhibition by sugar analogs with a nitrogen atom
in the ring is its slow onset in many cases, where K; is in the micromolar range or below.
This was first reported for B-D-glucosidase from almonds and nojirimycin by Grover and
Cushley [58], and for intestinal sucrase-isomaltase and nojirimycin, 1-
deoxynomirimycin, and acarbose, by Hanozet and coworkers [59]. The approach to the
steady-state inhibition took place on the time-scale of minutes. Studies with other
enzymes showed that this phenomenon is fairly widespread. In all cases, the enzyme-
inhibitor complex is formed at a rate that is of first order with respect to the inhibitor
concentration, with rate constants in the range [60-62] of 103 to 104 M'l's'l and thus three
to four orders of magnitude below those of the reactions controlled by the rate of

diffusion. As the association rate showed no saturation effects with inhibitor

26

concentrations up to 20 K, any loose, pre-steady-state complex must have a dissociation
constant K, at least 50-fold larger than the steady-state Ki. Inhibition constants for
intestinal sucrase measured after 2 and 10 s with nojirimycin, l-deoxynojirimycin, and
acarbose were found to be 134-, 59-, and 106-fold larger than those measured for 15
minutes. Inhibitors of this type were classified as slow, and slow, tight-binding inhibitors
by Morrison and Walsh [63]. Several possible models can be discussed for the molecular
basis of slow inhibition [63], but experimental evidence in support on one or the other is
still lacking for glycosidases. A reversible chemical reaction at the active site, for
example, formation of a cyclic imine 18 or a diffusion-controlled association with a trace
of imine in equilibrium with the 5-amino-5-deoxypyranose can be precluded, because
slow inhibition is also observed with l-deoxynojirimycin and its analogs and with

acarbose.

OH

H()' S E >
H
OH
18

Another possibility could be the association of the inhibitor with a high-affinity
conformer of the enzyme present in low, equilibrium concentration. Complex-formation
would then shift the conformational equilibrium towards the high-affinity state. In this
case, however, the rate constant should be independent of the inhibitor concentration,

because the high-affinity conformer would always be saturated with the inhibitor. A third

27

possibility is the formation of a loose complex, having a dissociation constant K, which
then undergoes a slow conformational change to form the tight complex. Reversibility of
the inhibition requires that the dissociation of the tight enzyme-inhibitor complex must

also be a slow process.

28

1.5. References

l.

2.

10.

ll.

12.

13.

14.

15.

16.

17.

18.

19.

Stryer, L. Biochemistry, Freeman, New York, 1988, pp 192-193.
Legler, G. Adv. Carbohydr. Chem. Biochem. 1990, 48, 319-384.

Truscheit, E.; Frommer, W.; Muller, L.; Junge, B.; Schmidt, D. D.; Wingender, W.
Angew. Chem. Int. Ed. Engl.1981, 20, 744-766.

Liu, P. S. US Patent 4,634,765, 1987.

Drug Data Report, 1987, 9(7), 603.

Drug Data Report, 1998, 20(9), 781.

Drug Data Report, 1992, 14(1), 46.

Drug Data Report, 1994, 16(1), 62.

Drug Data Report, 1986, 8(3), 276.

Rohrschneider, L. R.; Nichols, E. J. US Patent 4,83 7,23 7, 1989.

Tsuruoka, T.; Nakabayashi, S.; Fukuyasu, H.; Ishii, Y.; Tsuruoka, T.; Yamamoto, H.;
Inouye, S.; Kondo, S. US Patent 5,250,545, 1993.

Drug Data Report, 1993, 15(5), 487.

Drug Data Report, 1990, 12(2), 172.

Barre-Sinoussi, F.; Chermann, J .-C.; Rey, F.; Nugeyre, M. T.; Charnaret, S.; Gruest,
J .; Dauguet, C.; Axler-Blin, C.; Vezinet-Brun, F.; Rouzioux, C.; Rozenbaum, W.;

Montagnier, L. Science 1983, 220, 868-871.

Popovic, M.; Samgadharan, M. G.; Read, E.; Gallo, R. C. Science 1984, 224, 497-
500.

Gallo, R. C.; Montagnier, L. Sci. Am. 1988, 259, 40-48.

Coffin, J .; Hasse, A.; Levy, J. A.; Montagnier, L.; Oroszlan, S.; Teich, N.; Temin, H.;
Toyoshmia, K.; Varmus, H.; Vogt, P.; Weiss, R. Science 1986, 232, 697-697.

Levy, J. A. JAMA 1989, 261, 2997-3003.

Clavel, F .; Guyader, M.; Guetard, D.; Salle, M.; Montagnier, L.; Alizon, M. Nature
(London) 1986, 324, 691-695.

29

20. Guyader, M.; Emerrnan, M.; Sonigo, P.; Clavel, F.; Montagnier, L.; Alizon, M.
Nature (London) 1987, 326, 359-362.

21. Bowen, D. L.; Lane, H. C.; Fauci, A. S. Ann. Intern. Med 1985, 103, 704-709.

22. Embretson, J.; Zupancic, M.; Ribas, J. L.; Burke, L. Racz, P.; Tenner-Racz, K.;
Haase, A. T. Nature (London) 1993, 362, 359-362.

23. Garg, R.; Gupta, S. P.; Gao, H.; Babu, M. S.; Debnath, A. K.; Hansch, C. Chem. Rev.
1999, 99, 3525-3602.

24. Ratner, L.; vander Heyden, N.; AIDS Res. Human Retrov. 1993, 9, 1257-1262.

25. Sunkara, P. S.; Bowlin, T. L.; Liu, P. S. Biochem. Biophys. Res. Commun. 1987, 148,
206-210.

26. Tyms, R. Cancer 1987, 31, 1025-1026.

27. Walker, B. D.; Kowalski, M.; Goh, W. C.; Kozarski, K.; Krieger, M.; Rosen, C.;
Rohrschneider, L.; Sodroski, J. Proc. Natl. Acad. Sci. USA 1987, 84, 8120-8124.

28. Gruters, C. Nature 1987, 330. 74-77.
29. Drug Data Report, 1990, 12(6), 507.

30. Karpas, A.; Fleet, G. W. J .; Dwek, R. A.; Petursson, S.; Ramsden, N. G.; Jacob, G. S.;
Rohrschneider, L.; Sodroski, J. Proc. Natl. Acad. Sci. USA 1987, 84, 8120-8124.

31. Drug Data Report, 1991, 13(5), 429.

32. Drug Data Report, 1989, 11(11), 947.

33. Drug Data Report, 1994, 16(1), 100.

34. Drug Data Report, 1992, 14(3), 260.

35. Fleet, G. W. J. FEBS Lett. 1988, 237, 128-132.

36. Karpas, A.; Fleet, G. W. J.; Dwek, R. A.; Petursson, S.; Namgoong, K.; Ramsden,
N.G.; Jacob, G. S.; Rademacher, T. W. Proc. Natl. Acad. Sci. USA 1988, 85, 9229-
9233.

37. Hirsch, M. S.; Johnson, V. A. US Patent 5,011,829, 1991.

38. Drug Data Report, 1989, 11(10), 869.

30

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

S4.

55.

56.

57.

58

Blake, C. C. F.; Johnson, L. N.; Mair, G. A.; North, A. T. C.;Philips, D. C.; Sarma, V.
R. Proc. R. Soc. London, Ser. B 1967, 1 6 7, 378-385.

Ezake, S. J. Biochem. (Tokyo) 1940, 32, 104-111.

Hoikoshi, D. J. Biochem (Tokyo) 1942, 35, 39-42.

Conchie, J.; Levvy, G. A. Biochem. J. 1957, 65, 389-395.

Leaback, D. H. Biochem. Biophys. Res. Commun. 1968, 32, 1025-1030.
Levvy, G. A.; Snaith, S. M. Adv. Enzymol. 1972, 36, 151-181.

Pauling, L. Nature (London) 1948, 161, 707-710.

Kajimoto, T.; Liu, K. K.-C.; Pederson, R. L. Zhong, Z.; Ichikawa, Y.; Porco, J. A.;
Wong, C.-H. J. Am. Chem. Soc. 1991, 113, 6187-6196.

Niwa, T.; Inouye, S.; Tsuruoka, T.; Koaze, Y.; Niida, T. Agric. Biol. Chem. 1970, 34,
966-972.

Inouye, S.; Tsuruoka, T.; Niida, T. J. Antibiot. 1966, 19, 288-291.

Schmidt, D. D.; Frommer, W.; Muller, L.; Truscheit, E. Naturwissenschaften 1979,
66, 584-585.

Legler, G.; Julich, E. Carbohydr. Res. 1984, 128, 61-72.
Legler, G.; Pohl, S. Carbohydr. Res. 1986, 155, 119-129.
Kappes, E.; Legler, G. J. Carbohydr. Chem. 1989, 8, 371-388.
Legler, G. Biochim. Biophys. Acta 1978, 524, 94-101.

Legler, G.; Sinnott, M. L.; Withers, S. G. J. Chem. Soc. Perkin Trans. 1980, 2, 1376-
1383.

Legler, G.; Herrchen, M. Carbohydr. Res. 1983, 116, 95-103.

Schweden, J.; Borgrnann, C.; Legler, G.; Bause, E. Arch. Biochem. Biophys. 1986,
248, 335-340.

Liedtke, Thesis, University of Cologne, 1987.

Grover, A. K.; Cushley, R. J. Biochim. Biophys. Acta 1977, 482, 109-124.

31

59. Hanozet, H. P.; Pircher, H. P.; Vanni, P.; Oesch, B.; Semenza, G. J. Biol. Chem.1981,
256, 3703-3711.

60. Inouye, S.; Tsuruoka, T.; Ito, T.; Niida, T. Tetrahedron 1968, 23, 2125-2144.

61. Osiecki-Newman, K.; Fabgro, D. Legler, G.; Desnick, R. J.; Grabowski, G. A.
Biochim. Biophys. Acta 1987, 915, 87-100.

62. Donsimoni, R.; Legler, G.; pourbouze, R. Lalegrie, P. Enzyme 1988, 39, 78-79.

63. Morrison, J. F.; Legler, G.; Bause, E. Eur. J. Biochem. 1986, 15 7, 563-570.

32

Chapter 2

A General Approach to the Synthesis of Imino-Dideoxy and

-Trideoxy Alditols from B-D-Glyeosides

33

Abstract
Irnino-sugars (also called azasugars), a class of compounds of which the 1,5-dideoxy-1,5-
imino- and 1,5-imino-l,5,6-trideoxy alditols are members, are important glycosidase
inhibitors with very high potential as drugs. Their potential therapeutic applications range
from the treatment of diabetes to cancer and AIDS. We present here a general method for
the preparation of such compounds with the D-gluco and D-galacto configurations
starting from B-glycosides. The procedure is especially appealing because of its high
yield, stereoselectivity and straightforwardness. The key steps are the selective oxidation
of the glycosides to hexulosonic acids and reduction of the oxime derivatives to lactams
which are further reduced to the target compounds. The C6 position can be deoxygenated
during the reduction if it bears an acetoxy group. The trideoxy imino sugars are then
produced. This saves several steps in which a protection and deprotection strategy and the
installation of a halo or tosylate group at that position is avoided. Reduction of both the
E and Z forms of the oxime gives the correct D-stereochemistry. Deacetylation prior to

oxime reduction gives the dideoxy compounds.

34

2.1. Introduction

Over the last three decades there has been a continuous interest in natural and synthetic
imino-sugars because of their high potency as glycosidase inhibitors [1-18]. Glycosidases
catalyze the hydrolysis of glycosidic linkages and are the key enzymes in the degradation
of complex carbohydrates. One of their main metabolic roles is the conversion of
complex non-absorbable carbohydrates into absorbable mono- or oligosaccharides [19].
The rapid action of these enzymes can lead, however, to undesirable elevations in blood
glucose in diabetes. Iminosugars have been shown to act as glycosidase inhibitors and to
retard and regulate the intestinal carbohydrate digestion. They are therefore excellent
drug candidates for diabetes therapy [20]. N-Hydroxyethyl deoxynojirimycin,
commercialized under the name of Miglitol, has been launched in 1996 as a very potent
antidiabetic. [21]. An even more exciting potential use of iminosugars is in the treatment
of cancer and viral diseases [22]. It has been shown that modification of oligosaccharide
structures may alter metastatic capacity of cancer cells and l,5-diimino-l,5-
dideoxyglucitol (deoxynojirimycin) (1) [23] swainsonine (2) [24] and castanospermine
(3) [25] (Figure 2.1.1) can markedly inhibit metastasis of cancer cells. They might,
therefore, be used for the effective treatment of cancer. N-Butyl-deoxynojirimycin shows
excellent activity against herpes virus [26] whilst having low cyto-toxicity and no
inhibitory effect on the grth of normal cells. The greatest prospect for the use of
iminosugars as drugs is probably for the treatment of AIDS. Glycosidase inhibitors
prevent the processing of N-linked complex oligosaccharides. This results in the
disruption of the synthesis of viral coat glycoproteins such as the critical one called

gp120. This supposedly leads to the loss of recognition by the CD-4 receptor of the target

35

 

 

OH
NH
H
H
OH
1
Deoxynojirimycin (DNJ)
H0..."
3
Castanospermine
OH
NH
H
H
OH OH
5
Nojirimycin (NJ)

OH

,.... OH

'OH
2

Swainsonine

OH

N/BU
H
H

OH

N-Butyl-DNJ

NH

OH

6

l ,6-Dideoxynojirimycin

Fig. 2.1.] Iminosugars with therapeutic importance.

cell with concomitant reduction of syncytia formation resulting in the reduction of virus
infectivity and the inhibition of viral replication [27-29]. Clinical trials have been

launched for N-butyl-deoxynojirimycin (4) [30,31].

The iminosugars that have been the most investigated are deoxynojirimycin (DNJ) [32-
46] and its N-alkyl analogues [2,32,33,47]. l-Deoxynojirimycin was initially obtained
from nojirimycin (5) by catalytic hydrogenation or by reduction with NaBH4,
Nojirimycin was first described as an antibiotic produced by Streptomyces
roseochromogenes R-468 and Streptomyces lavendulae-SF-425 [48-50]. DNJ is produced
by strains of the genus Bacillus [51,52]. Nojirimycin (NJ) was found to be a potent
inhibitor of B—glucosidases (emulsin, fungal B-glucosidases) [53]. It also inhibits
microbial (rt-glucosidases, but to a much lesser degree than B-glucosidases.
Deoxynojirimycin is a much weaker inhibitor of emulsin than NJ but is a potent inhibitor
of trehalases and of Rhizopus niveus glucoamylase and exo-B-1-3-glucanase. NJ and DNJ
were found to be potent inhibitors of intestinal oligo- and disaccharidases of mammals
[51]. Nojirimycin and deoxynojirimycin are not the only potent glycosidase inhibitors.
Many other pyranoses and furanoses in which the ring oxygen has been replaced by an
amino group have biological activities. For example, l-deoxyrnannojirimycin inhibits 0L-
mannosidases and or-fucosidases [54-56], S-epi-L-deoxyrhamnojirimycin inhibits L-
rhamnosidase [57], deoxyrhamnojirimycin inhibits both lysosomal or-mannosidase and or-
fucosidase [58], deoxygalactonojirimycin inhibits galactosidases [59], and
deoxyfuconojirimycin is a powerful and specific inhibitor of several or-L-fucosidases

[60]. In addition to iminosugars, lactams have also been found to be good glycosidase

37

 

 

 

inhibitors, many times even better than the related azasugars. This may mean that the half
chair conformation of the lactams resembles more closely the structure of the transition
state for the enzymatic cleavage of glycosides. D-Mannolactam and D-rhamnolactam are

good inhibitors of or—mannosidase [61] and of apricot B-glucosidase [62].

It has been shown that alkylation of the ring nitrogen has different effects on different
glycosidases [63]. For in vivo inhibition, N-alkylation may facilitate transport of the
inhibitor across the cell membrane, thereby increasing its effectiveness. This might mean
that the more hydrophobic 6-deoxyiminosugars such as 1,5-imino-l,5,6-trideoxyglucitol
(6) (1,6-dideoxynojirimycin) (Figure 2.1.1) which in vitro experiments showed to be less
effective than its 1,5-dideoxy analogues [64], may be effective in vivo inhibitors. This is

still to be investigated.

We have been able to identify three chemical syntheses for 1,6-dideoxynojirimycin in the
literature. The first one involves enzymatic aldol condensation of dihydroxyacetone
phosphate (7) (DHAP) and enantiomerically pure 3-azido-2-S-hydroxypropanal (8),
using FDP-aldolase, followed by catalytic hydrogenation to give a 9:1 mixture of di-DNJ
(6) and its 05 epimer (13) [64] (Scheme 2.1.1). A key element used in this synthesis is
the preparation of the azidoaldehyde (8) as an optically pure compound. This is done by
peroxide oxidation of vinylaldehyde (11) followed by substitution with NaN; and
acetylation, to give a racemic mixture of azides that can be resolved with lipase to give an

approximatively equimolar mixture of the 2 enantiomers [65]. One disadvantage of this

38

FDP- 1d 1
PO\/u\/OH + H/u\/\N —>a 0 386

90%

PO

0 H ,Pd/C NH NH

OH N3 '2——" ”Om ”02%
950/ H H

° OH OH

 

 

OH OH
6 10
9
9 1
Pseudomonas
OEt 1) H202’ 2) NaN3 OAC lipase
N ’ N 3\/I\<OEt
>
013‘ OEt
75% 50%
11 12
OH QAC
N3\/'\<0Et + N3 § OEt
OEt Wore:
8 13
46% 47%

Scheme 2.1.1 Synthesis of di-DNJ by enzymatic aldol condensation.

39

method is the necessity of using flash chromatography after each reaction step. The
instability, high cost, difficulty in handling and narrow substrate specificity are other
drawbacks linked to the use of enzymes as synthetic catalysts [65]. The yield of the aldol
condensation reaction in the case of F DP-aldolases has been quite high, but this is not
consistent and yields of only 20-40% have been reported for rhamnulose-l-phosphate—

and firculose— 1 -phosphate aldolases [66].

The second method is based on an asymmetric Diels-Alder reaction of sorbaldehyde O-
methyloxime (14) or its dimethyl acetal (15) with a chiral chloronitroso derivative of
mannose (16) [67,68]. N-Protection, catalytic osmylation, configurational inversion using
the Sharpless method and then catalytic hydrogenolysis afford a 85:15 mixture of gluco-
gulo configuration in a 6% total yield (Scheme 2.1.2). The fuco, allo and tallo
configurations have also been prepared by slight modifications of this procedure. In
addition to the very low yield of this method, the use of toxic 0504 and ruthenium and

the multiple fractional separations does not make it appealing for large-scale synthesis.

The last approach uses protected 6-deoxy-5-keto-D-glucose (23) as substrate for a double
reductive amination [69] with benzhydrylarnine, NaCNBH3 and acetic acid in methanol at
—78°C. A diastereomeric mixture of gluco and ido piperidines in an 86:14 ratio was
obtained (Scheme 2.1.3). Benzylation, chromatographic separation followed by catalytic
hydrogenolysis affords 1,6-dideoxynojirimycin in a 20% total yield. The gulo and filCO

isomers have also been prepared. Disadvantages of this procedure are low yield, poor

40

R=NOMe, (OMe)2

14 15

Eya

o 0504, NMO
' ——>

N\
C023“

18 ———->

 

1) MeOH/CHCI3

 

2) BnCOCl

l7

 

of}

R
4
9 +
N4
CO2 Bn
0

l6 l7
éR
Ho...
., (')
JCM
HO E C0231)
Sharpless :
inversion
_____> 18
¢R
Ho
t?
.~‘ N\
H0“ C02 311
19
H2, Pd/C
19 ———->

 

gulo

Scheme 2.1.2 Synthesis of di-DNJ by asymmetric Diels-Alder reaction.

41

H H2COOEI CH3

 

   

O o O O
Bn m Bn DMSO,H20 O 0
———-—-—>
BF3 OEIZ 0 BIT
O 0 135 C
0% 0% 7C:
20 21 H
CH3 H NCHPh
2 2 Ho, = CH P
HCl,THF 0 0 NaCNBH3 N,CHzth U, 2 hz
——"> Bn OH ————-> +
AcOH,MeOH BnO :
OH -780C OH
23 24 86: 14 25
NaH
BnBr
“COONH4 ,CHzPhZ BnO.,,h * N,CH2P112
10% Pc/C £1
<———
MeOH 8110 E
OBn
total yield = 20%
ido
6 gluco
26 27

Scheme 2.1.3 Synthesis of di-DNJ by double reductive amination

42

stereoselectivity, harsh reaction conditions and toxic reagents, all of which is not
welcomed for large-scale production. In conclusion, although these three procedures of
synthesizing 1,5-imino-l,5,6-trideoxyalditols are important contributions, they generally
are not stereospecific for a given sugar derivative, require drastic reagents and/or reaction

conditions and the yields are below 20%.

The literature contains many syntheses for deoxynojirimycin and its analogues.
According to the methods that are involved they can be classified into: chemical,

chemical-microbiological and chemo-enzymatic syntheses.

Chemical synthesis

Many of the purely chemical methods that have been published involve many steps with
multiple protections and deprotections. They ofien utilize carbohydrates, lactones or other
chiral pool sources containing one or more chiral centers as starting materials. Yields are
generally low and multiple chromatographic separations are needed. An example of a
quite elaborate synthesis is the one published by Ikota in 1997 [39]. He used (S)-
pyroglutamic acid (28) as starting material. This is dihydroxylated using potassium
osmate and hydroquinidine-1,4-phthalazinediyl diether as a chiral ligand in the presence
of K3Fe(CN)6, K2CO3 and MeSOzNHz (Scheme 2.1.4). Protection of the alcohol groups
was followed by Cr03 oxidation and Grignard reaction to give a diastereomeric mixture
where the desired product was predominant. The major diastereomer was converted into

the corresponding MOM-ether and was treated with ozone followed by reductive work-

43

l) K0504, chiral ligand
K3Fe(CN)6

OT’ 2) MOMCl, base
HBoc
84%

M602

7

1) C103
—————>
HO on 2) CH2CHMgBr

HBoc CuBrz Megs
-78°C 51%

MOMO _OMOM

30

MOMO

l) MsCl, TEA

 

2) Kt-BuO, THF
70%

 

32

reduction

 

29

1) MOMCl
2) 03

on ———>
3) NaBH4

53%

MOMO _OMOM

  

H0

31

 

10% BC]

MeOI-I
80%

 

Scheme 2.1.4 Synthesis of DNJ from S-pyroglutamic acid.

0
Zn, NaCNBH3
——>
B 91%
OCH3
34
HgBr
S gNBn
Bn
+
OBn
36 3

l ) NaBH4'DW'02

2) H2, Pd/C

 

35

M

37

Scheme 2.1.5 Synthesis of DNJ by intramolecular aminomercuration.

up to give the alcohol in 53% yield. Mesylation followed by cyclization and deprotection
gave l-deoxynojirimycin in an 11% total yield. The disadvantages of this procedure are
the high number of reaction steps, low total yield, multiple separations, and the use of

toxic reagents. Other l-deoxyazasugars have also been synthesized by this procedure.

A different approach towards DNJ has been attempted by Ganem [46] who devised an
enantioslective synthetic route using intramolecular aminomercuration (Scheme 2.1.5). A
one-pot reductive ring opening and reductive amination of the pyranose were achieved by
heating tri-O-benzyl-6-bromopyranoside (34) with acid-washed zinc dust in 19:1
propanol-water containing benzylamine and NaBH3CN to afford an aminoalkene (35) in
91% overall yield. Reaction with mercuric trifluoroacetate in anhydrous THF gave a 3:2
mixture of bromomercurials afier ligand exchange with LiBr-THF. Following
chromatographic separation, the major cyclization product could be transformed to DNJ
by reductive oxygenation and hydrogenolytic deprotection. The total yield starting with
the bromobenzylglycoside was 35%, however the use of toxic cyanide and mercury

compounds excludes the procedure from large-scale synthetic considerations.

One method worthy of special mention is Behling‘s short and enantiospecific synthesis of
deoxynojirimycin starting from L-sorbose (38) [42]. This has been transformed to the
1,2-isopropyllidene sorbofuranose (39) and then selectively derivatized at the primary
position. Reduction, deprotection and catalytic hydrogenation leads to DNJ which has
been crystallized in a 17% total yield. The only drawback of this procedure is that it

cannot be extended to other sugars (Scheme 2.1.6).

46

 

OH OH 1) DMP o 0><
CHZOH —’H, H O
HOH 2) , I"120
2 HOH
0“ 2 0H
38 39
o 0>< o
o +
0 1) Ph P >< H
HO _1, Ho 0 —»
N3 2) H20 H N H N
H 2 OH 2
40 41
l Hz/Pd
OH
NH
H
H
OH

Scheme 2.1.6 Synthesis of DNJ from L-sorbose.

47

l) TIBSCl

TEA, Py
——>

2) NaN3
DMF

HO

OH

42

OH
OH

 

Chemo-microbiological synthesis

The chemo-microbiological method patented by Grabner [1,47] provides an elegant way
of transforming glucose into l-amino-l-deoxy-sorbitol (44), which is subsequently
transformed by bacteria into an amino hydroxyketone (45), that undergoes intramolecular
reductive amination to give DNJ in a good total yield of 45% (Scheme 2.1.7). The
extension of this method to deoxy sugars and various substituted sugars is limited by the
specificity of the organisms. So far, only mannose, allose, altrose, ribose and arabinose
could be used as starting materials, since gluconobacter selectively oxidizes the central of
three hydroxyl groups in D-erythro compounds. Since bacteria are being used, a

requirement of this procedure is avoidance of contamination of the reaction media.

 

 

 

CHO CHZN HBu HCl CHZNHBu
—OH bOH gluconobacter —OH
HO— BUNHZ, HCl HO-A OXidanS HO—
—OH H2, Pd/C —OH HOH
—-0H —0H pH=5 =
—-OH LOH —OH
43 44 45
1 H2, Pd/C
OH ,Bu
N
HO
HO
OH

Scheme 2.1.7 Chemo-microbiological synthesis of DNJ.

48

Chemo-enzymatic synthesis

The chemo-enzymatic approach has been extensively developed by C.-H. Wong and
involves aldolase promoted condensation reactions between DHAP (7) and a chiral N-
containing aldehyde (8) (Scheme 2.1.8). To date, three aldolases have been used to make
iminosugars and these are FDP (fructose-di-phosphate), Fuc-l-P (fiiculose-l-phosphate)
and Rha-l-P (rhamnulose-l-phosphate), the products being: D,L-gluco DNJ, D-galacto-
DNJ, D,L-manno-DNJ and D-talo-DNJ. In the case of FDP-aldolase and using a racemic
azide, a 1:4 mixture of gluco- and manna-DNJ has been obtained in a 59% yield based on
FDP. So far it is not clear whether the enzyme is available in sufficient quantity to allow

large-scale synthesis.

0 O OH OH
PO\/u\/0H 1)Pase
[KR/[WM 2) H2, Pd/C H0 NH
OH OH H OH
7
FDP
46 l
aldolase
——>
+
03 O OH 1)Pase OH
P0 ‘ 2 H ,Pd/C
EtO N3 MM )__2__, H OENH
H OH OH H
8 47

Scheme 2.1.8 Chemo-enzymatic synthesis of DNJ.

49

2.2. Results and discussions

As indicated in the previous chapter, most of the chemical methods for the preparation of
deoxynojirimycin, dideoxynojirimycin and their analogs are quite elaborate, require
multiple protection and deprotection steps, fractional separation, involve toxic reagents or
expensive chiral catalysts and provide low total yields. On the other hand, the alternative
microbiological and enzymatic methods need special care with respect to reaction
conditions, work-up and disposal of media. Finally, a common shortcoming to all these
procedures is their lack of generality. They are restricted to specific azasugars and there
is no single synthesis that can be applied for the preparation of all naturally occurring
imine carbohydrate analogs. Finding a general synthetic procedure would be extremely
appealing since the majority of N-sugar-analogs have glycosidase inhibitory action. Here
we present a general, simple and high yield new synthesis for iminosugars especially for
the 1,5,6-trideoxy analogs. The routes start from readily available B—glycosides and
require just a few steps with straightforward reaction conditions. They allow access to
both 1,5-dideoxy- and l,5,6-trideoxy—1,5-iminoalditols. The enantiochernistry does not
have to be induced but is derived from the starting sugar and is therefore always close to
100%. Hence the galacto- isomers can be obtained simply by using a galactoside. The
method also allows access to the S-amino-S-deoxy-aldonic acid 5—lactams. These are
excellent glycosidase inhibitors at concentrations 100 times lower than most of the other
inhibitors tested [64]. So far, the glucono-lactam has been made by the oxidation of
nojirimycin [64], a procedure that is quite expensive considering the high cost of
nojirimycin ($ 25 for 1 mg). The approach we use here is based on the chromium trioxide

oxidation of glycosides to yield ketoaldonic acid esters that are then converted to cyclic

50

iminosugars by reductive ring closure with a nitrogen species. Our research has focused

only on compounds with the D-gluco and D-galacto configurations.

The oxidation of the anomeric carbon of glycosides is well-documented in patents and
papers. Examples are the microbiological oxidation with gluconobacter oxidans [32, 47]
or the use of oxidants such as (Bu3Sn)2O [36] and RuO4 with or without partial protection
of the hydroxyl groups. The chromium trioxide oxidation we use here typically employs
a large excess of CrO3 [70] but we obtained excellent results using only 2 equivalents.
The peracetate of methyl-B-D-glucopyranoside 2a, on oxidation with chromium trioxide
gave the keto-ester 3a in quantitative yield (Scheme 2.2.1). Our first attempts at reductive
amination using NaCNBH3 and different amino-group sources such as NH4OH, NH4C1,
NH4OAc were unsuccessful. The use of the more nucleophilic hydroxylamine yielded the
oxime 4a, which crystallized as a mixture of cis- and trans-isomers in 95% yield.
Reduction of this oxime with hydrogen in the presence of palladium on carbon led to
formation of the 5-amino-5-deoxy derivative which spontaneously cyclized to form the
B—lactam. The hydrogenation conditions also led to reductive cleavage of the acetoxy
group to form a 6-deoxy fimction. This preceded the reduction of the oxime and allowed
access to the 1,5,6-trideoxyderivatives. Despite the presence of both the cis- and trans
oximes, no L-derivatives were formed. The desired isomer was formed exclusively.
Lactam 5a was formed quantitatively and was then reduced and deacetylated with
borane-THF, to afford the 6-dideoxynojirimycin 6a in 84.8% total yield, better than any

previously published synthesis.

51

OH
H O
H OMe
OH
la
OAc
0
Ac 0
Ac
OAc Me
3a
Ac NH
Ac
OAc 0
Sa

AC2O
Py

HzNOH
-——>

95%

BH3/THF
——>

95%

OAc

0 CrO
A 3
c OM e ———’
Ac AcOH

CM 100%
23
OAc OH
H
AC N, O 2
A Pd/C
c 0 Ac OMe
100%
4a

Scheme 2.2.1 Synthesis of l,5,6-trideoxy-1,5-iminoglucitol.

H OH
O
H OMe
OH
lb
Ac OAc
O
0
Ac Me
OAc
3b
Ac
NH
Ac
OAc 0
5b

AC2O
——>
Py

HzNOH
——->

FY
85%

BH,rrHF
——->

30%

A60 OAc

0
Ac OMe
OAc
2b
AcO 0 Ac
OH
N"
0
Ac Me
OAc
4b
H
NH
H
OH
6b

Scheme 2.2.2 Synthesis of 1,5,6-trideoxy-1,S-iminogalactitol.

53

C103
AcOH
l 00%

Pd/C

The same reaction sequence was applied to methyl-B-D-galactopyranoside lb. While the
first three reactions occurred with the same high yields, the catalytic reduction was
accompanied by partial deacetylation, making chromatographic separation necessary and
therefore decreasing the yield. Even with this inconvenience, this method affords a
convenient and high-yield route for preparing l,6-dideoxy-D-galactonojirimycin 6b
(Scheme 2.2.2). Access to the 6-hydroxy derivatives (deoxy-D-galacto and gluco-
nojirimycins) was readily achieved by deacetylating the oxime with hydrazine prior to
reduction. The deacylation yielded the acyl hydrazide 7a in quantitative yield (Scheme
2.2.3). Catalytic hydrogenation of the latter with Pd/C provided the 8—lactam 83. To
facilitate isolation, 8a was re-acetylated to 9a which was recovered in 35% total yield by

chromatography. This can be reduced with BH3/T HF, using the previously indicated

procedure to yield DNJ.
OAc OH
Ac N’OH N H N’OH H2
Ac 0 2‘4" HH 0 Pd/C
0
OAc OMe 100/0 0H NHNHZ
4a 711
CH OAc 0H
NH NH
Py AC H
OH 0 OAc 0 0H
83 93 10a

Scheme 2.2.3 Synthesis of gluconolactam and DNJ.

54

2.3. Conclusion

In summary, we have developed a facile and effective synthesis of 1,6-
dideoxynojirimycin, l,6-dideoxygalactonojirimycin and deoxynojirimycins as well as the
corresponding 5-amino-5-deoxy—aldonic acid 8—lactams using B-glycosides as starting
materials. B-Glycosides, in addition to the fact that they are relatively cheap and readily
available, have the advantage of containing the desired stereochemistry, making chiral
induction unnecessary. The synthesis is short and requires minimal or no
chromatographic separation, which accounts for the high yield and simplicity of the
procedure. The discovery that 6-acetoxy oximes can be deoxygenated is especially
important because the 6-deoxy-D-hexoses are extremely rare and alternative routes would
involve several protection and activation steps. This approach should be general to any [3-
glycoside, especially deoxy sugars and those with substituents such as ether groups, to
give general access to substituted or functionalized iminosugars. This is of great
importance since to date, there is no general procedure for synthesizing N-analogs of all
naturally occurring monosaccharides. The use of alkylamines instead of hydroxylamine

should give access to N-alkylated iminosugars.

2.4. Experimental section

General methods. Melting points were measured on a Fisher-Johns melting point
apparatus. Optical rotations were measured (2» = 589 nm) at room temperature using a
Perkin-Elmer 341 polarimeter in a 1 mL cell. The 1H and 13C NMR spectra were recorded

at 300 MHz on a Varian VXR spectrometer. The HRMS FAB mass spectra were

55

obtained using a JEOL HX-l 10 double focusing mass spectrometer operating in positive

ion mode.

Methyl 2,3,4, 6-tetra-O-acetyI-fl-D-glucopyranoside (2a)
The tetraacetate was prepared from methyl-B—D-glucopyranoside la, pyridine and acetic

anhydride according to standard procedures [71].

Methyl-2,3,4,6-tetra-O-acetyl-D-xylo-hex-5-ulos0nate (311)

To a solution of 2a (7.0 g, 19.3 mmol) in acetic acid (125 mL) and acetic anhydride (10
mL), CrO3 (3.8 g, 38.0 mmol) was added and the suspension was stirred at 50°C for 2
hours. The mixture was then poured slowly, with stirring into cold water (500 mL). The
water was extracted three times with chloroform. The combined chloroform layers were
decolorized with activated charcoal, filtered and washed with saturated NaHCO; solution
and then with water. After drying with Nast4 and rotary evaporation of the solvent the
ketone 3a (7.2 g, 100%) was obtained as a colorless syrup: [0t]23D -2.4° (c 1.67,CH3C1)
lit.-5.8°, (c 1.65, CH3C1) [21]; lH NMR(CDC13)8 2.08 (s, 3 H, OAc), 2.08 (s, 3 H, OAc),
2.10 (s, 3 H, OAc), 2.13 (s, 3 H, OAc), 3.70 (s, 3H, OCH3), 4.73 (d, l H, J6”), 17.3 Hz,
H-6a), 4.84 (d, 1 H, H-6b), 5.26 (d, 1 H, .123 4.4 Hz, H-2), 5.43 (d, 1 H, J3). 4.4 Hz, H-4)
5.65 (t, 1 H, H-3); l3'C NMR (CDC13) 5 20.1, 20.2, 20.3, 52.8, 66.4, 69.3, 69.6, 73.6,

167.0, 169.1, 169.3, 169.7, 197.0.

56

 

 

 

Methyl 2,3,4,6-tetra-O-acetyl—D-xylo-hex-S-ulosonate oxime (4a)

The ketone 3a (7.0 g, 18.6 mmol) was dissolved in pyridine (16 mL) and the solution
cooled to 0°C. Hydroxylamine hydrochloride (2.0 g, 28.7 mmol) was then added and the
solution stirred at 0°C for 15 minutes and then for another 2 hours at room temperature.
The mixture was poured onto ice and water and then extracted three times with
chloroform. The combined chloroform layers were subsequently washed with water,
dried with NaZSO4 and evaporated. Crystallization from hot ethanol afforded white
crystals of the oxime 4a (6.9 g, 95%) as a 3:2 mixture of cis-trans isomers: Isomer 1: 1H
NMR (CDC13) 5 1.93 (s, 3 H, OAc), 1.94 (s, 3 H, OAc), 2.00 (s, 3 H, OAc), 2.01 (s, 3 H,
OAc), 3.56 (s, 3H, OCH3), 4.36 (d, 1H, Jbabb 12.4 Hz, H6-a), 4.72 (d, 1H, H6-b), 4.99 (d,
1H, J3,4 2.6 Hz, H-4), 5.72 (dd, 1H, J33 7.8 Hz, H-3), 6.28 (d, 1H, H-2); l3C NMR(CDC13)
5 20.5, 20.4, 52.8, 61.3, 66.1, 69.5, 69.8, 149.9, 167.3, 169.4, 169.5, 170.1; HRMS
(M+H+) calcd. 392.1193, found 392.1198. Isomer 2: mp 121-122°C; lH NMR(CDC13) 5
1.88 (s, 3 H, OAc), 1.89 (s, 3 H, OAc), 1.98 (s, 3 H, OAc), 2.00 (s, 3 H, OAc), 3.56 (s,
3H, OCH3), 4.82 (s, 2H, H-6), 5.16 (d, 1H, J14 2.6 Hz, H-4), 5.62 (d, 1H, J33 8.5, H-2),
5.78 (dd, 1H, H-3); ”C NMR (CDC13) 5 20.5, 20.4, 52.8, 61.3, 66.1, 69.5, 69.8, 149.9,

167.3, 169.4, 169.5, 170.1.

2,3,4—Tri-O-acetyl-S-amino-S,6-dide0xy-D-gluc0n0-1,5-Iactam (5a)

A solution of 4a (6.9 g, 17.6 mmol) in glacial acetic acid (275 mL), containing 10% Pd/C
(2.7 g) was hydrogenated in a Parr reactor under a H2 pressure of 300-400 psi for 40
hours at 55°C. The reaction mixture was filtered through Celite and washed with ethanol.

The solvent was rotary-evaporated and the lactam 5a (5 g, 100%) was obtained as a light

57

yellow syrup: [611231) +70.0° (c 1.56, CHC13); 'H NMR (CDC13) 5 1.11 (d, 3H,.15,6 6.3 Hz,
H-6), 1.94 (s, 3 H, OAc), 1.98 (s, 3 H, OAc), 2.00 (s, 3 H, OAc), 3.51 (m, 1H, J25 9.7,
Hz, H-S), 4.94 (t, 1H, J2,3=J3,4 9.7 Hz, H-3), 4.96 (d, 1H, H-2), 5.40 (t, 1H, H-4); 13C
NMR (CDC13) 6 18.0, 20.3, 20.3, 48.7, 70.6, 70.9, 71.4, 166.7, 169.4, 169.6, 169.8;

HRMS (M+H+) calcd. 288.1083, found 288.1089.

1 ,5,6-trideoxy— l ,5-imino-D-glucitol (6a)

1M BH3/T HF (50 mL, 50.0 mmol) was added under N2 to a solution of 5a (5.0 g, 17.4
mmol) in THF (33 mL). The mixture was stirred at room temperature for 1.5 hours and
then refluxed for another 1.5 hour. After cooling to room temperature 9% methanolic
HCl (40 mL) was carefully added and the resulting solution was refluxed for 30 minutes.
The THF was removed by rotary evaporation and the reaction mixture was dissolved
repeatedly in methanol, followed by evaporation to remove borates. Water was added to
the dry crude product and the solution was passed through an anion exchange resin
(Amberlite IR-45 OH-form) and then dried on the rotary evaporator. To remove the last
traces of borates, a solution of 1M NaOH (15 mL) and methanol (6 mL) were added to
the crude product and the mixture was stirred overnight at room temperature. The
methanol was evaporated and the aqueous solution was lyophylized. A methanolic HCl
solution was added, which precipitated NaCl while the methanolic solution was dried, to
give product 6a (2.4 g, 95%): [61231) +155o (c 1.88, H20), lit. +13° (c 1.0, H20) [18]; 'H
NMR (D20) 5 1.25 (d, 3H, J56 6.3 Hz, H-6), 2.77 (dd, 1H, J18], 12.4 Hz, J13; 11.7 Hz, H-

la), 3.02 (dd, 1H, J45 10.0 Hz, H-5), 3.23 (dd, 1H, J3,4 9.0 Hz, H-4), 3.33 (dd, 1H, Jig; 5.1

58

Hz, H-le), 3.31 (dd, 1H, 12,3 9.2 Hz, H-3), 3.63 (ddd, 1H, H-2); l3c NMR (D20) 5 17.5,

49.5, 55.2, 71.4, 76.7, 79.0.

Methyl 2,3,4, 6—tetra-O-acetyl-fl-D-galactopyranaside (2b)
This was prepared from methyl-B-D-galactopyranoside hydrate lb (5.0 g, 24.6 mmol)

using the same procedure described for the gluco-compound.

Methyl 2,3,4,6-tetra-O-acetyI-L-arabino-hex-5-ulosonate (3b)

This was prepared from 2a using the same procedure as described for 311 except that the
reaction time was 3 hours. Yield is 100%: [on]23 D -24.3° (c 1.4, CHC13), lit.-12.0°, (c 2.8,
CHC13) [21]; 1H NMR (CDC13) 5 2.06 (s, 3 H, OAc), 2.10 (s, 3 H, OAc), 2.11 (s, 3 H,
OAc), 2.16 (s, 3 H, OAc), 3.71 (s, 3H, OCH3), 4.80 (d, 1H, .1635], 17.5 Hz, H-6a), 4.92 (d,
1H, H-6b), 5.20 (d, 1H, J23 8.7 Hz, H-2), 5.27 (d, 1 H, J3); 2.4 Hz, H-4), 5.66 (dd, 1 H, H-
3); 13C NMR (CDC13) 5 20.1, 20.2, 20.3, 20.4, 52.8, 67.1, 69.3, 69.5, 71.0, 167.0, 168.9,

169.4, 169.8, 198.1.

Methyl-2,3,4,6-tetra-O-acetyl-L-arabino-hex—S-ulosonate oxime (4b)

This was prepared from the ketone 3b as described for the corresponding D-xylo
compound. White crystals of 4a (85%) were obtained as a mixture of cis-trans isomers:
1H NMR (CDC13) 5 isomer 1: 1.98 (s, 3 H, OAc), 2.01 (s, 3 H, OAc), 2.08 (s, 3 H, OAc),
2.15 (s, 3 H, OAc), 3.70 (s, 3H, OCH3), 4.82 (d, 1H, J63“, 14.6 Hz, H6-a), 5.11 (d, 1H,

H6-b), 5.35 (d, 1H, J14 1.9 Hz, H-4), 5.68 (d, 1H, J33 9.0, Hz, H-2), 5.84 (dd, 1H, H-3);

59

l3C NMR (CDCl3) 5 20.2, 20.3, 20.4, 20.5, 52.6, 56.4, 68.7, 69.2, 69.6, 149.9, 167.5,

168.9, 169.3, 170.0, 170.3.

1,5,6-trideoxy-1 ,S-imino-D-galactitol (6b)

This was prepared from 4b (7.4 g, 18.9 mmol) as described for the corresponding gluco
compound 5a. 7.4 g of a crude product was obtained which was subjected to borane
reduction. The product was isolated as described for the gluco isomer. Flash column
chromatography using a chloroform-methanol (6:1) mixture gave the product 6b (1.5 g,
30%): [(1.]st +27.0° (c 1.3, CHC13) , lit. +490o (c 1, CHC13) [20]; 1H NMR (1320) 5 1.21
(d, 3H, .15], 6.6 Hz, H-6), 2.73 (t, 1H, Jlajc, Jla,2 11.9 Hz, H-la), 3.30 (dd, 1H, J13; 5.4 Hz,
H-le), 3.37 (m, 1H, H-S), 3.50 (dd, 1H, J23 9.6 Hz, J24 3.1 Hz, H-3), 3.90 (d, 1H, J45 3.1

Hz, H-4), 3.91 (ddd, 1H, H-2); l3c NMR (D20) 5 14.2, 46.1, 55.0, 64.4, 69.9, 73.1.

2, 3, 4, 6- Tetra-O-acetyl-S-amino-S-deoxy—D-glucono-1,5-lactam (9)

The acetylated oxime 4a (1.5 g, 3.8 mmol) was deacetylated with concomitant
conversion to the acyl hydrazide by treatment with anhydrous hydrazine (0.75 mL, 23.8
mmol) in methanol (15 mL) at room temperature for 2 hours. Evaporation of the solvent
gave crude 711: 1H NMR (D20) 5 4.18 (1H, dd, J4.6 Hz, J7.0 Hz) 4.51 (1H, (1, J65 Hz),
4.43 (1H, d, J 14.9 Hz), 4.53 (1H, d, J 14.8 Hz), 5.18 (1H, d, J4.6 Hz); 130 NMR (D20)
5 61.1, 69.1, 73.4, 73.5, 160.7, 173.4.

73 was hydrogenated in glacial acetic acid with 10%, Pd/C (0.4 g) at 50°C and 300 psi
pressure of H2 for 2 days. After filtration through Celite, the solution was dried on the

rotary evaporator and the crude product acetylated with acetic anhydride (15 mL) and

60

Ii

 

 

 

pyridine (15 mL) for 5 hours at room temperature. The mixture was poured into cold
water and extracted with chloroform. The chloroform layer was dried with Na2SO4.
Evaporation of the solvent gave crude product (1.4 g), which was subjected to flash
chromatography on silica (eluent hexane-acetone = 2: 1) to give the peracetylated lactam
9a (0.5 g, 34% total yield from 4a) and its C-S epimer: 9a: mp=177-178°C; [0t]23p +88.6°
(c 1.11, CHC13), lit. +104° (c 1.73, CHC13) [17]; 1H NMR (CDC13) 5 2.03 (s, 3 H, OAc),
2.06 (s, 3 H, OAc), 2.08 (s, 3 H, OAc), 2.10 (s, 3 H, OAc), 3.75 (ddd, 1H, J45 9.7 Hz, J5,6a
2.9 Hz, J 5,6b 6.5 Hz, H-5), 3.96 (dd, 1H, J6a,6b 11.7 Hz, H6-b), 4.22 (dd, 1H, H-6a), 5.06
(d, 1H, J32 9.5 Hz, H-2), 5.20 (t, 1H, J24 9.5 Hz, H-3), 5.53 (dd, 1H, H-4), 6.48 (s, 1H, 3,
NH) ;13C NMR(CDC13) 5 20.5, 20.5, 20.5, 20.6, 52.4, 62.7, 67.2, 70.4, 70.5, 166.2, 169.4,
169.6, 170.0, 170.4. HRMS (M+H*) calcd. 346.1060, found 346.1143. 9:1-epimer: [011230
+3.1° (c 1.81, CHC13); 'H NMR (CDC13) 1.98 (s, 3 H, OAc), 1.99 (s, 3 H, OAc), 2.00 (s,
3 H, OAc), 2.02 (s, 3 H, OAc), 3.88 (1H, m, H-5), 4.04 (dd, 1H, J63’6b 11.4 Hz, Jim, 6.3
Hz, H6-b), 4.18 (dd, 1H, J16, 3.9 Hz, H-6a), 5.15 (dd, 1H, J45 9.5 Hz, J24 7.5 Hz, H-4),
5.15 (d, 1H, J23 7.5 Hz, H-2), 5.39 (t, 1H, H-3), 7.27 (1H, s, broad, NH); 13C NMR

(CDC13) 5 20.2, 20.3, 20.4, 50.0, 62.0, 68.0, 69.8, 70.0, 166.7, 169.3, 169.7, 170.3, 170.6.

61

2.5. References

1.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

Grabner, R. W.; Landis, B. H.; Scaros, M. G.; Rutter, R. J. US Patent 5,695,969,
1997.

Boshagen, H.; Junge, B.; Paessens, A.; Schuller, M. US Patent 4,940, 705, 1990.
Shilvock, J. P.; Wheatley, B. D.; Nash, T. J.; Griffiths, R. C.; Jones, M. J.; Muller,
M.; Crook, S.; Watkin, D. J .; Smith, C.; Besra, G. S.; Brennan, P. J.; Fleet, G. W. J.
Tetrahedron Lett., 1996, 37, 8569-8572.

Rajanikanth, D. B.; Szarek, A. J. J. Chem. Soc. Perkin Trans. I 1992, 2151-2152.
Hussain, A.; Wyatt, P. B. Tetrahedron, 1993, 49 2123-2130.

Defoin, A.; Sarazin, H.; Streith, J. Tetrahedron Lett. 1993, 34, 4327-4330.

Defoin, A.; Sarazin, H.; Streith, J. Tetrahedron 1997, 53, 13769-13782.

Defoin, A.; Sarazin, H.; Strehler, C.; Streith, J. Tetrahedron Lett. 1994, 35, 5653-
5656.

Fleet, G. W. J .; Ramsden, R. J .; Witty, D. R. Tetrahedron Lett. 1988, 29, 2871-2874.
Fleet, G. W. J .; Ramsden, R. J .; Witty, D. R. Tetrahedron 1989, 45, 327-336.
Takahashi, S.; Kuzuhara, H. Chem. Lett. 1992, 21-24.

Takahashi, S.; Kuzuhara, H. J. Chem. Soc., Perkin Trans. 1, 1997, 607-612.

Hendry, D.; Hough, L.; Richardson, A. C. Tetrahedron Lett. 1987, 28, 4597-4600.
Hendry, D.; Hough, L.; Richardson, A. C. Tetrahedron Lett. 1987, 28, 4601-4604.
Straub, A.; Effenberger, F.; Fischer, P. J. Org. Chem. 1990, 55, 3926-3932.

Delinck, D. L.; Margolin, A. L. Tetrahedron Lett. 1990, 31, 3093-3096.

Look, G. C.; Fotsch, C. H.; Wong, C.-H. Acc. Chem. Res. 1993, 26, 182-190.

Kajimoto, T.; Chen, L.; Liu, K. K.-C.; Wong, C.-H. J. Am. Chem. Soc. 1991, 113,
6678-6680.

Truscheit, E.; Frommer, W.; Junge, B.; Muller, L.; Schmidt, D. D.; Wingender, W.
Angew. Chem. Int. Ed. Engl. 1981, 20, 744-761.

62

20.
21.
22.

23.

24.

25.

26.

27.

28.

29.

30.
31.
32.
33.
34.
35.
36.
37.

38.

Liu, P. S. US Patent 4,634, 765, 1987.
Minter, B. Decision Resources 1999, 1-24.
Rohrschneider, L. R.; Nichols, E. J. US Patent 4,837,237, 1989.

Tsuruoka, T.; Nakabayashi, W.; Fukuyasu, H.; Ishii, Y.; Tsuruoaka, T.; Yamamoto,
H.; Inouye, S.; Kondo, S. US Patent 5,250,545, 1993.

Dennis, J. W. Cancer Res. 1986, 46, 5131-5136.

Humphris, M. J.; Matsumoto, K.; White, S. L.; Olden, K. Cancer Res. 1986, 46,
5215-5312.

Jacob, G. S.; Tyms, S. A.; Rademacher, T. W.; Dwek, R. A. US Patent 4,957,926,
1990.

Walker, B. D.; Kowalski, M.; Goh, W. C.; Kozarsky, K.; Krieger, M.; Rosen, C.;
Rohrschneider, L.; Haseltine, W. A.; Sodroski, J. Proc. Natl. Acad. Sci. USA 1987,
8120-8124.

Karpas, A.; Fleet, G. W. J .; Dwek, R. A.; Petursson, S.; Namgoong, S. K.; Ramsden,
N. J.; Jacob, G. S.; Rademacher, T. W. Proc. Natl. Acad. Sci. USA 1988, 85, 9229-
9233.

Fleet, G. W. J.; Karpas, A.; Dwek, R. A.; Fellows, L. D.; Tyms, A. S.; Petursson,
S.;Namgoong, S. K.; Ramsden, N. G.; Smith, P. W.; Son, J. C.; Wilson, P.; Witty,
D. R.;Jacob, G. S.; Rademacher, T. W. FEBS Lett. 1988, 23 7, 128-132.
Rohrschneider, L. R. US Patent 5,643,888, 1997.

Hirsch, M. S. US Patent 5,011,829, 1991.

Schroder, T.; Stubbe, M. US Patent 4,806,650, 1989.

Koebernick, W. US Patent 4,611,058, 1986.

Anzeveno, P. B.; Creemer, L. J. US Patent 5,227,479, 1993.

Anzeveno, P. US Patent 4,908,439, 1990.

Tsuda, Y.; Okuno, Y.; Kanemitsu, K. Heterocycles, 1988, 27, 63-66.

Inouye, S.; Tsuruoka, T.; Ito, T.; Niida, T. Tetrahedron, 1968, 23, 2125-2144.

Vasella, A.; Voeffray, R. Helv. Chim Acta 1982, 65, 1 134-1 144.

63

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

Ikota, N.; Hirano, J .; Gamage, R.; Nakagawa, H.; Hama-Inaba, H. Heterocycles
1997, 46, 637-643.

Paulsenm, H.; Sangster, 1.;Heyns, K. Chem. Ber. 1967, 100, 802-815.

Rudge, A. J .; Collins, 1.; Holmes, A.B.; Baker, R. Angew. Chem. Int. Ed. Engl. 1994,
33, 2320-2322.

Behling, J.; Farid, P.; Medich, J. R.; Scaros, M. G.; Prunier, M. Synth. Commun.
1991, 21, 1383-1386.

Kinast, G.; Schedel, M. Angew. Chem. Int. Ed. Engl. 1981, 20, 805-806.
Pederson,R. L.; Kim, M.-J.; Wong, C.-H. Tetrahedron Lett. 1988, 29, 4645-4648.

Osten, C. H.; Sinskey, A. J .; Barbas, C. F.; Pederson, R. L.; Wang, Y.-F.; Wong, C.-
H. J. Am. Chem. Soc. 1989, 111, 3924-3927.

Ganem, B. Acc. Chem. Res. 1996, 29, 340-347.

Grabner, R. W.; Landis, B. H.; Wang, P. T.; Prunier, M. L.; Scaros, M. G. US Patent
5,610,039, 1997.

Inouye, S.; Tsuruoka, T.; Niida, T. J. Antibiot. 1966, 19, 288-292.
Paulsen, H.; Sangster, 1.; Heyns, K. Chem. Ber. 1967, 100, 802-812.
Inouye, S.; Tsuruoka, T.; Ito, T.; Niida, T. Tetrahedron 1968, 24, 2125-2144.

Schmidt, D. D.; Frommer, W.; Muller, L.; Truscheit, E. Naturwissenschafien 1979,
66, 584-593.

Frommer, W.; Muller, L.; Schmidt, D. D.; Puls, W.; Krause, H. P.; Herber, U. DOS
2,658,563, 1978.

Niwa, T.; Inouye, S.; Tsuruoka, T.; Koaze, Y.; Niida, T. Agric. Biol. Chem. 1970,
34, 966-968.

Truscheit, E.; Frommer, W.; Junge, B.; Muller, L.; Schemidt, D. D.;Wingender, W.
Angew. Chem. Int. Ed. Engl. 1981, 20, 744-761.

Fellows, L. E. Chem. Ber. 1987, 23, 842-844.

Straub, A.; Effenberger, F.; Fischer, P. J. Org. Chem. 1990, 55, 3926—3932.

64

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

71

Shilvock, J. P.; Wheatley, J. R.; Davis, B.; Nash, R.; Griffiths, R. C.; Jones, M. G.;
Muller, M.; Crook, S.; Watkin, D. J .; Smith, C.; Besra, G. S.; Brennan, P. J .; Fleet,
W. J. Tetrahedron Lett, 1996, 37, 8560-8572.

Fairvanks, A. J .; Carpenter, N. C.; Fleet, G. W. J.; Ramsden, N. G.; de Bello, C. 1.;
Winchester, B. G.; Al-Daher, S. S.; Nagahashi, G. Tetrahedron, 1992, 48, 3365-
3376.

Paulsen, H.; Matzke, M. Liebigs Ann. Chem. 1988, 1121-1126.

Fleet, G. W. J.; Petursson, S.; Campbell, A. L.; Mueller, R. A.; Behling, J. R.;
Babiak, K. A.; Ng, J. S.; Scaros, M. G. J. Chem. Soc. Perkin Trans. 1 1989, 665-666.

Tschamber, T.; Rodriguez-Perez, E.-M.; Wolf, P.; Streith, J. Heterocycles, 1996, 42,
669-676.

Niwa, T.; Tsurouka, T.; Goi, H.; Kodama, Y.; Itoh, J.; Inouye, D.; Yamada, Y.;
Niida, T.; Nobe, M.; Ogawa, Y. J. Antibiot. 1984, 37, 1579-1586.

Winchester, B.; Barkere, C.; Baines, S.; Jacob, G. S.; Namgoong, S. K.; Fleet,
G.;Biochem. J. 1990, 256, 277-282.

Kajimoto, T.; Kiu, K. K.-C.; Pederson, R. L.; Zhong, Z.; Ichikawa, Y.; Porco, J. A.;
Wong, C.-H. J. Am. Chem. Soc. 1991, 113, 6187-6196.

Wong, C.-H.;Halcomb, R. L.; Ichikawa, J.; Kajimoto, T. Angew. Chem. Int. Ed.
Engl. 1995, 34, 412-432.

Liu, K. K.-C.; Kajimoto, T.; Chen, L.; Zhong, Z.; Ichikawa, Y.; Wong, C.-H. J. Org.
Chem. 1991, 56, 6280-6289.

Defoin, A.; Sarazin, H.; Streith, J. Helv. Chim. Acta 1996, 79, 560-567.
Defoin, A.; Sarazin, H.; Streith, J. Tetrahedron, 1997, 53, 13783-13796.
Dhavale, D. D.; Saha, N. N.; Desai, V. N. J. Org. Chem. 1997, 62, 7482-7484.

Angyal, S. J.; James, K. Aust. J. Chem. 1970, 23, 1209-1221.

.Wolfrom, M. L.; Thompson, A. Methods in Carbohydrate Chemistry; Academic

Press Inc., 1963; pp 211—215.

65

Chapter 3

A Preparation and Screening Strategy for Glycosidase Inhibitors

66

Abstract
Glycosidases play important roles in numerous biological processes. Inhibitors of
glycosidases have the potential of producing several beneficial therapeutic effects.
However, their synthesis is not straightforward and their inhibitory activity can be
estimated only by testing them one by one, which proved to be laborious and time-
consuming. Finding a method of synthesis by which a large spectrum of compounds can

be produced and tested would be extremely advantageous.

Here we present a one-pot procedure that we have employed in the synthesis of a library
0f compounds belonging to 4 classes of substances: methyl glycosides, aldonic acid
lactams, aldonic acid lactones and l,S-dideoxy-l,5-iminosugars. The complex mixture
has been separated into individual components that have been screened for activity
against 01- and B-glucosidases. The active molecules have been identified by nuclear
magnetic resonance spectroscopy and high-resolution mass spectrometry. This
methodology provides a straightforward, general and time-efficient way to synthesize,
test and analyze an extensive number of glucosidase inhibitors. It can also be extended to

many other compounds.

67

3.1 . Introduction

Glycosidase inhibitors have the potential to produce several beneficial therapeutic effects
including reduction of the blood glucose level [1], inhibition of tumor metastasis [2] and
inhibition of viral replication [3-5]. Four of the most common and potent classes of
glycosidase inhibitors are glycosides, aldono-1,5-lactones, amino-deoxyaldonic acid
lactams, and 1,5-dideoxy-1,5-imino alditols (aza-sugars). There are several authoratative

reviews on this area [6-9]. Table 3.1.1 summarizes their inhibition of 01— and B-

glucosidases.
0H 0H 0H 0"
o o N N
”02% OR “0 "° "°
H H 0
OH 0H 0H 0 0"
glucosides aldono-1,5-hctone amino deoxyaldonic 1,5-dideoxy- 1,5

acid lactams -'nninoghc'toh

The classes of inhibitory molecules mentioned above are usually prepared using
unrelated synthetic routes although one way of preparing 5-lactams has been to oxidize
1 ,S-dideoxy-1,5-iminosugars such as nojirimycin (Scheme 3.1.1) [18]. However,
cOnsidering the high cost and instability of nojirimycin, this way of preparing lactams is

not at all appealing.

68

OH 0 H
NH 1 ,NaOH NH
”(34% _—_’2 “0
H20 H

OH OH OH 0

Scheme 3.1.] Preparation of 1,5-gluconolactam from nojirimycin.

The inhibitory activity of these classes of compounds can be modulated by substitution.
For instance, some O-alkylated and -arylated 5-lactams have been found to markedly
inhibit metastasis of cancer cells [19] (Figure 3.1.1). Substitution therefore affords a way
0f influencing the specificity, efficacy, potency and bio-availability of drug targets based
on these molecules. Routes for the syntheses of carbohydrate glycosidase inhibitors,
especially with O-substituents, are very involved. Because of this, it is not possible to
routinely prepare and screen a significant number of candidates. An integrated
combinatorial approach that gives access to candidates from all three classes mentioned

above and with variations in substitution would be an ideal and invaluable strategy.

c OY
H z 2
OH 0 oz 0 oz 0
Y=HR Y=H,RCO y=HaAr
’ z = H, RCO, ArCO Z = H, RC0

Figure 3.1.1 Biologically active 5-lactams.

69

Table 3.1.1. Inhibitory action of some glucosidases

 

 

Inhibitor Glucosidase Ki (uM) Reference
(Source)
a-D-Glucosidase
OH Rabbit,intestinal sucrase l 9000 10
“0% B-D-Glucosidase
HO Alcaligenesfaecalis 6400 l 1
OH 0“ Aspergillus wentii 2800 12
Sweet almonds, A 189000 13
B 80000 14
Human liver, cytosolic 55000 15
OL-D-Glucosidase
OH Rabbit, intestinal sucrase 10000 10
“Om B-D-Glucosidase
H0 0 Alcaligenesfaecalis 1.7 l 1
0” Aspergillus wentii 9.5 12
Sweet almonds, A 200 13
B 36 14
Human liver, cytosolic 15 15
OH OL-D-Glucosidase
NH Rabbit, intestinal sucrase 23000 16
[10% B-D-Glucosidase
”0 Aspergillus wentii 36 12
OH 0 Sweet almonds, B 37 13
a—D-Glucosidase
OH Rabbit, intestinal sucrase 0.032 16
fl Yeast 12.6 16
HO Rice 0.01 17
HO OH B-D-Glucosidase
Aspergillus wentii 2.7 12
Sweet almonds, B 47 13
Calf liver 210 15
NH a-D-Glucosidase
HOHO Brewers yeast 1560 18
OH B-D-Glucosidase
Sweet almonds 780 18

70

3.2. Results and discussions

In chapter 2 of this dissertation we have described in detail the procedure for the
transformation of methyl-B-D-glucopyranoside 1 into 5-amino-D-gluconic acid lactams
5, 8 and then to deoxynojirimycin 9 and dideoxynojirimycin 6 (Scheme 3.2.1) [20]. This
involves the chromium trioxide oxidation of methyl-B-D-glucopyranoside to form 5-keto
aldonic acid methyl ester, which is then converted to a cyclic arninosugar by reductive
ring closure with a nitrogen species. One important feature of this method is that it can
potentially afford access to all three of these classes of molecules. We used this
procedure as the basis for the design of a combinatorial approach to partially methylated
B-methyl-D-glucopyranosides, 5-amino-5-deoxy—D-gluconic acid lactams and 1,5-
dideoxy-1,5-imino glucitols. This method has the advantage of producing structural
diversity and a comprehensive spectrum of molecules through just a few simple and high
yield reactions. The strategy is illustrated in Scheme 3.2.2. Methyl-B-D-glucoside 1 was
partially methylated and then acetylated, giving an array of differently substituted
glucosides 11. This mixture of partially methylated, partially acetylated glucosides was
subjected to chromium trioxide oxidation to give 5-keto-gluconic acid methyl ester
derivatives 12. These were reacted with hydroxylamine to produce oximes 13. The
oximes were then converted to 5-lactams 14 by catalytic hydrogenation. Part of this
reaction mixture was then deacetylated to give 15, while another portion was reduced to
the 1,5-dideoxy-l,5-imino-D-glucitols 16 [21]. Both 15 and 16 were subjected to HPLC
separation. All of the individual components from HPLC separation of mixtures 15 and
16 were assayed for their inhibitory activity against 01- and [i-glucosidases. Six

compounds proved to be active. Structure determination using NMR spectroscopy and

71

OH OAC OAc

HO O AC2O, Py AC0 O Cr03, ACZO O O
HO OCH3 —-—> ACO OCH} ——-> AcAcO OCH3

OH OAC OAC O
1 2 3
OAc
H NOH H C
2 NOH H2. Pd/C 3 NH 3H,/ml: H C
——> . 3 NH
ACO OCH3 ——> AcO ——-> HO
Py AcO AcOH AcO H0
OAc 0 OAc O OH
4 5 6

lNzH4
OH

OH OH
NOH NH BH [THF NH
H%6&<NHNHZ H2, ME [lo/m 3 ’ ”(Ho
ACOH HO
OH 0 OH O OH

7 8

Scheme 3.2.1 Synthesis of 6—gluconolactam, dideoxynojirimycin and deoxynojirimycin.

72

OR

0H
0 CH31, Ago 0
H0 MeOH R0
0“ OR

1 10 R = H or CH3
OR
CTO3, ACZO R0 OCH} H2, Pd/C
RO ‘
OR 0
12 R = Ac or CH3
HZNOH
Py
OR
NOH
H , Pd/C
R30 . OCH3 2 ’
OR 0

l3 R= Ac or CH3

/K2CO3
MeOH

R*
NH
R0
R0
OR 0

15 R = H or CH3
R‘= H or OCH3

OR

Ac 0, Py 0
. __.2 Rem.

OR

11 R=Ac or CH3

OR

O
RO

ORO

14A R = Ac or CH3

Rt

NH
R0
116%

ORO

14 R = Ac or CH3
R*= H or OCH3

lam/THF

R.
N H
R0
Rm
OR

16 R = H or CH3
R*= H or OCH3

Scheme 3.2.2 Synthesis of partially methylated 5-amino-5-deoxy-gluconic acid lactams,

1,5-dideoxy-1,5-imino-glucitols and gluconic acid lactones.

73

high resolution mass spectrometry (HRMS) on the acetylated active substances led to the

structures in Table 3.2.1. A variety of positive candidate molecules were obtained from
the screen. Methyl-[i-D-glucosides 17, 18 with the hydroxyl at C-2 methylated were quite
common, indicating that the presence of the methoxy group retarded chromium trioxide
oxidation at the adjacent anomeric site. Three different lactams, the 5,6-dideoxy-1actam
19, the 2,3-dehydro-3,5,6-trideoxy lactam 20 and the enone 21, all displayed good to
excellent inhibitory activity despite their difference in functionality and the extent of
their structural planarity. The 2-oxo-lactone present as its hydrate 22 also shows good
inhibitory activity. The presence of this lactone indicates that some proportion of the 5-
keto-gluconic acid methyl ester was not oximated, and the carbonyl group was reduced
back to a mixture of D-gluco and L-ido aldonic esters. These subsequently cyclized to
produce the derivatized lactones. In the scheme described here, the proportion of lactam
species can be controlled by varying the amount of hydroxylamine used. For instance, the
use of half the required amount of hydroxylamine would increase the proportion of

lactone species produced.

3.3. Conclusions

Using a combinatorial library approach towards synthesis and screening, we have
identified methylated methyl-B-D-glucosides, lactams and lactones with inhibitory
activity towards or- and B-glucosidases. This method, which is primarily a discovery and
screening routine, is very convenient and practical. It circumvents the complexity and

tremendous effort needed to access these classes of molecules when treated separately

74

Table 3.2.1. Structure, characterization and inhibitory activity of some synthesized

glucosidase inhibitors.

 

C d Stru t l a Inhibitory
om un CUTE _
p0 H NMR (500 MHz) HRMS activityb
OAc
0 1.99 (s), 2.04 (s), 2.06 (s). 3.15 (dd, J=7.5, 9.3), calcd. 335.1264 8 5°/
,7 Ac 0011 3.48 (s), 3.54 (s), 3.63 (m),4.09 (dd.J=12.2,1.9), found 335.1353 “13)
Ac - 4.26 (dd, 1:122, 4.6). 4.98 (t, J=9.3), 5.07 (r, J=9.2).
OMe
OAc
. . . =9. . - . ,
o 2 10(s).213(s),305(dd,.1 5),3 26 (1, 1-9 :) calcd. 307.1393 20% (13)
18 M 3.40 (s). 3.48 (s), 3.51 (m), 3.54 (s), 4.24 (dd,J 250/ (or)
A OCH; 11.9, 5.1),427 (d, J=7.7), 4.36 (dd, 1=11.9, 2.2), [mm 307-1407 °
C 5 09 (t J=9 5)
OMe ' ‘ ‘ '
H3C
NH 1.21 (d, J=6.6), 2.12 (s), 2.17 (s), 3.45 (s). 3.52 (m), calcd- 260-1 '3‘ 25% (B)
'9 M 3.72 (t. J=8.6). 4.96 (1 .J=8.6). 5.16(d, J=8.6),5.4S found 260-1136
AC (5)
OAc 0
H3C
20 AcO N” 1.25 (d, J=6.8), 2.05 (s). 3.66 (s), 3.88 (m), 5.l9(m), calcd. 200.0923 15%(00
5.36(s). 5.58(d,.l=6.3). found ZOO-0923 25%(13)
/
OMe 0
OAc
NH
21 2.07 (s). 3.80 (s). 4.19 (dd, 1:121, 5.7). 4.30 (dd, 90% (a)
\ 1:121, 3.5), 5.10 (ddd, 1:2.0. 5.7, 3.5), 5.98 (d, 1:20).
0 O
OMe
OAc
2.10 (s), 2.18 (s). 3.74 (s). 4.46 (dd, J=9.1, 6.6), 4.50 calcd. 291.0954
22 AC (dd. 1:91, 7.9), 4.89 (ddd, J=7.9, 5.3, 6.6),5.20 ((1, found 291.0726 84% (a)
M J=2.8), 5.62 (dd, J=2.8, 5.3). +
HO 0H 0 M

 

a. Fast atom bombardment high resolution mass spectrometry (FABHRMS), M+H+ given.
b. % inhibition per lmM compound, measured for 0.5 mM 01- and B—p—nitmphenylglucoside respectively and 0.2 mg/mL of enzyme

at 37°C. In brackets. the tvoe of glucosidase is given.

75

and gives a wide cross section of candidates. This procedure is general and can be used to
prepare, isolate and screen other classes of alkylated and arylated carbohydrate

derivatives

3.4. Experimental section

3.4.1. Synthesis

In this procedure, the product of each reaction is used, after workup, without fiirther
purification or separation. In a 200 mL round-bottom flask, equipped with a condenser
and drying tube, methyl-B-D-glucopyranoside 1 (10.0 g, 49.2 mmol) was dissolved in
methyl iodide (30.0 mL, 492.0 mmol) and methanol (40 mL) and heated under reflux for
6 h. AgO (25.0 g, 107.8 mmol) is added to the solution in 10 equal portions at 0.5 h
intervals. After the final addition, heating is continued for 4 h, followed by stirring at
room temperature for another 12 h. The suspension was then filtered, extracted with

chloroform and dried on the rotary evaporator to give 10.

Acetylation of the partially methylated methyl-B-D-glucopyranoside was performed using

standard procedures to give 11.

To a solution of 11 (14.0 g) in acetic acid (280 mL) and acetic anhydride (24 mL), Cr03
(9.0 g) was added and the suspension was stirred at 50°C for 2 h. The mixture was then
poured slowly, with stirring into cold water (1000 mL). The water was extracted three

times with chloroform. The combined chloroform layers were decolorized with activated

76

charcoal, filtered and washed with saturated NaHC03 solution and then with water. Afier

drying with NaZSO4 and rotary evaporation of the solvent the ketone 12 was obtained.

This reaction mixture (4.0 g) was dissolved in pyridine (40 mL) and the solution cooled to
0°C. Hydroxylamine hydrochloride (5.0 g) was then added and the solution stirred at 0°C
for 15 minutes and then for another 2 h at room temperature. The mixture was poured
onto ice and water and then extracted three times with chloroform. The combined
chloroform layers were subsequently washed with water, dried with NaZSO4 and then

evaporated to give oxime 13.

A solution of 13 (12.5 g) in glacial acetic acid (300 mL), containing 10% Pd/C (5 g) was
hydrogenated in a Parr reactor under a H2 pressure of 300-400 psi for 40 hours at 55°C.
The reaction mixture was filtered through Celite and washed with ethanol. The solvent

was rotary-evaporated and the lactam 14 (15.0 g) was obtained.

Deacetylation was performed by dissolving 14 (1.0 g) into methanol (20 mL) and water (3
mL) to which K2CO3 (1.0 g) was added and stirred overnight at room temperature. The
suspension was evaporated, partially redissolved in ethanol and filtered. The filtrate was
then dried by rotary evaporation, after which the crude reaction mixture 15 was subjected

to HPLC separation.

77

A solution of the lactam 14 (1.2 g) in THF (10 mL) was reduced by adding 1M BH3/T HF
(13 mL) and stirring at room temperature for 1.5 h and then refluxing for another 1.5 h.
After cooling to room temperature 9% methanolic HCl (10 mL) was carefully added and
the resulting solution was refluxed for 30 minutes. The THF was removed by rotary
evaporation and the reaction mixture was dissolved repeatedly in methanol, followed by
evaporation to remove borates. Thus, 16 was synthesized which was then separated by

HPLC.

3.4.2. Separation of the reaction mixtures

The crude reaction mixtures 15 and 16 were separated via reverse-phase high
performance liquid chromatography (HPLC) using a Waters 600 multisolvent delivery
system with an ODS Ultrasphere 4.6 x 250 mm column, a Spectoflow 783 programmable
absorbance detector and a Waters 740 Data Module. The reaction mixture was dissolved
in a 5% solution of acetonitrile in water and 7011L (20 mg) loaded on the column.
Running time was set to 75 minutes, with a flow rate of 1.5 mL/min. The starting solvent
was 5% acetonitrile in water while the second solvent was water. Gradient curve shape 6
from the Waters gradient program was chosen to mix the two solvents. The absorbance
detector was set to 219 nm while the detection range was programmed to 0.2 AUFS. 28
fractions were collected from reaction mixture 15, and 52 fractions fiom reaction mixture

16. These were subjected to inhibition studies against 01- and B-glucosidase.

78

3.4.3. Inhibition studies

Materials

The enzyme, buffers and substrates were purchased from Sigma. These include: a—D-
glucosidase (maltase, 45U/mg), phosphate buffer, p-nitrophenyl or-D-glucoside, B-D-

glucosidase (from Almonds, 30U/mg), citrate buffer, p-nitrophenyl B-D-glucoside.

Preparation of solutions

(a) a—D-Glucosidase: The stock enzyme solution was prepared by dissolving 0.2 mg of
solid protein in 1 mL of buffer solution. This enzyme solution was diluted lO-fold for the
enzymatic assay. (b) B-D-Glucosidase: The stock enzyme solution was prepared by
dissolving 0.4 mg of solid protein in 1 mL of buffer solution. This enzyme solution was
diluted 10-fold for the enzymatic assay. (c) Phosphate buffer (0.5 M with 0.244 M
KzHPO4 and 0.256 M KHZPO4, pH 6.5): 24.4 mL of l M K2HPO4 and 25.6 mL of l M
KHzPO4 were mixed and diluted to 100 mL. ((1) Citrate buffer (0.3 M, pH 5) (e) p-
nitrophenyl or-D-glucoside (0.5 mM) and p-nitrophenyl B-D-glucoside (0.5 M) were
dissolved in the corresponding buffer solutions. (f) 1 mM solutions of inhibitors were

used.

General procedure for enzyme assay
To a 96-well disposable microtiter plate with a maximum volume per well being 300 11L,
the assay solutions and the potential inhibitors collected from reaction mixtures 15 and 16

were added as follows: 20 [LL of buffer, 20 [LL inhibitor, 10 1.1L enzyme and 40 1.1L of

water. The solutions were mixed and then incubated at 37°C for 15 minutes. Ten [J.L of

79

substrate were then added and the plate was incubated at 37°C for another 35 minutes.
After this, the molar extinction coefficient was read in a spectrophotometer with a
microtiter plate reader set at 420 nm. This procedure has been used for both 01- and B-
glucosidase assays, with the only difference being basification of the B-glucosidase

solutions with Na2CO3 prior to readings.

3. 4. 4. Structure identification
The lH-NMR spectra were recorded at 500 MHz on a Varian VXR spectrometer.
The fast atom bombardment high resolution mass spectra (FABHRMS) were obtained

using a J EOL HX l 10 double focusing mass spectrometer operating in positive ion mode.

Methyl 2-0—methyl-3,4,6-tri-0—acetyl-fl-D—glucopppyranaside (17)

'H-NMR (500 MHz, CDC13); 5 1.99 (s, 3H, OAc), 2.04 (s, 3H, OAc), 2.06 (s, 3H, OAc),
3.15 (dd, 1H, 1.; 7.57 Hz, 12,3 9.28 H2, H2), 3.48 (s, 3H, OCH3), 3.54 ( s, 3H, OCH3),
3.63 (m, 1H, H5), 4.09 (dd, 1H, 16,6, 12.2 Hz, 16.5 1.95 H2, H6.), 4.26 (dd, 1H, 16,6, 12.2
Hz, J6b,5 4.64 Hz, be), 4.98 (t, 1H, J3,2 = J3,4 9.28Hz, H3), 5.076 (t, 1H, J3,4 = .145 9.28 Hz,

H4); HRFABMS (M+H+) calcd. 335.1264, found 335.1353.

Methyl 3,6—di-0-acetyl-2,4—di-0-methyl -fl-D—g1ucopyranoside (18)
IH-NMR (500 MHz, CDC13); 5 2.10 (s, 3H, OAc), 2.13 (s, 3H, OAc), 3.05 (dd, 1H, 11,2 =
J23 = 9.5 Hz, H2), 3.26 (t, 1H, J33 = .14; 9.5 H2, H4), 3.40 (s, 3H, OCH3), 3.48 (s, 3H,

OCH3), 3.51 (m, 1H, H5), 3.54(s, 3H, OCH3), 4.24 (dd, 1H, 16,, 6, 11.93 Hz, 16,5 5.08 Hz,

80

 

H63), 4.27 (d, 1H, JL2 7.73 H2, H1), 4.36 (dd, 1H, J68], 11.93 Hz, .1655 2.21 Hz, H6b), 5.09

(t, 1H, J23 = 13,. 9.5 H2, H3); HRFABMS (M+H+)ca1cd. 307.1393, found 307.1407.

2, 3 -Di-0-acetyl-4-O—meth yI-5-amino-5, 6-dideoxy-D-glucono-1, 5-lactam (1 9)

lH-NMR (500 MHz, CDCl3); 5 1.21 (d, 3H, J36 6.62 Hz, CH3), 2.12 (s, 3H, OAc), 2.17
(s, 3H, OAc), 3.45 (s, 3H, OCH3), 3.52 (m, 1H, H5), 3.72 (t, 1H, J45 = .134 8.61 H2, H4),
4.96 (t, 1H, 123 = J34 8.61 H2, H3), 5.16 (d, 1H, 12.3 8.61 H2, H2), 5.45 (br s, 1H, NH);

HRFABMS (M+H+) calcd. 260.1134, found 260.1136.

4-0—Acetyl-2-O-methyl—2,3-dehydro -3,5,6-trideoxy-5—amino-1,5-lactam (20)
'H-NMR (500 MHz, CDCl3); 5 1.25 (d, 3H, 15,6 6.84 Hz, CH3), 2.054 (s, 3H, OAc), 3.66
(s, 3H, OCH3), 3.88 (m, 1H, H5), 5.19 (m, 1H, H4), 5.36 (br s, 1H, NH), 5.58 (d, 1H, .134

6.35 Hz, H3); HRFABMS (M+H+) calcd. 200.0923, found 200.0923.

6-0-Acetyl-3-O-methyl-5-amino-4,5—dideoxy-3,4-dehydro-D-gluco-2-ulosonic

acid lactam (21)

IH-NMR (500 MHz, CDCl3); 8 2.07 (s, 3H, OAc), 3.80 (dd, 1H, J6“, 12.15 Hz, .1635 5.74
Hz, H6a), 4.30 (dd, 1H, 163,1, 12.15 Hz, .1635 3.53 Hz, Hsb), 5.10 (ddd, 1H, J52 1.99 Hz, 1635

5.74 Hz, .1635 3.54 Hz, H5), 5.98 (d, 1H, J45 1.99 H2, H4).

81

 

4, 6-Di-0-acetyl-3-O-methyl-D-gluco-Z-ulosonic acid lactone hydrate (22)

1H-NMR (500 MHz, CDCl3); 8 2.10 (s, 3H, OAc), 2.18 (s, 3H, OAc), 3.74 (s, 3H,
OCH3), 4.46 (dd, 1H, J63], 9.06 Hz, J63; 6.63 Hz, H63), 4.50 (dd, 1H, J65], 9.06 Hz, .1655
7.95 Hz, H615), 4.89 (ddd, 1H, J6a,5 6.63 Hz, .1561, 7.95 Hz, J45 5.3, H5), 5.20 (d, 1H, J34

2.87 H2, H3), 5.62 (dd, 1H, 13‘. 2.87 Hz, 14,5 5.3 Hz, H4).

82

 

3.5. References

10.

ll.

12.

l3.

14.

15.

l6.

l7.

18.

Liu, P.S. US Patent 4,634,765, 1987.

Dennis, J .W. Cancer. Res. 1986, 46, 5131-5136.

Fleet, G.W.J.; Karpas, A.; Dwek, R.A.; Fellows, L.D.; Tyms, A.S.; Petursson, S.;
Namgoong, S.K.; Ramsden, N.G.; Smith, P.W.; Son, J.C.; Wilson, F.D.; Witty, D.R.;
Jacob, G.S.; Rademacher, T.W. FEBS Lett, 1988, 237, 128-132.

Rohrschneider, L.R. US Patent 5, 643,888, 1997.

Hirsch, M.S., US Patent 5,011,829, 1991.

Legler, G. Adv. Carbohydr. Chem. Biochem. 1990, 48, 319-384.

Ganem, B. Acc. Chem. Res. 1996, 29, 340-347.

Raadt, A.; Ekhart, C. W.; Ebner, M.; Stutz, A. E. Topics in Current Chemistry, 1997,
187, 157-186.

Stick, R. V. Topics in Current Chemistry, 1997, 187, 187-213.
Cogoli, A.; Semenza, G. J. Biol. Chem. 1975, 250, 7802-7809.
Day, A. G.; Withers, S. G. Biochem. Cell. Biol. 1986, 64, 914-922.

Legler, G.; Sinnott, M. L.; Withers, S. G. J. Chem. Soc., Perkin Trans. 2 1980, 1376-
1383.

Dale, M. P.; Ensley, ll. 13.; Kern, K.; Sastry, K. A. R.; Byers, L. D. Biochemistry
1985, 24, 3530-3539.

Legler, G. Struktur des aktiven Zentrums glykosidspaltender Enzyme, Forschungsber.
Nordrhein-Westfalen, Nr.2846, Westdeutscher Verlag, Opladen, FRG, 1979.

Legler, (3.; Bieberich, E. Arch. Biochem. Biophys. 1988, 260, 437-442.

Hanozet, G.; Pircher, H. P.; Vanni, P.; Oesch, B.; Semenza, G. J. Biol. Chem. 1981,
256, 3703-3711.

Inouye, S.; Tsuruoka, T.; Ito, T.; Niida, T. Tetrahedron, 1968, 23, 2125-2144.

Kajimoto, T.; Kiu, K.K.-C.; Pederson, R. L.; Zhong, Z.; Ichikawa, Y.; Porco, J. A.;
Wong, C.-H. J. Am. Chem. Soc. 1991, 113, 6187-6196.

83

 

 

19. Tsuruoka, T.; Nakabayashi, S.; Fukuyasu, H.; Ishii, Y.; Tsuruoaka, T.; Yamamoto,
H.; Inouye, S.; Kondo, S. US Patent 5,250,545, 1993.

20. Pistia, G.; Hollingsworth, R. I. Carbohydr. Res. 2000, 328, 467-472.

21. Wolfrom, M. L.; Thompson, A. Methods Carbohydr. Chem, 1963, 211—215.

 

84

Chapter 4

Chromium Trioxide Oxidation of Glycosides. A Comparative Study

85

Abstract
The chromium trioxide oxidation of glycosides has been known for a long time [1,2]. It
required large excesses of oxidant, which was a deterrent, considering the high toxicity of
Cr°+. This chapter presents a synthetic improvement over the old method, reducing the
amount of oxidant used by more than lO-fold and improving the yield of the reaction to
100%. We also describe the discovery that (Jr-benzyl glycosides are converted to or-
benzoates in quantitative yield. This reaction loses its specificity in the case of B—benzyl

glycosides where anomeric cleavage competes with benzyl oxidation. A detailed study of

this behavior is presented.

86

4.1. Introduction

In 1970 Angyal published his results on the chromium trioxide oxidation of
carbohydrates [1,2]. Chromium trioxide in acetic acid has been shown to oxidize acetals
of aldehydes, R'CH(OR2)OR3, to esters, R'COORZ, in which one of the alcohols
constituting the acetal has been retained while the other is lost during oxidation of the
alkylidene fragment to a ketone. Glycosides, being acetals, are oxidized according to this
general scheme, the aglycon being retained and the ring being ruptured. The products are
therefore esters of S-hexulosonic acids from pyranosides and 4-hexulosonic acids from
furanosides [3]. Scheme 4.1.1 exemplifies this chemistry for methyl-B-D-glucoside l. A
remarkable feature of the reaction is that it is specific to B-pyranosides but occurs with
both 0L- and B—furanosides. Since primary and secondary hydroxyl groups are rapidly
oxidized by the reagent, they must be protected, for example by esterification, since

esters are stable under these conditions. Acetals, being attacked, are not suitable for

OAc OAc
0 CrO Ac 0 o
AcO 3’ 2 A
AcmOCPB —> CAC OCH3
OAc OAc O
1 2

Scheme 4.1 .l Chromium trioxide oxidation of methyl 2,3,4,6-tetra-O-acetyl-glucoside.

87

protecting the hydroxyl groups. Methyl and benzyl ethers are also attacked by chromium
trioxide in acetic acid [4], and hence these are also unsuitable as protecting groups.
Methyl ethers are converted into formic esters, while benzyl groups are transformed to
benzoates. Disaccharides, being glycosides, are similarly oxidized. For example, the

octaacetate of lactose 3, which is a B-3-O-glucosyl-galactoside, is smoothly converted

into a keto ester, while maltose 4, being a (it—glycoside does not react.

OAc
AC OAc OAc Ac 0
AcO

O 0 Ac OAc
AGO A A00 0

c

AcO 0 Ac AcO Aco
OAc
3 4

The chromium trioxide oxidation has also been used to determine the anomeric natures of
sugar residues in oligo- and polysaccharides [5]. The materials are fully acetylated,
oligosaccharides having first been reduced to their alditols, and the products are then
treated with chromium trioxide in acetic acid in the presence of an internal standard. A
comparison of sugar analyses, performed before and after oxidation, reveals what sugar
residues have been oxidized. A prerequisite for obtaining reliable results is that the
carbohydrate does not contain free hydroxyl groups but is fully acetylated and that the
conformational stability of the chain forms having axially attached aglycons is large
enough to ensure that the proportion of the alternate form is negligible. This supposition

is valid for the or-pyranosides of the common hexoses, 6-deoxyhexoses, 2-acetamido-2-

88

deoxyhexoses and xylose. This technique has been applied for the elucidation of the
anomeric configuration of bacterial lipopolysaccharides from Salmonella typhi and
Salmonella Strasbourg [6]. The method may also, in favorable circumstances, be used for
sequence analysis, like in the case of the Pneumococcus type 2 capsular polysaccharide

[7,8].

The mechanism of the oxidation of acetals by chromium trioxide in acetic acid is not yet
known. Presumably the actual oxidizing agent is chromyl acetate, CrO2(OAc)2, since
chromium trioxide in water or pyridine will not give keto esters. It is not clear whether
the oxidation of the acetal to ester and that of one of the alkoxyl groups to ketone are
simultaneous or consecutive reactions. In order to decide this point, the oxidation of the
B-glucoside was carried out in acetic anhydride rather than in acetic acid according to the
reasoning that if the acetal were first oxidized to an ester, liberating a hydroxyl group, the
latter might be acetylated faster than it is oxidized. However, the same keto ester was
obtained and the experiment was inconclusive. It was then argued that if the two
reactions are concerted, oxidation of the glycopyranoside could not occur if there were no
hydrogen atom on C5. A suitable model compound that is a branched-chain glycoside is
methyl 6-deoxy-5-C-methyl-B-D-xylo-hexopyranoisde. The acetylated glycoside 5 was
treated with 003. Oxidation occurred at a slower rate than that of the unbranched B-
glycosides and two products were obtained: the minor one was the lactone 6 and the
major one was methyl 2,3,5-tri-O-acetyl-6-deoxy-5-C-methyl-L-threo-4-hexulosonate 7
(Scheme 4.1.2). It appears that acetyl migration occurred, presumably after oxidation of

the glycoside to ester, from O4 to 05, followed by oxidation of the free hydroxyl group

89

on C4. The experiment would suggest that removal of the two hydrogen atoms - from the
acetal and from the alkoxyl group - is not a concerted process: oxidation of the acetal will
occur even if there is no hydrogen atom on C5 (Scheme 4.1.3). However, this does not
preclude the possibility of the oxidation proceeding by two different mechanisms, a
concerted one operating with unbranched B-glycopyranosides. The ring-opening of the
branched-chain pyranoside (b) is, however, slower, allowing the competing reaction -
breaking of the bond between the anomeric carbon and the aglycon (a) to give the lactone
- to achieve some degree of prominence. It is possible that the first step in the oxidation is
the formation of a complex between the acetal and chromyl acetate; this complex then

breaks down at a rate, and in a direction, dependent on the nature of the two alkoxyl

groups of the acetal.
H3C H3C O H3C
were ”fir" M
—-> + OMe
CH3 OAc CH3 OAc 0 CH3 OAc 0
5 6 7

Scheme 4.1.2 Oxidation of methyl 6-deoxy-5-C-methyl-B-D-xylo-hexopyranoisde.

90

 

OAc

+ 0: Cr 2 —>
H I 0 GAO

\Cr
AcO
OAc
0 |
De + 0=r +
+H+ OAc 5
07° OR /n' 1
\J
<:><O a OAc + +
p\ / b
/Cr\ OAc
O OH |
ACO OR
+ O=Cr
0 1
OAc
KR
OAc H \\ /OAC
OH OR l ‘ O;C\{
0 + O=Cr=0 -———> O ——>
(LA -AcOH 0R
c o
(I)H
OR + 0: Cr
0 |
OAc

Scheme 4.1.3 Proposed mechanism for the oxidation of glucopyranosides by 003,

91

4.2. Results and discussions

4. 2.1. Improvements in the oxidation of methyl ,B-D-glycosides

An important previous disadvantage of this method has been the employment of large
excesses of CrO3. In his experiments, Angyal was using 28 equivalents of oxidant
obtaining a maximum of 70% conversion. In our efforts to minimize the quantity of
CrO3, we obtained complete conversion of B-methyl glucoside 1 (Scheme 4.1.1) and [3-
methyl galactoside 9 (Scheme 4.2.1) to products using only 2 equivalents of oxidant. This
represents a significant improvement from the initial 28 equivalents used, considering the
toxicity of the oxidant and the difficulty in removing traces of CrO3. In our procedure we
used a 12.5:1 (vol) mixture of glacial acetic acid and acetic anhydride. It is possible that
acetic anhydride removes the traces of water and improves the yield of the reaction. It
also seems that the 100% conversion is general and does not depend on the nature of the

B-glycoside substrate.

AcO OAc AcO OAc

O CrO3 O
_—’ OMe
Ac 0M6 Ac
OAc OAc 0
8 9

Scheme 4.2.1 Chromium trioxide oxidation of methyl 2,3,4,6-tetra-O-acetyl-galactoside.

92

 

4. 2. 2. Oxidation of benzyl glycosides to glycosyl benzoates

Angyal reported [4] that O-benzyl ethers are converted into benzoates. The substrate that

he analyzed was 1,2,4,6-tetra-O-acetyl-3-O-benzyl-B-D-glucopyranose 10 which he

transformed into 1,2,4,6-tetra-O-acetyl-3-O-benzoyl-B-D-glucopyranose 11 with a 50%

total yield and using 10 equivalents of chromium trioxide (Scheme 4.2.2). Benzyl groups

are important protecting groups that are orthogonal to esters. They are inert to acids and

bases and can be removed only by catalytic hydrogenolysis. A benzyl group in the

anomeric position constitutes a good protecting group that can be easily introduced. Once

the anomeric site needs to be activated, the protecting group has to be removed. This is

traditionally done by hydrogenolysis. However, if the benzyl group is transformed into a

benzoate, the latter can be either hydrolyzed under basic conditions or directly

transformed into a glycoside, following activation.

OAc

C6H5CH2 OAC ACZO

OAc

10

OAc
Ac 0
C6HSC OAc
ll OAc
O
l 1

Scheme 4.2.2 Oxdation of 1,2,4,6-tetra-O-acetyl-3-O-benzyl-B-D-glucopyranose.

93

 

By using the same reaction conditions as earlier described for the oxidation of methyl
glycosides and using again the same low amount of chromium trioxide, we succeeded in
transforming benzyl or-D-glucoside 12 into or-glucosyl benzoate 13 in 100% yield

(Scheme 4.2.3). The ester group can easily be removed by hydrolysis if needed.

OAc OAc
AC 0 A AC 0
AC ACOH ,ACzo AC
AcHN OCH2C6H5 ACHN OfiC5H5
O
12 13

Scheme 4.2.3 Oxidation of benzyl 2-acetamido-2-deoxy—3,4,6-tri-O-acetyl-0L-D-

glucopyranoside.

4.2.3. Chromium trioxide oxidation of benzyl ,B-glycosides. Competition between ring
cleavage and benzyl oxidation.

The synthesis of acyl glycosides is not very straightforward. The classical route for
making them involves benzyl protection of a glycoside, deprotection of the anomeric
center, acylation followed by catalytic hydrogenolysis (Scheme 4.2.4). We already
demonstrated the quantitative conversion of (Jr-benzyl glycosides into benzoates and we
also wanted to analyze the behavior of B-benzyl glycosides or analogs thereof. Since

glycosides with long chain aglycons are of interest because of their tendency to

94

 

aggregate, we considered synthesizing long-chain p-alkoxy—B-benzoyl glycosides. For
this purpose we chose a six-carbon alkyl chain. Scheme 4.3.5 outlines the synthesis. 4-
Hexyloxy benzylalcohol 21 was prepared from 4-hydroxy benzylalcohol l9 and l-
iodohexane 20. This was then reacted with 2,3,4,6-tetra-O-acetyl glucosylbromide 22, in
the presence of silver carbonate, to give p-hexyloxybenzyl-2,3,4,6—tetra-O-acetyl-B-D-

glucoside 23. On oxidation with chromium trioxide, we obtained, however, not only the

OH OBn
C
H NaH Bn OM e

OH OBn
14 15

OBn OBn
Bn 0 BU Ll Bn 0 H2, Pd
Bn 0“ RCOCI Bn ’

o R
03" OBnH in:

16 17
OH
H O
H 0 R
O

18

Scheme 4.2.4 Classical route for the synthesis of benzoyl glycosides.

95

 

KLCX)
HO-—©>—CH2—OH + IC6HI3 —-—Lz C6H130_©>_CH2_OH
EKJH

19 20 21
OAC OAC
Ac 0 + 21 AgZCO3 Ac 0
Ac "_‘" Ac O—CHz—‘©>_OC6HI3
ACO Br A00
22 23
OAc
Ac 0
AC O—fi—@A0C6Hl3
AcO O
24
CrO3
23 ——>
OAc
Ac 0
Ac O_CH2_@-OC6H13
AcO O
25

Scheme 4.2.5 Competition between benzyl oxidation and ring opening in the oxidation.

of benzyl B-D-glucopyranosides.

96

desired p-hexyloxybenzoyl-2,3,4,6-tetra-O-acetyl-B-D-glucoside 24 but also compound
25 with the ruptured ring. The two compounds are easily distinguished by their phenyl
proton signals 24 has two doublets at 6.96 and 7.89 ppm, while the same proton signals in
25 are shifted to 6.89 and 7.98 ppm respectively. Our next goal was to create conditions
in which formation of 25 is avoided or diminished. Since the first oxidation attempt was
performed using 2 equivalents of CrO3 and 50°C, we thought that reducing the amount of
oxidant might improve the yield of 24 but this did not happen. Next, the reaction was run
at different temperatures: 20 and 1°C. The course of the reaction was determined by
taking samples at different times and estimating the conversions from NMR spectra. The
values are indicated in the tables associated with the graphs in Figure 4.2.1. It can be
concluded that modification in the concentration of oxidant does not affect conversion
much, while the temperature has quite a large efiect. The best conversions are obtained at
the lowest temperature. However, none of the conditions accomplishes total
transformation of the benzyl into a benzoate and the ring cleavage cannot be totally

annulled. Therefore oxidation of B-benzyl glycosides proves to be less appealing than

that of or-glycosides.

4.3. Conclusions

The oxidation of B-D-glycosides to keto esters has been known since 1970. The
disadvantage of the method is the high excess of chromium trioxide used. By using small
volumes of acetic anhydride, we succesfully lowered the amounts of CrO3 used to two
equivalents and also obtained quantitative yields. This is a considerable improvement

over the original procedure, considering the high toxicity of the oxidant. Oxidation of

97

 

 

50° C

 

 

 

 

 

Concentration

 

1 + Compd 23 + Compd 24

 

 

 

 

 

 

Compd 25 ‘

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1°C
Time

Compound 0 30 60 90

23 100 70 40 0

24 0 25 30 50

25 0 5 30 50
1 .5
‘1 Ii
1
5
U

 

I+Compd 23 +Compd 24

 

 

, .Compd25.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

20° C
Time
Compound 0 20 40 60
23 100 75 60 10
24 0 10 20 40
25 0 5 20 50

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0 20 40 60
‘ Time
1 ETFEREHEEFEF—OHH-i: :0972129357
50° C (2eq)
Time
Compound 0 10 20 30
23 100 66 40 24
24 0 22 25 20
25 0 12 35 56
1 a 120
(l .2 '00 WW.-- _ -_ ._.-..__._-_- _W... ....W. W- . . .
11 E 80 \\
. ‘5 60 . _.W
1 l 2 40 . . . WWW-e... .-W\. «.2.- he.--
I: c 2:: L.;/keg- _-
X 0 10 20 30
Time
[+Compd 23 +Compd 24‘ .2

Figure 4.2.1. Oxidation of benzyl-B-D-glucopyranoside.

98

 

 

 

- 22222222

benzyl or-D-glycosides to (Jr-glycosyl benzoates can be achieved in quantitative yield by
using the same procedure. Acid or base catalyzed hydrogenolysis provides an alternative
to the traditional debenzylation via catalytic hydrogenolysis. Chromium trioxide
oxidation of B-benzyl glycosides, however, targets both the benzyl group and the sugar

ring, making the reaction less appealing from the standpoint of conversion.

4.4. Experimental section

Methyl-2,3,4,6-tetra-O-acetyl-D-xylo-S-hexulosonate (2) and Methyl-2,3,4,6-tetra-O-
acetyl-L-arabino-S-hexulosonate (9)

The procedure and spectral characterizations are described in the experimental section of

chapter 2.

Benzoyl-2-acetamido-2-deoxy—3,4, 6-tri-0-acetyl- at-D-glucose (13)

To a solution of benzyl-2-acetamido-2-deoxy-3,4,6-t1i-O-acetyl-a-D-glucose 12
(140.00 mg, 0.34 mmol) in acetic acid (1.5 mL) and acetic anhydride (0.1 mL), CrO3
(180.00 mg, 1.80 mmol) was added and the suspension was stirred at 50°C for 30
minutes. The mixture was then poured slowly, with stirring into cold water (10 mL).
The water was extracted three times with chloroform. The combined chloroform
layers were decolorized with activated charcoal, filtered and washed with saturated
NaHCO3 solution and then with water. After drying with Na2SO4 and rotary
evaporation of the solvent the ester 13 (143.00 mg, 100%) was obtained as a colorless
syrup: lH NMR (CDCl3) 5 1.87 (3 H, s, NHAc), 2.03 (6 H, 3, OAc), 2.04 (3H, 8,

OAc), 4.04 (1H, dd, 16,... 12.2 Hz, 16..., 12.45 Hz, H-6a), 4.04 (1H, m, H-5), 4.21 (1H,

99

dd, H-6b, 16,3 3.17 Hz, H-6b), 4.57 (1H, ddd, 112 8.79 Hz, 12,, 9.77 Hz, 12,.” 2.93, H-
2), 5.25 (1H, t, J45 =12. 9.77 Hz, H-4), 5.37 (1H, t, H-3), 5.68 (1H, d, H.), 6.41 (1H,
d, NH) 7.50 (2H, dd, 1,1,... 7.57 Hz, 11.5,, 7.08 Hz, H-Arb), 7.63 (1H, t, H-Arc),
8.06(2H, d, H- c); 13C NMR (CDCl3) 5 20.5, 20.6, 20.7, 22.9, 57.2, 61.3, 67.4, 70.0,

70.7, 91.1, 128.6, 128.8, 129.8, 134.0, 164.1, 169.0, 170.0, 170.6, 171.6.

Hexyloxy benzylalcohol (21)

To a solution of 4-hydroxy benzylalcohol 19 (6.0 g, 48.3 mmol) and l-iodohexane 20
(8.6 mL, 58.0 mmol) in ethanol (40 mL), potassium carbonate (26.8 g, 193.9 mmol)
was added and the suspension was stirred at 55°C for 15 h. A second amount of
iodohexane (1.5 mL, 10.0 mmol) was added and stirring was continued for another 5
h at 55°C and then15 h at room temperature. The suspension was filtered, the
precipitate washed with ethanol and the mother liquor rotary evaporated. The
potassium iodide that is soluble in ethanol, was removed from the product 21 by
precipitation from chloroform/hexane solution. After filtration, the solution was
evaporated and the hydroxyether 21 (9.9 g) was obtained in 100% yield. lH NMR
(CDCl3) 8 1.04 (3 H, t, J”- 6.84 Hz, H-j) 1.46 (2H, hex, J23. 3.66 Hz, H-i), 1.48 (2H, p,
H-h), 1.89 (2H, p, J65 6.59 Hz, , H-f ), 2.98 (1H, s, OH), 4.04 (2H, t, H-e), 4.46 (2H,
d, H-d), 6.97 (2H, d, J 8.56 Hz, H-Ara), 7.33 (2H, d, H-Arb); l3C NMR (CDCl3)

5 13.9, 22.4, 25.5, 29.0, 31.4, 64.5, 67.8, 114.2, 128.4, 132.8, 158.4.

100

p-Hexyloxybenzyl-2,3,4,6-tetra-O-acetyl-fl-D-glucoside (23)

Hexyloxy benzylalcohol 21 (5.0 g, 24.0 mmol) was dried for 1 h in the vacuum oven
and then dissolved in a 1:1 mixture of dry benzene and nitromethane. Calcium
sulphate was added and the suspension was stirred at room temperature for l h. To
this, silver carbonate (1.4 g, 5.0 mmol) and 2,3,4,6-tetra-O-acetyl glucosylbromide 22
(2.0 g, 4.8 mmol) were added and stirred for 20 h. Some more 22 (1.2 g, 3.0 mmol)
was added and the temperature was increased slowly to 55°C and stirring was
continued for another 24 h. The suspension was filtered over Celite, washed with
chloroform and the solution then rotary evaporated. The crude product was purified
by column chromatography using a 1:6 acetone-hexane mixture as the eluent. Pure p-
hexyloxybenzyl-2,3,4,6-tetra-O-acetyl-B-D-glucoside (23) (2.4 g) was obtained in a
57% yield. lH NMR (CDCl3) 8 0.88 (3H, t, J”- 7.08 Hz, H-j), 1.23 (2H, m, H-i), 1.31
(2H, s, J 3.42 Hz, H-h), 1.43 (2H, m, H-g), 1.76 (2H, m, J 6.59 Hz, J 7.3 Hz, H-f,),
1.97 (6H, 5, OAc), 1.99 (3H, s, OAc), 2.08 (3H, 8, OAc), 3.6 (1H, ddd, .154 9.28 Hz,
156,, 2.44 Hz, J5“, 4.64 Hz, H-5), 3.92 (2H, t, J52 6.59 Hz, H-e), 4.13 (1H, dd, J65],
12.21 Hz, J5b,5 2.44 Hz, H-6b), 4.75 (1H, dd, H-6a), 4.48 (1H, d, J1,2 7.81 Hz, H-l),
4.53 (1H, d, H-ad), 4.79 (1H, d, H-db), 5.01 (1H, dd, H-2), 5.07 (1H, t, J45 9.28 Hz,
H-4), 5.13 (1H, t, J3,4=J2,3 9.28 Hz, H-3), 6.84 (2H, d, Jab 8.3 Hz, H-a), 7.16 (2H, d, H-
b); 13C NMR (CDCl3) 5 14.0, 20.6, 20.7, 20.8, 22.5, 25.6, 29.1, 31.5, 61.9, 68.0. 68.3,

70.4, 71.2, 71,7, 72.8, 98.7, 114.3, 128.2, 129.5, 159.0, 163.4, 170.2, 170.7.

101

p-hexyloxybenzoyl—2,3,4,6—tetra-O-acetyl-fl-D-glucoside (24) and p-hexylbenzyloxo-
2,3,4,6-tetra-O-acetyl-D-xylo-S-hexulosonate (25)
p-Hexyloxybenzyl-2,3,4,6-tetra-O-acetyl-B-D-glucoside (23) (0.15 g, 0.27 mmol) was
dissolved in a mixture of glacial acetic acid (1.5 mL) and acetic anhydride (0.2 mL)
and chromium trioxide (0.14 g, 1.40 mmol) was added. The suspension was stirred at
50°C for 30 minutes and samples were taken at 10 minute intervals, and were worked

up using the following procedure. The suspension was poured over cold water (15
mL) and extracted three times with chloroform (10 mL). The solution was rotary
evaporated and the crude solid was analyzed by NMR. 24: 13C NMR (CDCl3) 514.0,

20.5, 20.7, 22.5, 25.5, 28.9, 31.4, 61.2, 66.3, 67.1, 69.7, 70.9, 90.1, 114.1, 114.7,

137.2, 161.9, 169.3, 170.1, 170.7, 176.7.

102

4.5. References

l.

2.

Angyal, S. J.; James, K., Aust. J. Chem. 1970, 23, 1209-1221.

Angyal, S. J.; Evans, M. 13., Aust. J. Chem. 1972, 25, 1513-1520.

. Angyal, S. J .; James, K., Chem. Commun. 1969, 617-621.

Angyal, S. J.; James, K., Carbohydr. Res. 1970, 12, 147-149.
Hoffman, J .; Lindberg, B.; Svensson, S., Acta Chem. Scand. 1972, 26, 1209-1217.

Hellerqvist, C. G.; Hoffman, J.; Lindberg, B.; Pilotti, A.; Lindberg, A. A., Acta
Chem. Scand. 1971, 25, 1512-1519.

Larm, O.; Lindberg, B.; Svensson, S., Carbohydr. Res. 1973, 31 , 120—133.

Larm, O.; Lindberg, B.; Svensson, S., Carbohydr. Res. 1975, 40, 69-82.

103

Chapter 5

Development of a New Glycosylation Procedure for 2-Aamino-2-Deoxy

Sugars

104

Abstract
Amino sugars occur in a host of different compounds mainly as components of
poysaccharides and sometimes in low molecular weight metabolic products, but rarely in
the free form. The finding of aminosugars in antibiotics, in microorganisms, higher
plants, invertebrates and human body has deepened the interest in this class of
biologically important compounds. It appears that in these natural products the amino
groups of the sugars are always protected, most often by acetyl, and occasionally by
forrnyl or methyl groups. N-Acetylglucosamine is the most commonly occurring amino
sugar. Therefore, glycosylation reactions involving them are of high interest. There are
only two procedures that involve N-acetylated glucosamine as starting material: the
Koenigs-Knorr and the oxazoline methods. Here we present a different approach that

consists of activation of a nitro-phenyl leaving group with ZnCl2 to afford high yields of

N-acetamide B-D-glucosides.

105

5.1. Introduction

Amino sugars represent a class of compounds that in diversity and ubiquity has few
parallels in the biological field. The number of amino sugars identified in living
organisms is greater than that of all other sugars. Amino sugars have also been found in
all tissues and fluids of pluricellular organisms. Their structure, which blends those of
both amino acid and common sugar, gives them a marked lyophilic character that may
explain their association with most proteins and with many lipids. Glycoconjugates such
as the bacterial lipopolysaccharides, cell surface glycoproteins and glycolipids, as well as
other complex proteoglycans of immunological significance are involved in a host of
biological processes. Their oligosaccharide parts have received much attention and have
acquired an importance as great as that of proteins and nucleic acids. Their functions are
less well understood, but it is clear that the oligosaccharide moieties of biological
membranes play a decisive role for specific immune reactions [1]. Moreover, they fulfill
important functions in intercellular recognition and interaction, in the control of cell
growth and thus tumor formation, and in the interaction with biologically active factors
such as enzymes, hormones, bacteriotoxin, and viruses. This may be due to the fact that
strongly hydrophilic glycan chains will normally be located at the outer surface of
molecules in aqueous environments, which render them available for interactions with
other molecules. The wide structural variety makes sugars and in particular
oligosaccharides ideal as carriers of biological information. In contrast to peptides and
nucleotides, in which the informational content is determined solely by the number and
sequence of different monomer units, the informational content of oligosaccharides is

fixed additionally by the site of coupling, by the configuration of the glycosidic linkage,

106

and by the occurrence of branching. Thus, polymers made up of carbohydrates can carry
considerably more information per building block then proteins and nucleic acids [2]. N-
Acetylglucosamine is an important component of oligosaccharides with biological
importance. Glycosides of N-acetylglucosamine are widely distributed in living
organisms where they constitute building blocks of glycoconjugates such as
peptidoglycans, mucopolysaccharides (hyaluronic acids, keratan sulfates and inner and
outer core regions of glycoproteins). They are also encountered in human milk, in blood
group substances, in bacterial lipopolysaccharide antigens where they constitute part of
the epitopes, and in plants [3]. A sulfated and acetylated glucosamine oligosaccharide is
recognized by root cells of leguminous plants and acts as a signal for symbiotic host-
specificity of Rhizobium bacteria. This shows that receptors for amino sugars are also
present in plant cells. Of special interest is the occurrence of aminosugars in a variety of
useful antibiotics, such as kanamycin C, paromomycin, hydroxyrnycin, trehalosamine

and zygomycin A [4].

The construction of each individual oligosaccharide poses a new challenge, requiring
knowledge of methods, experience, and experimental dexterity, especially because there
are no universal methods for their synthesis. This chapter will focus on 1,2-trans-
glycosylations of N-acetylglucosamine. Glycosylation reactions are not restricted only to
the actual process of coupling of the donor D with the acceptor A in the presence of
different promoters P, but are constituted from a series of reaction steps: protection of
both the acceptor and the donor, introduction of a participating or nonparticipating group

at C-2 of the donor, activation of the anomeric position of the donor, glycosylation in the

107

presence of a suitable protecting group P and finally deprotection of C-2 and its
retransforrnation in an amido group. Protecting groups are used to block certain hydroxyl
groups in a carbohydrate, so that these will not interfere in reactions. In the case of the
donor sugar, all hydroxyl groups have to be protected, to avoid its self-condensation. On
the other hand, in the case of the acceptor sugar, it has to be ensured that only the oxygen
atom in the desired position forms the glycosidic bond. This can be achieved if that
particular hydroxyl group is the most reactive in the molecule, or it is the only one
available for reaction. If the synthesis of a reducing di- or oligosaccharide is desired, the
anomeric site not involved in glycosylation has to be protected, since it is the most
reactive position in the molecule. There are three major protecting agents: esters, ethers
and cyclic acetals, which can be used alternatively to achieve the desired protection
pattern. The suitable protection of the OH groups determines the regioselectivity of the
reaction. The diastereoselectivity of the glycosylation, that is the or or [3 configuration of
the newly formed glycosidic bond, is determined to a large extent by the blocking groups
of the donor sugar, but also by the type of promoter and solvent used. In order to achieve
1,2-trans glycosylation, the most often encountered linkage for NAG, two approaches are
of general interest (Figure 5.1.1). The most widely used method involves a glycosylation
donor containing a C-2 participating group as the amino protective function. The
formation of a cyclic intermediate (D) by anchimeric assistance results in a shielding of
the or—face of the donor, allowing the reaction of the acceptor alcohol on the B—face only,
thus affording the 1,2-trans glycoside (F) with a high degree of stereoselectivity. This
approach has been used in reactions in which the assisting group is N-acetyl or

phthalimido and the leaving group is a halogen. A second approach involves 1,2-cis-2-

108

 

men

3
1
11

 

 

OR'
RZN x RzN
E’ F
E _
INSOLUBLE _ _
PROMOTER +Y 2 ’ X
o ——+
PARTLY Emfl *— m”
2 - - -
SOLUBLE X +X2 Y RZN
a l
c
R'OH ——>
RZN V RZN OR'
C G
if R is a
participating
amp

11

OR'

X
—->
N
R/ \R RzN

D F

Figure 5.1.1 General scheme of glycosylations.

109

amino-2-deoxy-or-D-glycopyranosyl halides (A) (having an amino nonparticipating
protecting group) and an insoluble promoter able to shield the or-face of the donor (E)
during the substitution with the acceptor alcohol. Selectivities are generally lower than
those observed with the donors containing C-2 participating groups. An important issue is
the survival of the N-protecting group under the conditions necessary for activation of the
anomeric position. The most often employed leaving-groups are: halogens, acetyls, thiols,
trichloroacetimide and 4-pentenyl. The promoters can be used either in catalytic amounts
or in stoechiometric proportions. For oxygenated leaving groups, Lewis acids have
proven adequate (ZnCl2, SnCl4, AlCl3, FeCl3, BF3) and for the rest, silver and mercury
compounds have been used (Ag2O, Ag2CO3, Hg(CN)2, HgBr2, F3C-SO3-Ag, AgClO4, the
activity increasing in the order given). The role of the promoter is both as and activator as
well as an acid scavanger. The promoter should assist the departure of the anomeric
leaving group (X) in a way which avoids the formation of the oxocarbonium ion, which
results, most of the time, in a lack of stereoselectivity and affords mixtures of or— and B-
glycosides (G, F). Thus, after choosing the proper N-protecting group, the promoter and
solvent, the glycosylation reaction is performed to give exclusively or predominantly the
desired anomeric glycosidic linkage. The final step of this procedure is deprotection.
Table 5.1.1 gives a summary of the main leaving groups, N-protecting groups, and the
corresponding promoters. Also, the advantages and limitations of the particular methods
are indicated [32-34]. There are many glycosylation methods but each of them has
disadvantages or limitations. One major disadvantage common to all glycosyl halides is

their low thermal stability and high sensitivity towards hydrolysis.

110

Table 5.1.1 Conditions, advantages and limitations for the most common glycosylation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

procedures.
Leaving group N-protecting gr. Promoters Advantages, limitations & ref
Koenigs -Cl -N-CO-CH3 Ag salt //Hg salts only active alcohols/llow yields
Knorr [5-8]
Oxazoline -Cl -N-CO-CH3 ZnCl2//SnCl4//A1Cl3// harsh cond, not for sec OH.//
-Br F eCl2//F eCl3 migration of acid sensitive gr.//
partial anomerization [9-11]
-Cl ‘ AgTfl-collidine // Ag good for both primary and
N Zeolite // Ag salicilate // secondary OH; deprotection:
OC/ \ CO AgClO4-AgCO2//HgBr2 H2N-NH2, EtOH // NaBH4
[12,13]
'0AC SHC14 // BF 3.EtzO //
M63Si-OSOZ-CF3 good yields [14,15]
-S-R MeTfl // NIS+CF3SO3H good yields // MeTfl is toxic
AgTfl+Br2 [16,17]
-O-CNH-CC13 BF 3'Et2O//Me3SiTfl good yields and selectivity;
low temp [18-20]
-O(CH2)3CHCH2 [(coll)2I]ClO4 good yields [21,22]
-Cl -N3 insol. Ag promoters deprotection: reduction [23-26]
-O-CNH-CC13 BF3 [27,28]
-S-R MeTfl Ot-selectivity only // low yields
for oligosaccharides [29,30]
-O-Ac -NH-CO—OR Me3Si-OSO2CF 3 good yields for primary &

 

 

 

 

secondary OH [31]

 

lll

 

Usually, Cl is favored over Br, which is more reactive but less stable. The Koenig-Knorr
and oxazoline methods do not use N-protecting groups different from acetyl, but this
advantage is counteracted by the fact that only reactive acceptors in excess can react. The
phthalimido method [7] is the method of choice so far, giving 1,2-cis NAG disaccharide
in good yields and high stereoselectivity, but it has the inconvenience of requiring
removal of the phthalimido group, once the reaction is completed. Fraser-Reid discovered
not long ago that 4-pentenyl glycosides can be activated by NBS to give a bromonium
ion, which will be attacked by the alcohol nucleophile to give the trans 1,2-glycoside, but
this procedure also requires the presence of protecting groups. Even though a multitude
of procedures are employed in the synthesis of N-acetylglucosamine glycosides, none of
them is a ‘perfect choice’. This is why the search for better activating groups and

promoters is still going on.

5.2. Results and Discussions

We have demonstrated that the o-nitrophenyl can act both as a protecting group of the
anomeric center and as a leaving group, upon activation with a Lewis acid. The promoter
used is ZnCl2, which is a mild Lewis acid. This constitutes an advantage over many other
highly toxic or highly unstable reagents. The o-nitrophenyl glycoside of NAG can easily
be synthesized by the reaction of 2-acetamido-2-deoxy-3,4,6-tri-O-acetyl-OL-D-
glucopyranosylchloride 1 with o-nitrophenol 2 in the presence of a base like KH and 18-
crown-6-ether (Scheme 5.2.1). The role of the crown ether is to complex the potassium
cation so that the hydride is free and available to abstract the phenolic proton. The 0-

nitrophenylglycoside 3 is stable at room temperature and also in the presence of bases.

112

OAC OZN OAC OzN

 

Ac 0 KH, 18 crown 6 Ac ()
+ H = 0
Ac CHZCI Ac
AcNH Cl AcNH

1 2 3
OAc O2N OH OZN 1‘
AC 0 K2CO3 H O 1
Ac ——’ H .
AcNH M60“, H20 AcNH 1

3 4

Scheme 5.2.1 Synthesis of nitrophenyl-2-amino-2-deoxy—glucoside.

This is a big advantage, considering the lability of the very often-used sugar halide
donors and how easily they decompose in the presence of moisture or even in an
improper solvent system. Under standard deacetylation conditions, 3 is converted to o-
nitrophenyl-2-amino-2-deoxy-glucoside 4, that has a long shelf stability. The suitability
of the o-nitrophenyl group as a leaving group was evaluated by reacting glycoside 4 with
two secondary alcohols: isopropanol and ethyl lactate (Scheme 5.2.2). Heating the
solution of glycoside in excess alcohol at 70°C in the presence of ZnCl2 afforded high
yields of isopropyl-2-amino-2-deoxy-B-D-glucoside 5 and ethyl lactyl-2-amino-2-deoxy-
B-D-glucoside 6 respectively. The B-isomer is formed exclusively. It can be concluded

that o-nitrophenyl is a good protecting/activating group. It can easily be activated by

113

0H 02N H3C—CH2-CH3 0“
H O OH H O

 

 

W CH
H O r H O\ / 3
CH
AcNH ZnClz AcNH \
CH3
4 S
OH OzN H3C—(|ZH2-COOE1 OH
OH 0
H 0 _ H CH3
H r H O\C11
ZDCIZ ACNH \
A°NH coon
4 6

Scheme 5.2.2 Transglycosylation between o-nitrophenylglycosides and alcohols.

ZnCl2 and substituted by primary and secondary alcohols but is not removed under
alkaline conditions. The glycosylation reaction occurs in high yields, with total B-
diastereospecificity. If the synthesis of the N-acetyl glycoside is desired, this procedure is
advantageous since it employs acetyl as a starting protecting group. Thus, it circumvents
the introduction of other miscellaneous protecting groups that need further deprotection,

followed by N-acetylation, in order to obtain the desired product.

The mechanism that we propose for the transglycosylation reaction is depicted in Scheme
5.2.3. The acetarnido functionality is a very good participating group, and alter zinc
chloride mediated removal of the o-nitrophenyl leaving group, the oxazolinium ion 7 is

formed. Since the conditions of the reaction are acidic and Zn coordinates to both the

114

 

nitro- and phenyl hydroxyl groups, the oxazoline is not formed by abstraction of the

proton from the nitrogen. This explains the absence of oxazoline byproducts. This cyclic
intermediate formed has the oc—face shielded, allowing the attack of the acceptor alcohol

on the B—face only and thus affording the trans-glycoside 8 with total stereoselectivity.

o“ + .-0
OH .31 NV OH
Zr:
HO O ‘- HO 0 OR
H HO
. NHAc
PIN.
0;)
6
H3C ..
RQIV
4 OH
HO 0
HO OzN
HN\ o
69‘c’
I
CH3
5

Scheme 5.2.3 Mechanism of transglycosylation reaction.

115

5.3. Conclusions

Glycosylations are important reactions in carbohydrate chemistry. They are the key step
in the synthesis of complex oligosaccharides and glycoconjugates. Since the outcome of
the reaction is determined by the structure of the substrates, the nature of the substituents,
and the desired stereochemistiy in the products, it is not possible to use one generally
applicable technique. 2-Amino-2-deoxy sugars do not form glycosides when treated with
the conventional acid-alcohol mixtures, owing to the electrostatic shielding effect of the
-NH3+ group, which prevents the approach of cationic groups to the glycosidic center.
Glycoside formation takes place normally when the amino group is suitably blocked, as
with the N-acetyl, N-(benzyloxycarbonyl) or phthalimido derivatives. If an N-acetyl-
glucosamine glycoside is desired, the last two N-protecting groups mentioned need to be
removed so that the amino group can be transformed into an acetamide. Only two
procedures use acetyl protected glucosamine as starting material. The Koenigs-Knorr
method has the disadvantage of using toxic silver or mercury salts and giving only low
yields. The oxazoline method requires harsh reaction conditions, leading to migration and
partial anomerization and affords satisfactory results only for primary alcohols. We
developed a new glycosylation procedure where the leaving group is o-nitrophenyl. This
provides good protection of the anomeric center against bases and can be activated with

ZnClz to render B-glucosamine glycosides in high overall yield.

116

 

5.4. Experimental section
For the subsequent reactions, all solvents were distilled in order to remove traces of
water. The solid reactants were dried in a vacuum oven at room temperature. Potassium

hydride was washed with hexane before weighing.

o-Nitrophenyl-2-acetamido—2-deoxy-3, 4, 6-tri- O-acelyl-fl-D-glucoside (3)

In a 2 neck round bottom flask equipped with a dropping funnel and a nitrogen tube

 

was added potassium hydride (12.8 g, 0.2 mol) as a 60% suspension in mineral oil,
over dichloromethane (80 mL), keeping the suspension in an ice-water bath. To this, a
solution of 18-crown-6 (26.5 g, 0.1 mol) and o-nitrophenol (28.0 g, 0.2 mol) in
dichloromethane (100 mL) was and added through the dropping funnel over 1 h. The
suspension was kept stirring for another 2 h, afier which chloroglucosamine (36.6 g,
0.1 mol) was added and stirring was continued at room temperature for another 2 h.
The reaction was quenched with glacial acetic acid (12 mL) during 0.5 h of mixing.
For workup, 5% NaHCO3 solution (l L) was added and the product was extracted into
the organic layer and dried on the rotary evaporator to give crude 3 (45.0 g), which

was used without further purification in the deacetylation reaction.

o-Nitrophenyl-Z-acetamido-Z-deoxy—fl-D-glucoside (4)

To a solution of 3 in methanol (1050 mL) and water (70 mL) potassium carbonate (55.0
g, 0.4 mol) was added and the suspension was stirred overnight at room temperature.
After removal of the salts by filtration, the solution was rotary evaporated. More
methanol was added to precipitate the excess salt which was then filtered off. By

dissolving the crude reaction mixture in water and rotary evaporating it at 40°C, the o-

117

nitrophenol was removed. Recristallization from hot ethanol gave 4 (18.4 g) in a 54%
overall yield starting from 1. lH-NMR (CDCl3) 8 1.85 (3H, s, NHAc), 3.48 (1H, t, J 8.31
Hz, H-4), 3.55 (1H, t, J2,3=J3,4 8.61 Hz, H-3), 3.57 (1H, m, H-S), 3.72 (1H, dd, J68, 6., 12.37
Hz, J635 5.3 Hz, H-6a), 3.88 (1H, dd, J6” 2.43 Hz, H6-b), 3.92 (1H, dd, H-2), 5.11 (1H,
(1, 11,2 10.16 Hz, H-l), 7.17 (1H, t, J=8.62 Hz, H-c), 7.36 (1H, d, H-a), 7.57 (1H, t, J=8.63,

H-b), 7.77 (1H, d, H-d).

IsopropyI-Z-acetamido-Z-deoxy-fl-D-glucoside (5)

A suspension of o-nitrophenyl-2-acetamido-2-deoxy-B-D-glucoside (4) (0.10 g, 0.29
mmol) and Zan (0.10 g, 0.70 mmol) in dry i-propanol (5 mL) was heated at 70°C, when
the solution became homogeneous and stirring was continued for 22 h. The solvent was
removed and the crude reaction mixture was dissolved in water and rotary evaporated to
distill off the o-nitrophenol. Column chromatography with a chloroform-methanol (6:1
vol) eluent was used to remove the traces of ZnClz. The product 5 (0.068 g) was
recovered in 95% yield. lH-NMR (CDCl3); 5 1.10 (3H, d, J=6.1 Hz,CH3), 1.16 (3H, (1,
CH3), 2.01 (3H, s, NHCH3), 3.40 (2H, m, H-6a, H-6b), 3.46 (1H, m, H-S), 3.56 (1H, dd, J
8.05 Hz, J 11.97 Hz, H-2), 3.62 (1H, dd, J 12.65 Hz, J 7.2 Hz, H-4), 3.71 (1H, dd, H3),

3.98 (1H, m, H-S), 4.56 (1H, d, H-l).

Ethyl lactyl-Z-acetamido-Z-deoxy-fl-D-glucoside (6)

A suspension of o-nitrophenyl-2-acetamido-2-deoxy—B-D-glucoside (4) (0.10 g, 0.29
mmol) and Zan (0.10 g, 0.70 mmol) in dry ethyl lactate (2 mL) was heated at 70°C,
when the solution became homogeneous and stirring was continued for 20 h. The

solvent was removed and the crude reaction mixture was dissolved in water and

118

 

rotary evaporated to distill off the o-nitrophenyl. Column chromatography with a
chloroform-methanol (7:1 vol) eluent was used to remove the traces of ZnClz and
separate the product from the ethyl lactate dimmer. The product 6 (0.06 g) was
recovered in 63.78 % yield. lH-NMR (D20); 5 1.05 (3H, t, J 7.08 Hz, CH3), 1.14 (3H,
d, J 6.84 Hz, CH3), 1.78 (3H, s, NHAc), 3.03 (1H, m, H-5), 3.08, (1H, m, J“ 8.06 Hz,
JNHJ 1.71 Hz, H-2), 3.26 (1H, dd, J23 8.06 Hz, J3,4 10.25 Hz, H-3), 3.37 (1H, dd, H-
4), 3.43 (1H, dd, J6“, 11.97 Hz, 163.5 5.61 Hz, H-6a), 3.65 (1H, dd, H-6b), 3.96 (2H, 2,
J 7.08 Hz, H-d), 4.26 (1H, d, H-a), 4.28 (1H, 2, J 6.84 Hz, H-b); l3C-NMR (D20)

5 8.7, 13.3, 17.7, 50.9, 56.0, 57.8, 65.2, 68.8, 69.0, 71.3, 95.6, 169.9, 170.0.

119

- ;ii;_::‘q

5.5. References

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

. Hakamori, S. Annu. Rev. Biochem. 1981, 50, 733-785.

Hakamori, S. Annu. Rev. Biochem. 1981, 50, 733-785.
Schmidt, R. Angew. Chem. Int. Ed. Engl. 1986, 25, 212-234.
Banoub, J .; Boullanger, P.; Lafont, D. Chem. Rev. 1992, 92, 1167-1195.

Jeanloz, R. W.; Balazs, E. A. The Amino Sugars, Vol II A, Academic Press, 1965, p.
32. ...‘
F

 

1
Koenigs, W.; Knorr, E. Chem. Ber. 1901, 34, 957-968. E
Kunh, R.; Kirschenlohr, W. Chem. Ber. 1953, 86, 1331-1342. "
Lemieux, R. U.; Bundle, D. R.; Baker, D. A. J. Am. Chem. Soc. 1975, 97, 4076-4082.

King, R. R.; Bishop, C. T. Carbohydr. Res. 1974, 32, 239-253.

Salo, W. L.; Fletcher, H. G. J. Org. Chem. 1969, 34, 3189-3195.

Stockl, W. P.; Weidman, H. Carbohydr. Chem. 1989, 8, 169-185.

Kiso, M.; Anderson, L. Carbohydr. Res. 1979, 72, C12.

Lemieux, R. U.; Takeda, T.; Chung, B. Y. A.C.S. Symp. Ser. 1976, 39, 90-104.

Akiya, S.; Osawa, T. Chem. Pharm. Bull. 1960, 8, 583-594.

Baker, B. R.; Joseph, J. P.; Schaub, R. E. Williams, J. H. J. Org. Chem. 1954, 19,
1786.

Paulsen, H.; Paal, M. Carbohydr. Res. 1982, 101, 271-299.

Nicolau, K. C.; Seitz, S. P.; Papahatjis, D. P. J. Am. Chem. Soc. 1983, 105, 2430-
2443.

Lonn, L. Carbohydr. Res. 1985, 139, 105-125.
Schmidt, R. R.; Grundler, G. Angew. Chem, Int. Ed. Engl. 1983, 22, 776-787.
Nakano, T.; Ito, Y.; Ogawa, T. Tetrahedron Lett. 1990, 31 , 1597-1601.

Grundler, (3.; Schmidt, R. R. Carbohydr. Res. 1985, 66, 1651-1671.

120

 

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

Mootoo, D. R.; Fraser-Reid, B. Tetrahedron Lett. 1989, 30, 2363-2367.
Mehta, S.; Pinto, B. M. Tetrahedron Lett. 1991, 32, 4435-4439.
Lemieux, R. U.; Ratcliffe, R. M. Can. J. Chem. 1979, 57, 1244-1249.
Grundler, G.; Schmidt, R. R. Liebigs Ann. Chem. 1984, 1826-1837.
Kanie, 0.; Takeda, T.; Ogihara, Y. Carbohydr. Res. 1989, 190, 53-77.
Paulsen, H.; Stenzel, W. Chem. Ber. 1978, 111, 2334-2341.

Kinzy, W.; Schmidt, R. R. Liebigs Ann. Chem. 1985, 1537-1556.
Schmidt, R. R.; Michel, J. Tetrahedron Lett. 1984, 25, 821-823.

Lonn, L. Carbohydr. Res. 1985, 139, 105-124.

Andreson, F.; Fugedy, P.; Nashed, M. Tetrahedron Lett. 1986, 27, 3919-3923.
Zervas, L.; Konstas, S. Chem. Ber. 1960, 93, 435-456.

Paulsen, H.; Bunsch, A., Carbohydr. Res. 1982, 100, 143-158.
Wang, Y.; Hollingsworth, R. 1., Biochemistry 1996, 35, 5647-5652.

Lemieux, R. U.; Huber, G. Can. J. Chem. 1953, 31-39.

121

 

Chapter 6

Approaches Towards the Total Synthesis of the Disaccharide Head-

Group of Lipid A of Rhizobium T rifolii

122

Err-+494

Abstract
Lipopolysaccharides (LPS) are a unique class of glycolipids found on the surface of
Gram-negative bacteria. LPS are capable of eliciting a wide array of biological responses
when they interact with cellular systems of animals. Recently, the LPS from Rhizobium
T rifolii has been isolated from bacterial extracts and its structure elucidated. The total
synthesis of the disaccharide head-group of lipid A has been attempted in order to prove
the originally assigned structure. The most successful method involves coupling of a
trichloroacetimidate donor sugar with an acceptor benzyl glucosamine having an
unprotected primary hydroxyl, followed by deprotection and radical oxidation. Three

different approaches are described.

123

 

 

6.1. Introduction

Lipopolysaccharides (LPS) were first discovered by Richard Pfeiffer in 1892. He found
that Gram-negative bacteria produce toxins that are distinct from the more commonly
known exotoxins in that they were heat-stable and remained associated with the bacteria.
Due to these unique properties, he called them endotoxins [1]. Since this first discovery, a
tremendous amount of effort went into elucidating the chemical structure of endotoxins.
It was found that they consisted of three regions. The O-antigen region is involved in the
host-to-parasite interactions [2]. The distal core region serves as the linker arm
connecting the O-antigen to the rest of the LPS molecule. The inner core region (rich in
charged groups) appears to maintain the barrier properties of the outer membrane, and,
finally, the lipid A is the hydrophobic moiety that anchors the LPS to the outer

membrane.

Lipid A has a highly conserved structure. When the structure of this glycolipid from
many bacterial sources is examined, certain common features are revealed. They appear
to have the following: (1) B—(l-6)-linked disaccharide backbone containing amino
sugars, (2) five to seven fatty acyl groups of which the sugar-linked acids are 3-hydroxy
fatty acids, (3) acyloxyacyl fatty acids, and (4) one to two phosphate groups at the 1- and
4’-positions. The structural role of lipid A as a hydrophobic anchor is provided by the
abundant fatty acyl groups attached to the sugar. The presence of the acyloxyacyl group
further enhances the hydrophobicity of the lipid A. It is not known why a disaccharide
would be needed instead of a monosaccharide. Perhaps a higher degree of hydrophobicity

is required and can be achieved by a disaccharide, due to the greater capacity for acyl

124

 

 

groups. Amino sugars could provide the more stable amide bond with the fatty acids
rather than the labile ester bond provided by neutral sugars. This might be an advantage
for survival of bacteria. The reason for the presence of hydroxy fatty acids is not clear.
They might be involved in intermolecular hydrogen bonding (i.e. with outer membrane
proteins) to provide firrther stability to the outer membrane by cross-linking adjacent lipid
A molecules via divalent cations. It appears that the entire lipid A molecule is important
for biological activities, since alteration of the previously mentioned characteristics alter

biological activity [3].

The pathophysiological effects of lipid A on the mammalian system have created great
interest [4]. LPS is known to elicit a variety of pathological effects as a result of an
adverse host inflammatory response [5] and lipid A is responsible for triggering these
events [6]. Recently, nontoxic lipid A molecules were isolated from bacterial extracts and
shown to exhibit LPS antagonistic properties [7, 8]. These properties are useful for
investigations into the mechanisms of LPS action and have served as the basis for
speculation about strategies for the possible therapeutic intervention in LPS-related
disease states. It is tempting to develop structure-relationships among lipid A analogs as
an aid toward understanding LPS action at the molecular level. However, such an
approach critically depends upon firm structural information and supplies of
homogeneous materials. Hollingsworth and Wang [9] have recently solved the structure
of Rhizobium Trifolii lipid A. On the basis of degradation and spectral studies, they have
suggested the structure depicted in Figure 6.1.1. Our attempt was the total synthesis of

the head group of lipid A, in order to prove the initially assigned structure.

125

COOH
O
O
O
M
NH
H
H H

Figure 6.1.1 Structure of Rhizobium T rifolii lipid A.

126

“‘51

T"

6.2. Results and Discussions
Three attempts were made to prepare the lipid A head group 1 (Figure 6.2.1) and each

will be described with their successes and failures.

6.2.1. Method 1

The first procedure was designed to take advantage of the new anomeric nitrophenyl
activating group described in chapter 5. The idea was to use o-nitrophenyl 2-acetamido-
2-deoxy-B-D-glucopyranoside 3, and react it with 2,3,4,6-tetra-O-acetyl-0L-D-

galactopyranosyl bromide 4 in the presence of a promoter, to obtain an or/B mixture of

H
OOH
O
H
HO
H 0 CH3
H /
RH
IRNH

COOH

Figure 6.2.1 Assigned structure for the lipid A head-group of Rhizobium T rifolii.

127

 

 

 

 

N
OAc 02 OH OZN
Ac 0 1) H KB, 18 crown6 H 0
Ac > H
AcNH C1 2) K2C039 MCOH, H20 ACNH
2 3
A00 OAc
O
ACO OAc Ac
AcO

O + 3 promoter 0 N + B— 5

Ac 2
O
AcO H
Br H O
4 AcNH
5
l) KZCO3
2) Lactate
ZDCIZ
HO OOH H OH
0 o
H H
HO oxidation H
<— 0
O
H H 0\ /CH3 H O O /CH3
ACNH \ \
COOH AcNH COOH
1 6

 

 

 

Scheme 6.2.1 First attempt at the synthesis of the disaccharide head-group of lipid A.

disaccharides 5. The a—isomer would then be deacetylated and transglycosylated with
lactic ethyl ester, to give disaccharide 6, which would be oxidized to the final product 1,
according to Scheme 6.2.1. The synthesis of compound 3 occured in 65% overall yield
and has been described in detail in chapter 5. It is known that the reactivity of the
pyranose hydroxyl groups decreases in the order 6 OH >> 3 OH > 2 OH > 4 OH. Since
primary hydroxyl groups are much more reactive than secondary ones, we expect the
reaction of the unprotected nitrophenyl glycoside 3 with acetobromogalactose 4 to afford
only the 1-6 disaccharide. Because the sugar acceptor is very polar, the only acceptable
solvent for the reaction proved to be DMF. Unfortunately, acetobromogalactose
decomposes under these conditions before it gets a chance to react with 3. HgO and
Hg(CN); have been used as promoters, but only unreacted nitrophenyl glycoside and

2,3,4,6-tetra—O-acetyl galactose have been isolated from the product mixture (Scheme

6.2.2).
AcO OAc
Hg(CN)2,HgO
DMF
ACO Br
4 3

Scheme 6.2.2 Unsuccessful glycosylation between acetobromogalactose and o-nitro-

phenyl glucosamine.

129

Ac OAc

0
Ac
ACO Br
4
AngO3
CH3N02
OH
H O
H
ACNH ocuzcfin5
7
Ac OAc
0
Ac
Ac 0
AGO AcO
ACNHOCH2C6H5
8-A

Ac OAc
O
A
c AcO H O
H
ACNHOCH2C6H5
8
Ac OAc
O
SHC14 AC
+> AcO
CHC13
Ac 0
Ac
ACNHocnzcéu5
9

Scheme 6.2.3 Synthesis of B-disaccharide 8 and unsuccessful anomerization towards 9.

130

Because of the difficulties described above, the use of a different acceptor molecule was
considered. Benzyl 2-acetarnido-2-deoxy-0L-D-glucopyranoside 7 was reacted with the
same sugar donor 4 in nitromethane and in the presence of silver carbonate [10], to give
solely the B—disaccharide 8 in a 33% yield (Scheme 6.2.3). Other promoters such as
Hg(CN)2-HgO or Hg(CN)2-Ag2COg were considered but no improvement in yield was
obtained. In the first case, using a 3:1 :1 :2 ratio between 4, 7, Hg(CN); and HgO the yield
was 20%, while in the second case, a 24% yield was obtained using a 2.5: 1 :2:2 ratio of 4,
7, Hg(CN)2 and Ag2CO3, We tried to invert the stereochemistry of the disaccharide
linkage by refluxing the peracetylated disaccharide 8-A in chloroform or THF in the
presence of Lewis acids like TiCl4, ZnClz, BF3, SnClz, SnCl4. However, unlike the
anomerization of methyl 2,3,4,6-tetra-O-acety1-B-D-glucopyranoside [11], these reactions
proved unsuccessful (Scheme 6.2.3). Since the B-stereoselectivity of the reaction is due to
the activating effect of the C-2 acetyl group, we considered using a different donor
molecule that has a less participating C-2 substituent, in order to favor
a—stereoselectivity. The compound chosen was 3,4,6-tri-O-acetyl-2-O-trichloroacetyl-[3-
D-glucopyranosyl chloride 11, obtained from pentaacetyl-B-D-galactoside 10 by
treatment with PC15 in CC14 [12-14] (Scheme 6.2.4). In this case, the assisting effect of
the tn'chloroacetyl group at C-2 is minimized by the inductive effect of the three chlorine
groups. This new glycosyl donor was reacted first with the unprotected nitrophenyl
glycoside 3 in the presence of AgOTf, and TMU [15] and also with the ethyl lactyl 3,4-
tri-O-acetyl-2-acetamido-2-deoxy-B-D-glucoside 13, in the presence of AgClO4 and

silver carbonate [16]. Unfortunately no product has been isolated, since 11 decomposed.

131

AC0 0A0 A OAc
0 PC15

.___, 0
Ac 0’“ ca. A, CI
AcO OCOCCl3
A00 OAc
0
Ac
OH
11 H o TMU 02N
+ H + H O
AcNH CH3N02 H O 2: :>
AcNH
3
12
Ac OAc
O
AgC104 c1 coc
11 + A“ O 0 ,CH3 Ag2C03 3
\
A“ R” +>
AcNH COOEt CH Cl Ac 0 [CH3
2 2 AC ‘CH
\
ACNH c0015:

l3
14

Scheme 6.2.4 Synthesis of glycosyl donor 11 and attempted glycosylation reactions.

132

 

 

on OzN
H O
H
AcNH
3
OCPh3
H H O 0 /CH3
\CH
AcNH \
COOH

H3C—$H2-C00Et

 

 

0“ H 0 CH3 Ph3CCl
J; H 0\ / ———>
CH
2102 AcNH \ Py
C003
15
OCPh3
ACZO AC 0 CH3
Ac 0\ /
Py CH
AcNH \
C008
17
HBr
AcOH
OH
Ac 0 0 [CH3
Ac \CH
ACNH \COOEt

 

 

 

Scheme 6.2.5 Synthesis of lactyl 2-acetamido-2-deoxy—3,4-di-O-acetyl-B-D-gluco-

pyranoside ethyl ester 13.

133

A
fl)

 

Compound 13 was synthesized (Scheme 6.2.5) fiom nitrophenyl glycoside 3, which in
ethyl lactate and in the presence of ZnClz gives the B-lactyl glycoside 15. Its primary
hydroxyl group was protected by tritylation [17], followed by acetylation of the
secondary hydroxyls. The intermediate 16 was not isolated because the one pot reaction
gives a higher yield of 17. Deprotection in the presence of HBr and acetic acid gave 13. It
is very important to conduct the detritylation very rapidly, otherwise the peracetylated

side-product will predominate.

6. 2. 2. Method [1

Because the former approach proved unsuccessful, a totally different strategy was used.
2,3,4,6-Tetra-O-benzyl-B-D-glycopyranosyl trichloroacetimidate 18 was chosen as donor,
while benzyl-2-acetamido-2-deoxy-3,4-di-O-acetyl-B-D-glucopyranoside 19 served as
acceptor. Scheme 6.2.6 summarizes their synthesis. Methyl a—D-galactopyranoside 20
was perbenzylated with benzyl chloride in the presence of sodium hydride and then the
anomeric hydroxyl group deprotected [18] under acidic conditions. Reaction of benzyl
hemiacetal 22 with trichloroacetonitrile in the presence of potassium carbonate gave
trichloroacetimidate 18 in very good yield. The acceptor 19 was obtained from
benzylglucosamine 7 after tritylation, acetylation and detritylation using NaI and
trimethylsilyl chloride [19]. This new detritylation method proved to be much better than
the one that involves HBr and AcOH, since peracetylation is avoided. It seems that a
(trimethylsilyl)t1ityloxonium ion intermediate is formed, that is displaced by iodide. The
trichloroacetimidate 18 was reacted with the benzyl glucosamine 19 in the presence of

trimethylsilyl triflate [20] to give the a-disaccharide 25 in 20.3% yield afier

134

H OH BnO OBn

 

 

 

 

 

st04
O BnBr,NaH O
H DMF Bn ACOH
OHOMe Bno OMe
20 21
En OBn
Bn OBn C13CCN o IH
o K2C03, Bn o-—c—cc13
Bn c112c12
BnO
B110 OH
13
22
0“ ooph3
0 Ph CCl Ac 0
H 3 H O 2 I
H H py
AcNH Py AcNH
an OBn
7 23
oc1>h3 on
Ac 0 Nal, Me3SiCl Ac 0
Ac ——> Ac
AcNH CH3CN A
OBn C OBn
24 19

 

 

 

Scheme 6.2.6 Synthesis of donor 18 and acceptor sugar 19.

135

NH
0 II
B O—C—CC13
Bn
l8
Me3Sin
OH '
Et20/CH2CIZ
Ac 0 20%
A
ACNHOBn
19
OH OH
O
H
OH
H , Pd/C
2 O ACCI
MeOH O 63%
EtOAc AC
88°/ Ac
0
AcNH OH
26
OAc OAC
O
ACO 1) K2CO3
Lactate Ac MeOH
—-—> O —)
AngO3 2) Oxidation
AgZO Ac 0 0 /CH3
Ac \CH
AcNH COOEt
28

8:001)

25
OAc OAc

AcNH

 

 

OCOOH
H Hol

CH3
“mac/H
\
M H coon

 

Scheme 6.2.7 Second attempt at the synthesis of the disaccharide head-group of lipid A.

136

 

chromatographic separation (Scheme 6.2.7). Catalytic hydrogenolysis removed all the
benzyl groups and also deprotected the anomeric center. Peracetylation with acetyl
chloride was accompanied by activation of the reducing end of the disaccharide,
affording 27. This has been glycosylated with ethyl lactate in the presence of AgZO and
AngO3 to afford B-lactyl galactosyl glucoside 28. Unfortunately, the total synthesis
could not be completed because compound 28 was not isolated in sufficient amount to
undergo deacetylation and oxidation. However, oxidation can easily be accomplished by
using TEMPO (2,2,6,6-tetramethyl-l-piperidinyl-l-oxy) in the presence of NaBr and
NaOCl [21]. TEMPO is a stable radical, which can be oxidized by several agents to give
a nitosonium ion. The latter is a strong oxidant and shows selectivity towards primary
hydroxyl functions over secondary ones. The reaction has been tested on methyl-(x-D-

galactopyranoside 20 and produced 29 in quantitative yield (Scheme 6.2.8).

OH OH H
o TEMPO 00%
H OH NaBr, NaOCl H OH
OMe OMe
20 29

Scheme 6.2.8 Oxidation of methyl OL-D-galactoside with TEMPO.

137

 

 

 

6. 2.3. Method 111

Since the glycosylation with ethyl lactate gives only low yields, it would be better if it
would occur early in the synthesis, so that after multiple reaction steps, there will be
enough material for charachterization at the end of the synthesis. The idea was to couple
the trichloroacetimidate 18 with the lactyl glycoside 34 to give the a-disaccharide,
followed by catalytic hydrogenation, deacetylation and oxidation (Scheme 6.2.9). A

different approach for the synthesis of the sugar acceptor started from 2-acetamido-2-

 

deoxy-3,4,6-tri-O-acetyl-a-D-glucopyranosyl chloride 2 which is transformed into the [3-
lactyl glycoside 31 in the presence of Ag2C03 and AgzO. Afier deacetylation, the lactic
acid is reesterified with MeOH and BF3 to give 33. Tritylation, peracetylation followed

by detritylation with NaI affords 34. The rest of the synthesis is still to be completed.

 

 

 

Bn OBn on
NH
0 ll 0 '
O—C—CCb + A 0 ,CH3 M_°35‘_T‘,
B Ac \CH EtZO
3“ AcNH \COOMe
18 34
H
Bn OBn COOH
‘ o o
B 1) H2, Pd/C H
Bn 2) K2C03 H
3) TEMPO
——>
0
Ac 0 CH3 H \ [CH3
Ac O\C’H H CH
\ AcNH \COOH
AcNH COOMe
30 1

 

Scheme 6.2.9 Third attempt at the synthesis of the disaccharide head-group of lipid A.

138

0A0 H3C—CIZH2-COOEt 0A0

 

0 K co
Ac 0 OH ‘ Ac 0\ [CH3 2 3
Ac 7 AC C\H MeOH
AcNH 1 AgZCO3 ACNH c0013: H20
AgzO
2 31
OH OH
H H O 0 /CH3 MeOH,BF3 H O 0 [CH3
———>
A NH ‘RH H \R“
C
COOH AcNH COOMe
32 33

(l) Ph3CCl, Py
(2) AC20, Py

(3) HBr, AcOH
or NaI, Me3SiCl

 

OH
0
Ac 0 [CH3
Ac \CH
\
AcNH COOMe
34

 

 

 

Scheme 6.2.10 Synthesis of lactyl 2-acetamido-2-deoxy—3,4-di-O-acetyl-B-D-gluco-

pyranoside methyl ester 33.

139

F"

hum
I. A
.

 

5‘

 

6.3. Conclusions

Lipopolysaccharides are a unique class of glycolipids found on the surface of Gram-
negative bacteria. LPS are capable of eliciting a wide array of biological responses when
they interact with cellular systems of animals. Recently, the structure of lipid A of
Rhizobium T rifolii has been determined by degradation of bacterial extracts and
spectroscopic studies. Its headgroup consists of a galacturonic acid 1(a)-6 linked to
glucosamine. The reducing end of the glucosamine is coupled to lactic acid. We
attempted to chemically synthesize this disaccharide. In one method we tried to couple
tetraacetobromogalactose to an o-nitrophenyl glucosamine acceptor. The high polarity of
the acceptor molecule and the B-orienting effect of the donor made this synthesis
unsuccessful. A different approach used tetrabenzyl B-D-galactopyranosyl trichloro-
acetimidate as a nonparticipating donor while benzyl 2-acetamido-2-deoxy-3,4-di-O—
acetyl-a-D-glucopyranoside with only the 6-hydroxyl group unprotected was the
acceptor. The a—disaccharide was produced, but glycosylation with lactic ethyl ester

proved to be a very low yield reaction, preventing us fiom reaching the final target.

6.4. Experimental section

o-Benzy, - 6- 0— (2, 3, 4, 6-tetra-0-acetyl— a—D—galactopyranosyl)-2-acetamido-2-deoxy-fl-D-
glucopyranoside (8)

In a 25 mL stoppered round bottom flask, dried benzyl 2-acetamido-2-deoxy-0t-D-
glucopyranoside 7 (0.10 g, 0.32 mmol) was dissolved in dry nitromethane (6 mL). Silver
carbonate (0.22 g, 0.79 mmol) and calcium sulphate (0.50 g) were added and the

suspension was stirred in the dark at room temperature for 5 h. A solution of bromo

140

2,3,4,6-tetra-O-acetyl-0t-D-galactopyranoside 4 (0.39 g, 0.95 mmol) dissolved in
nitromethane (0.5 mL) was added at 60°C over a period of 10 min., after which the
temperature returned to room temperature. Stirring was continued for 48 h. The
suspension was filtered through Celite and washed several times with chloroform and
acetone. The solution was rotary evaporated and the crude product stirred with
chloroform. The disaccharide and unreacted or decomposed bromoacetogalactose
dissolved in chloroform while the benzylglucosamine precipitated and was filtered off.
The solution was rotary evaporated and the product was isolated by flash column
chromatography, using 5:1 (vol) eluent of acetone-hexane as the eluent. After
evaporating the fractions, 3 (0.07 g) was obtained in 33.35% yield. lH-NMR(CDC13)
5 1.95 (3H, 5, OAc), 2.01 (3H, s, NHAc), 2.02 (3H, 3, OAc), 2.09 (3H, 3, OAc), 2.14 (3H,
5, OAc), 3.45 (1H, m, H-5), 3.65 (1H, m, H-S’), 3.70 (1H, dd, H-6a), 3.75 (1H, dd, H-6b),
3.76 (1H, m, H-2), 3.90 (1H, t, .123 6.7 Hz, H-3), 4.03 (1H, t, 13,4 6.7 Hz, H-4), 4.10 (1H,
dd, 1635 6.98 Hz, 16m, 11.16 Hz, H-6a’), 4.16 (1H, dd, Jams 7.26 Hz, H-6b’), 4.41 (1H, d,
J 11.72 Hz, H-bn), 4.54 (1H, d, J 7.82 Hz, H-l’), 4.71 (1H, d, I 11.72 Hz, H-bn), 4.83
(1H, d, .1 3.63 Hz, H-l), 5.00 (1H, dd, J33 10.3 Hz, J34 3.35 Hz, H-3’), 5.21 (1H, dd, H-
2’), 5.35 (1H, dd, H-4’); l3C-NMR (CDCl3) 820.4, 20.4, 20.5, 20.6, 23.1, 53.4, 61.1,
66.9, 68.6, 68.9, 69.2, 69.4, 70.4, 71.6, 74.0, 96.0, 101.6, 128.0, 128.2, 128.4, 137.0,

170.0, 170.3, 170.4, 170.9, 172.0.

3, 4, 6-tri- O-acetyl-Z- 0-trichIoroaceozl-fl-D-glucopyranosyl chloride (1 1)
A suspension of B-pentaacetylgalactose (3.0 g, 7.6 mmol), phosphorus pentachloride (6.0

g, 28.0 mmol) and carbon tetrachloride (1.5 mL, 15.5 mmol) was refluxed for 5 h. The

141

mixture was poured into ice and water (30 mL) and extracted three times with
dichloromethane (10 mL). The combined organic extracts were successively washed with
water, saturated aqueous NaHCO3 solution and then again with water. After drying over
Na2504, the solution was filtered and the filtrate was evaporated under reduced pressure
at 40°C to give a residue that was subjected to column chromatography. The eluent was a
1:11 solution of EtOAc—CHCI3. After recrystallization from ether, 11 (0.36 g) was
obtained in 10.33% yield. lH-NMR(CDC13) 5 1.99 (3H, 5, OAc), 2.07 (3H, 8, OAc), 2.20
(3H, 8, OAc), 4.07 (1H, dd, J 6.59 Hz, H 7.08 Hz, H-6a), 4.19 (1H, d, H-6b), 4.19 (1H, d,
‘ J 6.59 Hz, H-l), 5.18 (1H, m H-5), 5.40 (1H, t, H-2), 5.41 (1H, dd, H-3), 5.48 (1H, dd, J
3.17 Hz, H-4); l3C--NMR (CDCl3) 8 20.1, 20.4, 20.5, 60.9, 66.7, 70.2, 74.6, 75.1, 86.7,

160.3, 169.3, 169.8, 170.1.

(D-ethyl propionyl) 2-acetamido-2-de0xy-3,4-di-O-acetyl-6-triphenylmethyI-fl-D-
glucopyranoside (17)

In a round bottom flask, lactyl glucosamine 15 (1.34 g, 4.17 mmol) was dissolved in
pyridine (7.5 mL). Triphenyl methylchloride (1.22 g, 4.30 mmol) and dimethyl
aminopyridine (DMAP) (0.046 g, 0.370 mmol) were added, and the solution was stirred
at 40°C for 14 h. 16 was obtained, and used without isolation in the following reaction.
After cooling the reaction mixture to room temperature, acetic anhydride (1.9 mL) was
added and stirring was continued for another 2 h. The solution was poured slowly and
with stirring over ice and water (100 mL), extracted with chloroform, washed several
times with ice/water and dried with Na2S04. The chloroform layer was rotary evaporated

and the product was recrystallized from pure ethanol to give 17 (2.27 g) in an 84.34%

142

 

 

total yield. 'H-NMR (CDCl3) 81.30 (3H, t, J 7.33 Hz, OEt), 1.52 (3H, d, J 7.08 Hz,
OCHCH3), 1.72 (3H, s, NHAc), 2.00 (3H, 5, OAc), 2.02 (3H, s, OAc), 3.08 (1H, dd, 16,,
10.74 Hz, J6,,5 4.88 Hz, H-6a), 3.23 (1H, dd, 16,5 2.24 Hz, H-6b), 3.49 (1H, m, H-5), 4.06
(2H, q, OCHz), 4.21 (1H, dd, J 8.31 Hz, J 8.06 Hz, H-2), 4.48 (1H, q, OCHCH3), 4.57
(1H, d, H-l), 5.06 (1H, t, J 9.77 Hz, H-3), 5.17 (1H, t, H-4), 5.94 (1H, d, NH), 7.26 (3H,

t, H-ph), 7.44 (1H, d, H-ph).

(2-Ethy1 propionyl) 2-acetamido-2-deoxy-3,4-di-O-acetyl—fl-D-glucopyranaside (13)

The tritylated lactylglucosamine 17 (2.27 g, 3.50 mmol) was dissolved on a steam bath in
glacial acetic acid (13.6 mL). The clear solution was cooled to 10°C and HBr (1.5 mL)
was added, shaking the flask for 40 seconds. The trityl alcohol fell out of solution and the
solution was filtered immediately through Celite and washed with acetic acid into a filter
flask that contained ice and water (100 mL). The detritylated product was extracted with
chloroform, washed with water several times, dried with Na2SO4 and rotary evaporated at
35°C. Chromatographic separation using a 1:1 eluent of chloroform-ethyl acetate gave 13
(0.3 g) in a 21.4% yield. lH-NMR (CDCl3) 8 1.26 (3H, t, J 7.08 Hz, OCHZCH3), 1.40
(3H, d, J 6.84 Hz, OCHCH3), 1.94 (3H, s, NHAc), 2.00 (3H, s, OAc), 2.01 (3H, s, OAc),
32.14 (3H, s, OAc), 3.46 (1H, m, H-5), 3.67 (1H, dd, H-6a), 3.9 (1H, dd, J 1.95 Hz, J
12.69 Hz, H-6b), 3.95 (2H, q, OCH2CH3), 4.16 (1H, m, H-2), 4.39 (1H, q, OCHCH3),
4.58 (1H, d, J 7.57 Hz, H-l), 4.99 (1H, t, J 9.77 Hz, H-3), 5.15 (1H, t, J 10.50 Hz, H-4),
6.29 (1H, d, J 7.81 Hz, NH); 13C--NMR(CDC13)5 14.1, 18.5, 20.6, 70.7, 23.3, 54.2, 61.4,

62.0, 68.2, 72.2, 72.4, 73.3, 99.8, 127.2, 127.9, 143.1, 170.8, 170.9, 173.1.

143

Methyl 2,3,4, 6-tetra-0-benzyl—a—D—galactopyranoside (21)

NaH in mineral oil (6.72 g, 0.28 mol) was added to a 1000 mL round bottom flask. After
washing 5 times with hexane to remove the mineral oil, anhydrous DMF (200 mL) was
added and the suspension was stirred at 60°C for 45 min. The methyl galactoside 20 (7.76
g, 0.04 mol) was added and the reaction mixture stirred for 3 h, lowering the temperature
gradually to 15°C. Benzyl bromide (23.75 mL, 0.2 mol) was slowly added from a
dropping funnel with the flask cooled in an ice-water bath. The temperature of the
reaction mixture was increased to 40°C and stirring continued overnight. The solvent was
rotary evaporated and the oily mixture separated by flash chromatography using 10:1
chloroform-hexane as the eluent. 21 (19.9 g) was obtained in 90.5% yield as a yellowish
syrup. lH-NMR (CDCl3) 8 3.37 (3H, s, OCH3), 3.52 (1H, d, J 6.50 Hz, H-2), 3.91 (1H,
dd, H-6a), 3.92 (1H, m H-S), 3.95 (1H, m, H-3), 4.04 (1H, dd, J 11.03 Hz, J 3.42 Hz, H-
6b), 4.39 (1H, d, J 11.72 Hz, H-bn), 4.48 (1H, d, H-bn), 4.56 (1H, d, J 3.42 Hz, H-4),
4.58 (1H, d, J 11.47 Hz, H-bn), 4.69 (1H, d, H-l), 4.69 (1H, d, J 12.2 Hz, H-bn), 4.74
(1H, d, H-bn), 4.84 (1H, d, J H-bn), 4.85 (1H, d, H-bn), 4.95 (1H, d, H-bn), 7.30 (1H, d,
H-bn); l3C-NMR(CDC13)5 55.0, 69.2, 73.2, 73.4, 74.9, 75.2, 78.8, 78.9, 79.0, 79.2, 98.9,

127.6, 127.8, 128.1, 137.9, 138.4, 138.5, 138.7.

2,3, 4, 6- T etra-O-benzyl- a-D-gacactopyranoside (22)

Methyl 2,3,4,6-tetra-O-benzyl-a—D-galactopyranoside (21) (17 g, 0.03 mol) was
dissolved in glacial acetic acid (315 mL) and heated at 100°C [22]. A solution of 2N
H2804 (68 mL) was added and refluxed for 1 h. A second portion of sulfuric acid was

added and the solution was left at 100°C with stirring, overnight. The reaction mixture

144

 

 

was cooled and poured into water (2.5 L) and left at RT for 2 days. After extraction with
chloroform and drying, 22 (16.2 g) was obtained in 100% yield. ”C-NMR (CDC13)
555.7, 69.4, 69.5, 73.5, 73.7, 73.8, 75.1, 75.4, 78.0, 79.3, 99.1, 127.8, 127.9, 128.0,

128.1, 128.4, 128.5, 128.6, 128.6, 128.7, 130.4, 138.2, 138.8, 138.9, 139.1.

2, 3, 4, 6- T etra-O-benzyl- a—D-galactopyranosyl trichloroacetimimdate (18)

To a solution of 22 (9.0 g, 0.016 mol) in dry CHzClz, K2CO3 (8.65 g) and C13CCN (8.65
mL) were added [23]. The suspension was vigorously stirred at room temperature for 17
h. The reaction mixture was filtered through Celite and washed with chloroform. The
mother liquors were rotary evaporated and 18 (11.4 g) was obtained as a yellow syrup in
quantitative yield. l3C-NMR(CDC13)8 66.2, 68.0, 72.9, 73.2, 73.3, 74.2, 74.6, 75.1, 77.9,
82.0, 98.6, 127.4, 127.6, 127.7, 127.7, 127.9, 128.0, 128.1, 128.3, 128.4, 137.7, 138.1,

138.3, 161.4.

Benzyl 2-acetamido-2-deoxy-3,4-di-0-aceo/l-6-0-tri021-OLD-glucopyranoside (24)

Benzyl 2-acetamido-2-deoxy-a-D-glucopyranoside 7 (10.0 g, 32.0 mmol), DMAP (0.35
g, 2.80 mmol) and Ph3CC1 (10.0 g, 35.9 mol) were dissolved in pyridine (50 mL) and the
resulting solution was stirred at 40°C for 48 h. The reaction mixture was cooled and after
addition of acetic anhydride (30 mL), stirring was continued at room temperature for 4 h.
The solution was poured over ice and water (800 mL) with strong agitation. The crude
product was extracted into chloroform and the chloroform layer dried over Na2S04. The
mixture of 24 and peracetylated benzylgalactoside was not purified any more, but used in

the following reaction without purification. IH-NMR (CDC13) 5 1.93 (6H, 8, OAc), 2.06

145

(3H, s, NHAc), 3.32 (1H, dd, 16,, 10.01 Hz, 16,, 5.86 Hz, H-6a), 3.40 (1H, dd, 161,, 2.93,
H-6b), 3.55 (1H, 1,13,. 8.79 Hz, 14,5 9.77 Hz, 114), 3.72 (1H, dd, 13,2 10.5 Hz, H-3), 3.86
(1H, m, H-5), 4.15 (1H, m, J21»: 8.78 Hz, 12,, 3.91 Hz, H-2), 4.50 (1H, d, J 11.72 Hz, H-
bn), 4.83 (1H, d, H-bn), 4.95 (1H, d, J 391,111), 6.23 (1H, d, NH), 7.28 (20H, m, H-bn);
”C-NMR (CDCl3) 523.1, 53.4, 63.5, 69.0, 71.0, 71.8, 73.3, 81.6, 86.4, 96.3, 124.0,

126.0, 172.0.

Benzyl 2-acetamido-2-deoxy-3,4-di-O-acelyl- a—D—glucopyranoside (l9)

Crude 24 (12.4 g, 19.4 mmol) was dissolved in acetonitrile (145 mL). Afier addition of
NaI (5.1 g, 34.0 mmol) and Me3SiCl (4.37 mL, 34 mmol), the solution was stirred at
room temperature for 15 min. Water (115 mL) was added and the mixture stirred for
another 15 minutes at 0°C. The precipitated trityl alcohol was removed by filtration and
the filtrate extracted with CHzClz, after which the organic phase was washed with
aqueous 10% sodium thiophosphate. The aqueous phase was washed 2 times with
dichloromethane, and then the combined CHzClz layers were rotary evaporated.
Chromatographic separation with 4:1 hexane-acetone gave pure 19 (3.7 g) in 50.4% total
yield. lH-NMR (CDCl3) 6 1.87 (3H, 3, OAc), 1.99 (3H, 3, OAc), 2.02 (3H, s, NHAc),
3.56 (1H, dd, J63), 12.69 Hz, J63; 4.15, H-6a), 3.61 (1H, dd, J6b,5 2.2 Hz, H-6b), 3.77 (1H,
m, H-5), 4.31 (1H, ddd, H-2), 4.49 (d, 1H, J 11.72 Hz, H-bn), 4.70 (1H, d, H-bn), 4.92
(1H, (1, J13 3.67, H-l), 5.04 (1H, t, J 10.01 Hz, H-4), 5.26 (1H, t, H-3), 5.78 (1H, d, J 9.52
Hz, NH), 7.3 (5H, m, H—bn); l3C-NMR (CDCl3) 6 20.5, 20.6, 23.0, 51.8, 60.8, 68.4, 69.9,

70.0, 70.9, 96.4, 128.0, 128.2, 128.5, 136.5, 169.9, 170.2, 171.2.

146

Benzyl-6—0-(2,3,4, 6-tetra-0-benzyl—a-D-galactopyran0sy1)-2-aceatamido-2-deoxy-3, 4-
di-O-acetyl- a—D-glucopyranoside (25)

The trichloroacetimidate donor 18 (1.4 g, 3.5 mmol) and the acceptor sugar 19 (5.0 g, 7.3
mmol) were dissolved in a mixure of diethylether (30 mL) and dichloromethane (5 mL).
Me3Sin (0.3 mL, 1.6 mmol) was added and the solution stirred at RT for 10 h. The
reaction mixture was treated with excess solid NaHCO; and then with ether/NaHCO3
solution in water. The ether extract was washed with water, dried with Na2804 and then
concentrated. The crude product was separated by chromatography with a 3:1 hexane-
acetone eluent. Pure 25 (0.65 g) was obtained in 20.21 % yield. lH-NMR (CDCl3) 5 1.89
(3H, 8, OAc), 1.96 (3H, 5, OAc), 2.00 (3H, s, NHHc), 3.46 (1H, dd, J6a,b 10.7 Hz, J6a,5
2.0 Hz, H-6a), 3.50 (1H, d, J 6.35 Hz, H-4’), 3.74 (1H, dd, J 10.98 Hz, H-3’), 3.80 (1H,
m, H-5), 3.95 (1H, dd, J 2.69 Hz, H-6a’), 3.97 (1H, m, H-5’), 4.07 (1H, dd, J 4.2 Hz, J
10.6 Hz, H-6b’), 4.30 (1H, dd, J 3.66, H-2), 4.37 (1H, d, J 11.72 Hz, H-bn), 4.47 (1H, d, J
11.96 Hz, H-bn), 4.57 (1H, d, J 11.47 Hz, H-bn), 4.67 (1H, d, J 11.48 Hz, H-bn), 4.71
(1H, d, J 9.52 Hz, H-bn), 4.79 (1H, d, H-2’), 4.81 (1H, d, J 3.42 Hz, H-1 ’), 4.85 (1H, d, J
3.66 Hz, H-l), 4.93 (1H, d, J 11.48 Hz, H-bn), 5.09 (1H, t, J 9.77 Hz, H-4), 5.22 (1H, t,
H-3), 5.64 (1H, d J 9.53 Hz, NH); l3C-NMR(CDC13) 8 20.5, 20.6, 23.0, 51.6, 66.3, 68.6,
68.7, 68.8, 69.4, 69.5, 71.5, 73.0, 73.3, 74.6, 75.0, 76.2, 78.6, 95.9, 97.6, 127.3, 127.4,
127.6, 127.6, 127.7, 128.1, 128.2, 128.3, 124.4, 128.5, 136.4, 137.8, 138.5, 138.6, 169.5,

169.4, 171.3.

147

 

 

6-0-(a-D-galactopyranosyI)-2-aceatamido-Z-deoxy-3,4-di—0-acelyl-a-D-
glucopyranoside (26)

In a Pair reactor, benzyl disaccharide 25 (0.65 g, 0.7 mmol) was dissolved in a mixture
of ethyl acetate (50 mL) and methanol (50 mL), to which Pd/C (0.15 g) was added. The
suspension was stirred at room temperature and 200 psi H2 for 18 h. The reaction mixture
was filtered through Celite and the filtrate rotary evaporated to give crude product.
Recristallization from ethanol afforded pure 26 (0.35 g) in 88.40 % yield. ”C-NMR
(CDCl3) 5 20.6, 20.7, 53.0, 62.5, 66.5, 68.7, 70.1, 70.9, 72.2, 72.5, 78.8, 79.1, 79.6, 92.3,

100.0, 171.7, 179.9.

6- 0- (2, 3, 4, 6-tetra- O-acetyl- a-D-galactopyranosyD-2-aceatamido-2-deoxy-3 , 4 -di- 0-
acetyl- a-D-glucopyranosyl chloride (27)

In a 25 mL round bottom flask equipped with a condenser and drying tube, the dry
disaccharide hemiacetal 26 (0.35 g, 0.62 mmol) was dissolved in acetyl chloride (2 mL)
and the solution was stirred at room temperature for 40 h. After adding dichloromethane
(5 mL), the solution was poured over ice and water (5 mL) in a separatory funnel and
shaken. The organic layer was poured rapidly into another separatory funnel containing
ice and saturated sodium bicarbonate solution, and extracted without delay. The organic
layer was then poured into a flask that contained MgSO4 for drying. After decanting, the
solution was dried on the rotary evaporator at 50°C. Over the still warm solution, ether
was added and crystals of 27 formed (0.3 g) in 62.9 % yield in a 1:1 mixture with

peracetylated disaccharide.

148

 

(Z-Ethyl propionyl) 6-0-(2,3,4,6-tetra-0-acetyl-a—D-gaIactopyranosyl)-2-acetamido-2-
deoxy-3,4—di—0-acetyl- a—D-glucopyranoside (28)

To a 25 mL round bottom flask equipped with a drying tube, glycosyl chloride 27 (0.30
g, 0.39 mmol) was dissolved in lactic acid ethyl ester (7 mL) and CaS04 (0.1 g) was
added to remove any residual moisture. After stirring for 30 minutes, AgZO (0.15 g, 0.64
mmol) and Ag2C03 (0.3 g, 1.1 mmol) were added and stirring was continued at room
temperature for 48 h. The suspension was filtered through Celite, washed with
dichloromethane and then evaporated. The crude product was purified using column
chromatography with a 2:1 mixture of hexane-acetone as the eluent. Pure lactyl glycoside
(0.15 g) was obtained. l3C-NMR (CDC13) 5 14.0, 20.6, 22.5, 23.0, 29.6, 50.9, 51.7, 61.3,

61.7, 66.3, 67.4, 67.9, 68.2, 68.7, 69.6, 71.2, 72.4, 96.0, 99.7, 169.2, 169.9, 171.3.

I-O-Methyl- a—D-galacturonic acid (29)

Methyl a-D-galactopyranoside 20 (0.1 g, 0.5 mmol) was dissolved in water (10 mL). To
this was added NaBr (0.025 g, 0.25 mmol), NaClO (0.525 mL, 1.1 mmol) and TEMPO
(0.0008 g, 0.005 mmol) and the solution was strongly stirred at 0°C and pH 11. The pH
was checked every 5 minutes and NaClO was added to maintain pH 11. After 20 minutes,
the pH was stabilized and the reaction completed. The reaction was quenched with
ethanol (0.5 mL) and then neutralized to pH 7 by adding 4 N HCl. The solution was
concentrated at 35°C to about 1 mL and then ethanol was added to precipitate the
product, which was isolated by centrifugation. After drying pure l-O-methyl galacturonic
acid (0.1 g) was obtained in quantitative yield. lH-NMR (CDC13) 5 3.4 (3H, s, OMe), 3.6

(1H, dd, H-6), 3.7 (1H, dd, H-2), 3.72 (1H, m, H-5), 3.75 (1H, m, H-3), 3.8 (1H, dd, H-

149

 

6), 4.24 (1H, dd, H-4), 4.9 (1H, d, H-l); l3c-NMR (CDC13) 5 51.0, 63.5, 64.3, 66.0, 66.3,

95.0, 172.0.

2-acetamid0-2-deoxy—3, 4, 6-tri-0- a—D-glucopyranosyl chloride (2)

2-Acetamido-2-deoxy glucose (50.0 g, 226.0 mmol) and acetyl chloride (100 mL) were
added to a 500 mL round bottom flask equipped with a magnetic stirrer, reflux condenser
and drying tube. The mixture was stirred at room temperature for 16 h. Dichloromethane
(400 mL) was added through the top of the condenser and the resulting solution was
poured with vigorous stirring on ice (400 g) and water (100 mL) in a 3 L beaker. The
organic solution was drawn off without delay into a 3 L beaker containing ice and
saturated sodium bicarbonate solution (400 mL). After neutralization wass complete, the
organic layer was run directly into a flask containing anhydrous MgSO4 (25 g) and stirred
for 10 minutes. The suspension was filtered and washed with dry dichloromethane
directly into a 1 L round bottom flask, and then concentrated on a rotary evaporator at
50°C to about 75 mL. Dry ether (500 mL) was added to the warm solution, rapidly and
with swirling, and the product crystallized out (63 g) in a 76 % yield. lH-NMR (CDCl3)
5 1.96 (3H, 3, OAc), 2.02 (3H, 8, OAc), 2.03 (3H, s, OAc), 2.08 (3H, s, OAc), 4.12 (1H,
dd, H-6a), 4.24 (1H, m, H-5), 4.25 (1H, dd, H-6b), 4.51 (1H, ddd, J 9.77 Hz, J 8.3 Hz, J
3.9 Hz, H-2), 5.19 (1H, J 9.77 Hz, H-4), 5.29 (1H, t, H-3), 5.8 (1H, (1, NH), 6.15 (1H, d, J
3.41 Hz, H-l); l3C-NMR (CDC13) 5 20.4, 20.6, 23.0, 53.3, 61.0, 66.8, 70.0, 70.8, 93.5,

169.1, 170.1,170.5, 171.4.

150

 

 

(Z-Ethyl propionyl) 2-acetamido-2-de0xy-3,4,6-tri-O-acetyl-fl-D-glucopyranoside (31)

In a round bottom flask covered with aluminum foil, AgzO (7 g, 30 mmol), AngO3 (15.0
g, 54.4 mmol) and CaSO4 (15 g), all dried in the vacuum oven for l h, were added to
lactic acid ethyl ester (200 mL) and the suspension was stirred for 1.5 h at room
temperature. N-acetylglucosamine 2 (15.0 g, 41 mmol) was added and stirring is
continued for 24 h. The reaction mixture was filtered through Celite and washed with
chloroform. To eliminate all Ag salts, the residue was dissolved in chloroform and
extracted 2 times with the same volume of aqueous 2 % NHaOH and then with water.
Afier drying over Na2804, the solution was rotary evaporated. Afier one day, pure 31
crystallized out (3.3 g) in 17 % yield. lH-NMR (CDCl3) 5 1.29 (3H, t, OCH2CH3), 1.44
(3H, d, OCH), 1.98 (3H, 8, OAc), 2.00 (3H, s, OAc), 2.02 (3H, s, OAc), 2.08 (3H, 5,
OAc), 3.64 (1H, m, J 2.14 Hz, J 4.39 Hz, J 9.77 Hz, H-5), 4.01 (1H, q, J 8.30 Hz, H-lac),
4.10 (1H, dd, J 12.45 Hz, H-6a), 4.21 (1H, ddd, H-2a), 4.24 (1H, dd, H-6b), 4.40 (2H, q, J
7.08 Hz, H-lac), 4.56 (1H, d, J 8.55 Hz, H-l), 5.09 (1H, t, J 9.52 Hz, H-4), 5.13 (1H, t, J
9.79 Hz, H-3), 6.10 (1H, t, J 7.82 Hz, NH); l3C-NMR (CDC13) 5 14.1, 18.5, 20.5, 20.7,

23.3, 54.1, 61.3, 61.9, 68.2, 72.1, 72.4, 73.2, 99.7, 169.2, 170.8, 173.0.

Lactyl 2-acetamido-Z-deoxy—fl-D-glucopyranaside (32)

To a solution of 31 (3.34 g, 7.7 mmol) in methanol (67 mL) was added a solution of
K2CO3 (3.34 g) in water (5 mL) and the suspension was stirred at room temperature for
24 h. Afier drying the solvent on a rotary evaporator at 30°C, water was added and the
solution was passed over a cation exchange strongly acidic resin. Lyophilization afforded

pure product (2.25 g) in quantitative yield. lH-NMR (D20) 5 1.21 (3H, d, J 7.08 Hz, H-

151

 

lac), 1.83 (3H, s, OAc), 3.22 (1H, m, H-3), 3.23 (1H, m, H-S), 3.34 (1H, m, H-5), 3.47
(1H dd, J 1.46 Hz, J 6.60 Hz, H-6a), 3.49 (1H, dd, J 2.93 Hz, H-6b), 3.70 (1H, d, H-4),
4.32 (1H, q, J 7.08 Hz, H-lac), 4.38 (1H, d, J 8.31 Hz, H-l); l3c-NMR (1320) 513.4,

17.6, 50.8, 56.0, 65.1, 68.8, 69.0, 71.2, 95.7, 170.1, 171.8.

(2-Methyl propionyl) 2-acetamid0-2-deoxy-flD-glucopyranoside (33)

32 (2.26 g, 7.7 mmol) was dissolved in methanol (70 mL) and BF3 etherate (7 mL) was
added. The mixture was stirred at room temperature for 2 h. The solution was dried on a
rotary evaporator several times by adding larger amounts of methanol. Afier
lyophylization, crude product (2.2 g) was obtained. Recrystallization from 1:1
ethanol/ether gave pure 33 (1.2 g) in a 50 % yield. ”C-NMR (1)20) 5 13.4, 17.6, 48.16,

50.8, 56.0, 65.1, 68.8, 69.0, 71.2, 95.7, 170.1, 171.8.

152

 

 

6.5. References

1.

2.

10.

11.

12.

13.

14.

15.

16.

17.

18.

Rietschel, E. T.; Brade, H. Scientific American, 1992, Aug., 54-61.

Nikaido, H.; Vaare, M., Cellular and Molecular Biology, vol. 1, Neidhardt, F. C., Ed.,
American Society for Microbiology, Washington, DC, 1987, 7.

Morrison, D. C.; Ryan, J. L., Bacterial Endotoxic Lipopolysaccharides, vol. 1, CRC
Press, Inc., Boca Raton, 1992.

Nowotny, A., Beneficial Effects of Endotoxin, Plenum Press, New York, 1983.

Raetz, C. Annu. Rev. Biochem. 1990, 59, 129-170.

Homma, J.; Kanegasaki, S.; Luderitz, O.; Shiba, T.; Westphal, O. Bacterial
Endotoxin: Chemimcal, Biological and Clinical Aspects; Verlag Chemie, Basel,

1984.

Qureshi, N.; Honovich, J. P.; Hara, H.; Cotter, R. J.; Takayama, K. J. Biol. Chem.
1988, 263, 5502-5504.

Qureshi, N.; Takayama, K.; Kurtz, R. Infect. Immunol. 1991, 59, 441-444.

Wang, Y. ‘Analysis of the Molecular Basis for the Supramolecular Structure of
Lipopolysaccharides ’ PhD Dissertation, 1995, pp 86-138.

King, R. R. Can. J. Chem. 1975, 53, 1978.

Janson, J .; Lindberg, B. Methods in Carbohydrate Chemistry, Academic Press Inc.,
New York 1963, vol. 11, p.380.

Lemieux, R. U.; Huber, G. Canadian J. Chem. 1953, 31.
Hickenbottom, W. J. J. Chem. Soc. 1929, 1676.

Lemieux, R. U.; Howard, J. Methods in Carbohydrate Chemistry, Academic Press
Inc., New York 1963, vol. II, p. 400.

Saito, S.; Ichinose, K.; Sasaki, Y.; Sumita, S. Chem. Pharm. Bull. 1992, 40, 3261.
Lemieux, R. U. Justus Liebigs Ann. Chem. 1974, 1514.
Rashid, A.; Mackie, W. Can. J. Chem. 1990, 68, 1122-1127.

Lemieux, R. U.; Hendriks, K. 8.; Stick, R. V.; James, K. J. Am. Chem Soc. 1975, 97,
4056-4062.

153

 

l9. Aspinall, G. O.; Carpenter, R. C.; Khondo, L. Carbohydr. Res. 1987, 165, 281-298.
20. Wegmann, B.; Schmidt, R. R. J. Carbohydr. Chem. 1987, 6, 357.

21. de Nooy, A. E. J.; Besemer, A. C.; van Bekkum, H. Reel. Trav. Chim. Pay. B. 1994,
113, 165-166.

22. Glaudeman, P. J.; Fletcher, H. G. Methods in Carbohydrate Chemistry, Academic
Press, New York 1972, vol VI, p.374.

23. Wegmann, B.; Schmidt, R. R. J. Carbohyd. Chem. 1987, 6, 357-375.

154

IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

llllllllillllllllllllllllllll1111111111