Synthesis of a library of variously modified 4-methylumbelliferyl xylosides and a structure-activity study of human β4GalT7

Organic &

Biomolecular Chemistry

View Article Online

View Journal

PAPER

Synthesis of a library of variously modiﬁed

-methylumbelliferyl xylosides and a

4

Cite this: DOI: 10.1039/c7ob02530k

structure–activity study of human β4GalT7†

a

b

Samir Dahbi,

Jean-Claude Jacquinet,

Isabelle Bertin-Jung, Anne Robert,

b

Nick Ramalanjaona, Sandrine Gulberti, Sylvie Fournel-Gigleux* and

a

Chrystel Lopin-Bon

*

Proteoglycans (PGs) are complex macromolecules that are composed of glycosaminoglycan (GAG)

chains covalently attached to a core protein through a tetrasaccharide linker. The biosynthesis of PGs is

complex and involves a large number of glycosyltranferases. Here we present a structure–activity study of

human β4GalT7, which transfers the ﬁrst Gal residue onto a xyloside moiety of the linkage region. An

eﬃcient and regiocontrolled synthesis of a library of modiﬁed analogs of 4-methylumbelliferyl xyloside

(XylMU) is reported herein. Hydroxyl groups at the position C-2, C-3 or C-4 have been epimerized and/or

replaced by a hydrogen or a ﬂuorine, while the anomeric oxygen was replaced by either a sulfur or a

sulfone. The eﬀect of these compounds on human β4GalT7 activity in vitro and on GAG biosynthesis

in cellulo was then evaluated.

Received 12th October 2017,

Accepted 27th October 2017

DOI: 10.1039/c7ob02530k

rsc.li/obc

1

Introduction

Proteoglycans (PGs) consist of linear polysaccharide chains

called glycosaminoglycans (GAGs), covalently attached to

serine residues of a core protein through a tetrasaccharide

linker (Fig. 1a). PGs are mainly located in the extracellular

matrix and at the cell plasma membranes, and are also found

intracellularly. There are two major classes of GAG chains:

chondroitin sulfate (CS)/dermatan sulfate (DS) and heparin/

heparan sulfate (HS). Because of the structural diversity and

anionic characteristics of GAG chains, PGs interact with signal-

ing proteins, growth factors, immuno-modulators and a variety

of ligands, and thus play an important role in numerous physio-

logical processes, such as cell diﬀerentiation, proliferation and

1

migration, and tissue morphogenesis and architecture.

However, deregulation of PG metabolism is involved in degen-

2

erative pathologies such as cancer, osteoarticular and cardio-

Fig. 1 (a) Biosynthesis of the tetrasaccharide linker of HS and CS/DS

PGs. (b) Galactosylation of XylMU 1 to form Gal-XylMU.

a

Univ. Orléans et CNRS, ICOA, UMR 7311, F-45067 Orléans, France.

E-mail: chrystel.lopin-bon@univ-orleans.fr; http://www.univ-orleans.fr/icoa;

Fax: +33-2-3841-7281

b

3

UMR 7365 CNRS-Université de Lorraine, Biopôle-Faculté de Médecine, CS 50184,

vascular disorders, as well as genetic diseases, like the

4

5

4505 Vandoeuvre-lès-Nancy Cedex, France.

E-mail: sylvie.Fournel-gigleux@univ-lorraine.fr

Electronic supplementary information (ESI) available: Experimental procedures

Ehlers–Danlos syndromes.

PG biosynthesis is initiated by the formation of a tetra-

saccharide linker, GlcA(β1–3)Gal(β1–3)Gal(β1–4)Xylβ– on the

PG core protein, which is used as a primer for the elongation

of the CS/DS and HS chains. Among the enzymes involved,

†

1

19

13

and analytical data as well as copies of H, F and C NMR spectra of new key

intermediates and the final compounds and immunoblots from the in cellulo

screening. See DOI: 10.1039/c7ob02530k

This journal is © The Royal Society of Chemistry 2017

Org. Biomol. Chem.

View Article Online

Paper

Organic & Biomolecular Chemistry

β1,4-galactosyltransferase 7 (β4GalT7) catalyses the transfer of

the first Gal residue of the linkage region from the activated

5

sugar UDP-galactose (UDP-Gal) onto the xylose residue (Fig. 1a).

Because all GAGs share the same stem core tetrasaccharide,

β4GalT7 is a key enzyme and a prime target for the study of

GAG biosynthesis and the development of specific therapeutic

agents. Interestingly, GAG biosynthesis can be modulated by

β-D-xylosides carrying a hydrophobic aglycon, such as 4-methyl-

umbelliferyl β-D-xylopyranoside (XylMU, 1). Hydrophobic xylo-

side derivatives can act as acceptor substrates of β4GalT7 and

6

initiate GAG chain biosynthesis (Fig. 1b). Several chemical

syntheses of xyloside analogs as potential inhibitors of GAG

7a

synthesis have also been reported. Tsuzuki et al. and Garud

7b

et al. have described the synthesis of triazole analogs which

were able to inhibit GAG formation with various eﬃciencies

depending on the hydrophobicity of the aglycone group.

8

Siegbahn et al. reported a collection of xylosides linked to

diﬀerent aromatic aglycons with anomeric oxygen or sulfur. The

ability of modified xylosides to modulate β4GalT7 enzymatic

eﬃciency depends on the nature of the substitution and its

position on xylose, and on the anomeric center.

We recently reported a structure-guided approach for the Scheme 1 Reagents and conditions: (a) (i) Bu

2

SnO (1.5 eq.), iPrOH,

reﬂux, 3 h, (ii) ClCH C(O)Cl (1 eq.), CH Cl , 0 °C, 1 h; (b) BzCl (3 eq.),

9

design of β4GalT7 inhibitors. That study showed that the

2

CH

d) 2,2,3,3-tetramethoxybutane (2 eq.), (MeO)

MeOH, reﬂux, 12 h; (e) BzCl (1.1 eq.), Et N, DMAP, CH

2 h; (f) 95% TFA, CH Cl , rt, 1 h; (g) BzCl (1.5 eq.), Me SnCl

2

Cl

2

, pyridine, 0 °C → rt, 12 h; (c) thiourea (5 eq.), MeOH, reﬂux, 6 h;

CH (4 eq.), CSA (0.05 eq.),

Cl , 0 °C → rt,

(0.05 eq.),

aglycon 4-methylumbelliferyl (MU) produced better xylosides

in terms of in vitro enzymatic eﬃciency and in cellulo GAG

priming, in comparison with naphthyl xylosides for instance.

(

3

2

1

2

This observation could be explained, at least in part, by the DIPEA, THF, rt, 2 h.

structural organisation of the β4GalT7 active site comprising

1

95

His whose nitrogen backbone was predicted to interact with

the carbonyl group of the coumarinyl molecule. We have also

Protected xyloside 4, having the free C4-OH, was prepared

1

shown that 4-methylumbelliferyl 4-deoxy-4-fluoro-xyloside according to the literature

using selective dibutyltin-

(

4F-XylMU 18) inhibited β4GalT7 activity in vitro and GAG mediated chloroacetylation of 1 followed by subsequent ben-

priming in recombinant cells. Here, we present an eﬃcient zoylation. After dechloroacetylation with thiourea in methanol,

9

and regiocontrolled synthesis and the functional evaluation of 4 was isolated in 37% yield in four steps.

the eﬀect of a large library of variously modified analogs of

In order to isolate the C2-OH, we used butanediacetal (BDA)

XylMU 1 on the β4GalT7 activity, using a double screening as a protecting group, which has already been employed to

strategy: (i) we tested the compounds towards the in vitro protect xylosides, leading to mixtures of 2,3- and 3,4-protected

8

c,12

activity of the purified recombinant human β4GalT7, and (ii) sugars.

we determined their potency to regulate the glycanation of the refluxing a mixture of 1 in methanol for 12 h, in the presence of

core protein of decorin, used as a model PG in cellulo.

2,2,3,3-tetramethoxybutane (TMB), trimethylorthoformate and a

Introduction of the BDA group was achieved by

13

catalytic amount of camphor sulfonic acid (CSA). The two

separable BDA-protected xylosides 5a (3,4-protected) and 5b

(

2,3-protected) were obtained in a 3 : 2 ratio in 96% yield, and

2

Results

allowed us to isolate the C2- and C4-OH groups in one step.

Furthermore, we took advantage of the preparation of 3,4-

2

.1 Synthesis

Starting from XylMU 1, our goal was to isolate each OH group protected xyloside 5a to isolate the C3-OH. First, 5a was ben-

of the xylose moiety and then perform chemical modifications zoylated to give 6 in 85% yield, and then the BDA group was

(

fluorination, deoxygenation). Although 1 was commercially removed using 95% TFA to aﬀord 7 in 71% yield. Finally, selec-

1

0

available, it was prepared as reported in the literature.

tive benzoylation of the C4-OH with benzoyl chloride in the

SnCl led to derivative 8 with the free C3-OH

2

.1.1 Isolation of the OH groups at C-2, C-3 and C-4. presence of Me

2

1

4

Isolation of each one of the three hydroxyl groups of D-xylose in 95% yield.

can prove diﬃcult, as they are all secondary, equatorial and

Having isolated the OH groups at C-2, C-3 and C-4, we per-

therefore similar in reactivity. Hence, there is no general pro- formed either fluorination (two epimers) or deoxygenation of

tecting group strategy for all positions of D-xylose. We thus these positions.

used diﬀerent strategies to selectively isolate each hydroxyl

group of XylMU 1 (Scheme 1).

2.1.2 Chemical modifications at C-4. The fluoro- and

deoxy-analogs at C-4 were prepared from either 2,3-di-O-ben-

Org. Biomol. Chem.

This journal is © The Royal Society of Chemistry 2017

View Article Online

Organic & Biomolecular Chemistry

Paper

zoylated xyloside 4 or 2,3-BDA-protected xyloside 5b, both with in two steps. Finally, the OH group at C-3 was selectively ben-

the free C4-OH. zoylated at low temperature (−35 °C) to aﬀord 11 in 40% yield

Fluorination reactions were envisaged using diethyl- in four steps. This revealed to be a more eﬃcient and straight-

aminosulfur trifluoride (DAST). This reagent has been exten- forward route to obtain our desired precursor the free axial C4-

sively applied to the fluorination of carbohydrates, converting OH. In comparison, 11 was prepared from xyloside 4 in 24%

alcohols to fluorides with overall inversion of stereo- yield in six steps, while BDA-protected arabinoside 12 was

1

5

chemistry. Therefore, in order to introduce an equatorial obtained from xyloside 5b in 27% yield in three steps.

fluoride at C-4, the corresponding OH group on precursors 4

and 5b had to be first inverted.

Arabinosides 11 and 12, with the free C4-OH, were then

engaged in fluorination reactions to introduce an equatorial

Epimerization of C4-OH was achieved by using the Lattrell– fluorine atom (Scheme 3).

1

6

Dax reaction, i.e. displacement of a 4-O-triflyl group by tetra-

Reaction of 11 and 12 with DAST in dichloromethane, from

butylammonium nitrite (TBAN) (Scheme 2). By reaction with −40 °C to room temperature (method A), aﬀorded the desired

trifluoromethanesulfonic anhydride, triflates 9 and 10 were 4-deoxy-4-fluoro-xylosides 16 and 17 in 26% and 49% yields,

prepared from 4 and 5b, respectively. While triflate 9 had to be respectively. When the reaction was carried out in diglyme at

used rapidly after a quick work-up to avoid the possible dis- 100 °C (method B), as described in the literature for the syn-

1

8

placement of the neighboring benzoyl group at C-3, triflate 10 thesis of 2-deoxy-2-fluoro-D-glucose, xyloside 16 was obtained

could be isolated in 88% yield after flash chromatography. in a better yield of 48%, while xyloside 17 was obtained in a

Consecutive inversion was carried out with TBAN in toluene at lower 19% yield. 16 was then deprotected with sodium metho-

room temperature to aﬀord arabinosides 11 and 12, with the xide to aﬀord 4F-XylMU 18 in 95% yield, while the BDA group

free C4-OH, in 65% yield from 4 and 74% yield from 5b, on 17 was removed with 95% TFA, to give 18 in 53% yield.

respectively.

When method A or B was applied to fluorinate xyloside 4,

Since this inversion step corresponds to the conversion of a only traces of the desired 4-deoxy-4-fluoro-arabinoside 19 were

D-xyloside to an L-arabinoside, we also thought of working observed (Scheme 4). However, starting from BDA-protected

directly on L-arabinose (Scheme 2). Although it was commer- xyloside 5b, fluorination with DAST aﬀorded the targeted

cially available, 4-methylumbelliferyl α-L-arabinopyranoside 13 fluorinated arabinoside 20, along with the elimination product

was prepared in the laboratory by glycosylation of peracetylated 21. Upon deprotection of the BDA group, 4F-AraMU 22 was

α-L-arabinopyranosyl bromide with 4-methylumbelliferone obtained from 20 in 67% yield. As fluorination of xyloside 4

(66% yield), followed by transesterification under the Zemplén with DAST failed, we decided to use instead a triflate/fluoride

conditions (81% yield), using the same procedures as those tandem sequence, with tetrabutylammonium fluoride (TBAF)

1

0

described for the synthesis of XylMU 1. Arabinoside 13 was as fluorinating agent (Scheme 4). 2,3-Di-O-benzoylated triflate

then protected at C-3 and C-4 with an isopropylidene acetal, to 9 was successfully fluorinated by reaction with an excess of

give 14 in 76% yield. Next, benzoylation of the free C2-OH and TBAF in acetonitrile at 75 °C. To our surprise, it also resulted

1

7

removal of the isopropylidene aﬀorded diol 15 in 93% yield in the removal of the benzoyl groups, which directly aﬀorded

desired 4F-AraMU 22 in 66% yield from 4. Such lability of ester

1

9

groups in the presence of TBAF has already been reported.

We also tested this method for the fluorination of BDA-pro-

tected xyloside 5b, which led to arabinoside 20 in 36% yield

and the elimination product 21 in 15% yield, a result compar-

able to those obtained with the fluorination with DAST. The

BDA group was then removed to aﬀord 22 in 67% yield.

Scheme 2 Reagents and conditions: (a) Tf

°C, 10 min; (b) TBANO (5 eq.), toluene, rt, 20 min; (c) 2-methoxy-

propene (4 eq.), ACS, pyridine, rt, 2 h; (d) BzCl (1.1 eq.), Et N, DMAP,

CH Cl , rt, 3 h and then 70% AcOH, 70 °C, 1 h; (e) BzCl (1.1 eq.), pyridine,

35 °C, 4 h.

2 2 2

O (1.5 eq.), CH Cl , pyridine,

0

2

3

Scheme 3 Reagents and conditions: method A: DAST (4.5 eq.), CH

−40 °C → rt, 4 h; method B: DAST (3 eq.), diglyme, 100 °C, 7 min; (a)

MeONa, MeOH, rt, 30 min; (b) TFA/H O, rt, 2 h.

2 2

Cl ,

2

−

2

This journal is © The Royal Society of Chemistry 2017

Org. Biomol. Chem.

View Article Online

Paper

Organic & Biomolecular Chemistry

benzoylated xyloside 4 or BDA-protected xyloside 5b. Although

no thorough conformational study was performed, analysis of

the NMR coupling constants suggest that β-glycosides 18

(J1,2 = 6.9 Hz), 22 (J1,2 = 7.3 Hz) and 27 (J1,2 = 7.2 Hz) are in the

4

normal C chair conformation. Interestingly, the use of BDA

1

4

as protecting group locked the carbohydrate ring in this C

1

chair conformation during the syntheses starting from 5b,

while the use of benzoyl groups induced a deformation of the

chair, as most of the intermediates starting from 4 were not in

a normal chair conformation (J1,2 < 5 Hz) until the final

4

deprotection restored the C conformation.

1

2.1.3 Chemical modifications at C-2. The fluoro- and

deoxy-analogs at C-2 were prepared from BDA-protected xylo-

side 5a, with the free C2-OH.

Here again, the C2-OH on 5a had first to be inversed in

order to introduce an equatorial fluorine atom. However, when

we used the triflate/TBAN reaction sequence, lyxoside 30 was

obtained in low 36% yield from triflate 28 (Scheme 6). We thus

2 2

Scheme 4 Reagents and conditions: method A: DAST (4.5 eq.), CH Cl ,

−

40 °C → rt, 4 h; method B: DAST (3 eq.), diglyme, 100 °C, 7 min; (a) decided to investigate another method to epimerize the C2-

Tf O (1.5 eq.), CH Cl

, pyridine, 0 °C, 10 min; (b) TBAF (9.5 eq.), MeCN, OH. Manner et al. have recently reported a Swern oxidation–

5 °C, 1 h; (c) TFA/H O, rt, 2 h.

2

7

2

epimerization-reduction sequence to selectively epimerize any

position in isopropylidene-protected xylopyranosides, thus

forming arabinosides, ribosides or lyxosides. Following their

strategy, ketone 29 was prepared from xyloside 5a by Swern oxi-

dation, and then reduced in situ with sodium borohydride to

aﬀord lyxoside 30 in an overall 65% yield.

2

0

The deoxygenation of xylopyranosides by the reaction of

xanthates intermediates with Bu SnH and AIBN in a Barton

McCombie radical deoxygenation reaction has been reported.

3

8

c

A similar methodology was used to deoxygenate xylosides 4

and 5b (Scheme 5). First, thionocarbonates 23 and 24 were pre-

pared from 4 and 5b in 68% and 76% yield, respectively, by

reaction with O-phenyl chlorothionoformate. Then, Barton–

McCombie radical deoxygenation of 23 and 24 in presence of

We then carried out fluorination reactions on both BDA-

protected lyxoside 30 and xyloside 5a (Scheme 7). Fluorination

of 30 with DAST using method A aﬀorded xyloside 31 in 54%

yield, while method B gave only 21%. Subsequent removal of

the BDA group aﬀorded 2F-XylMU 33 in 77% yield.

Fluorination of 5a with DAST only led to recovery of the start-

ing material. However, fluorination of the corresponding tri-

flate 28 with an excess of TBAF aﬀorded desired lyxoside 32 in

Bu SnH and ACBN in toluene at reflux aﬀorded 4-deoxy-xylo-

3

sides 25 and 26 in 88% and 62% yield, respectively. After de-

protection, 4H-XylMU 27 was obtained in 87% yield from 25

and in 33% yield from 26.

39% yield, which was then deprotected to give 2F-LyxoMU 34

In conclusion, 4F-XylMU 18 and 4F-AraMU 22, as well as

in 88% yield.

4

H-XylMU 27, were successfully prepared from either 2,3-di-O-

Derivative 5a was also engaged in deoxygenation reaction

(

Scheme 8). Thionocarbonate 35, prepared from 5a in 59%

yield, was deoxygenated using Barton–McCombie reaction to

give 2-deoxy derivative 36 in 44% yield over two steps.

Scheme 6 Reagents and conditions: (a) Tf

0 °C, 10 min; (b) TBANO (5 eq.), toluene, rt, 20 min; (c) 2 M oxalyl chlor-

ide (3 eq.), DSMO (6 eq.), CH Cl , −78 °C, 2 h and then DIPEA (10 eq.),

−78 °C → rt, 30 min; (d) NaBH (2.7 eq.), MeOH, 0 °C, 4 h.

2 2 2

O (1.5 eq.), CH Cl , pyridine,

Scheme 5 Reagents and conditions: (a) PhO(CS)Cl (3.2 eq.), CH

2

Cl

2

,

2

pyridine, rt, 1.5 h; (b) Bu SnH (1.5 eq.), ACBN, benzene, 80 °C, 3 h; (c)

3

2

MeONa, MeOH, rt, 30 min; (d) TFA/H

2

O, rt, 2 h.

4

Org. Biomol. Chem.

This journal is © The Royal Society of Chemistry 2017

View Article Online

Organic & Biomolecular Chemistry

Paper

step route was longer than the direct deoxygenation of 38, the

overall yield was better (63%).

In conclusion, 2F-XylMU 33 and 2F-LyxoMU 34, as well as

2H-XylMU 37, were successfully prepared from xyloside 5a. In

particular, the synthesis of 37 highlighted the incompatibility

of acid-labile protecting groups with 2-deoxy derivatives.

Analysis of the NMR coupling constants indicates that

β-glycosides 33 (J₁_,₂= 7.2, J₁_,_F= 3.8 Hz), 34 (J₁_,₂= 1.9, J

=

1,F

4

1

10.9 Hz), and 37 (J1,2 = 7.3 and 2.8 Hz) are in a C confor-

mation, as it was observed on C4-modified analogs, with the

same conclusion regarding the influence of BDA and benzoyl

groups on the conformation of the synthetic intermediates.

2.1.4 Chemical modifications at C-3. The fluoro- and

deoxy-analogs at C-3 were prepared from xyloside 8, with the

free C3-OH.

In order to introduce an equatorial fluorine atom, the C3-

OH was first epimerized (Scheme 9). Triflate 41 was prepared

from 8 and then reacted with TBAN to give riboside 42 in 39%

yield over two steps. Next, fluorination of 42 with DAST using

method A led to xyloside 43 in 54% yield, while reaction using

method B only yielded traces of the desired product. Finally,

deprotection aﬀorded 3F-XylMU 44 in 68% yield.

2 2

Scheme 7 Reagents and conditions: method A: DAST (4.5 eq.), CH Cl ,

−

40 °C → rt, 4 h; method B: DAST (3 eq.), diglyme, 100 °C, 7 min. (a)

TBAF (9.5 eq.), MeCN, 75 °C, 1 h; (b) TFA/H O, rt, 2 h.

2

Direct fluorination of 8 with DAST only led to traces of

desired riboside 45 when method A was used. Using method

B, the product was obtained in 20% yield. On the other hand,

reaction of triflate 41 with an excess of TBAF directly yielded

desired 3F-RiboMU 48. However, the compound was isolated

with inseparable TBAF-derived salts that could not be removed

by purification. Since the use of an excess of TBAF (9.5 equiv.)

only led to impure deprotected product 48 (Table 1, entry 1),

we decided to test the reaction with 4.5 equiv. of TBAF. After

10 min at 75 °C, TLC showed complete conversion, and a

separable mixture of di- and mono-benzoylated 3-deoxy-3-

Scheme 8 Reagents and conditions: (a) PhO(CS)Cl (3.2 eq.), CH

pyridine, rt, 1.5 h; (b) Bu SnH (1.5 eq.), ACBN, benzene, 80 °C, 3 h; (c)

5% TFA, CH Cl , rt, 10 min; (d) BzCl (3 eq.), pyridine, rt, 12 h; (e)

2 2

Cl ,

3

9

2

MeONa, MeOH, rt, 30 min.

However, removal of the BDA group in acidic conditions (95%

TFA) led to the degradation of 36 by hydrolysis of the aglycon.

We thus revised our synthetic strategy and decided to remove

the BDA group before the deoxygenation. Deprotection of

thionocarbonate 35 aﬀorded diol 38 in 79% yield. However,

subsequent radical deoxygenation led to 2H-XylMU 37 in only

3

8% yield after isolation from a complex mixture. Therefore,

we decided to reprotect diol 38 with basic-labile benzoyl

groups to give 39 in 82% yield, which was deoxygenated to

aﬀord deoxy-derivative 40 in better 85% yield. Consecutive de-

protection of 40 using sodium methoxide aﬀorded 2H-XylMU

Scheme 9 Reagents and conditions: method A: DAST (4.5 eq.), CH

40 °C → rt, 4 h; method B: DAST (3 eq.), diglyme, 100 °C, 7 min. (a)

Tf O (1.5 eq.), CH Cl , pyridine, 0 °C, 10 min; (b) TBANO (5 eq.),

toluene, rt, 20 min; (c) MeONa, MeOH, rt, 30 min; (d) TBAF (1.1 eq.),

2 2

Cl ,

−

2

3

7 in 91% yield, without any degradation. While this three- MeCN, 75 °C, 10 min.

This journal is © The Royal Society of Chemistry 2017

Org. Biomol. Chem.

View Article Online

Paper

Organic & Biomolecular Chemistry

Table 1 Removal of Bz groups during ﬂuorination of 8 with TBAF

Our retrosynthetic strategy was to first deoxygenate the posi-

tion 2, and then either fluorinate or deoxygenate the position

4

(Scheme 11). Arabinoside 14 was selected as precursor

because it had the following characteristics: (a) the C2-OH was

already diﬀerentiated; (b) the equatorial C3-OH and the axial

C4-OH were diﬀerentiable; (c) the C4-OH was already in the

right conformation to introduce an equatorial fluorine atom.

As presented on Scheme 12, thionocarbonate 54 was first

prepared in 75% yield from 14. The isopropylidene group was

then removed to give diol 55 in 95% yield. Direct deoxygena-

Yield (%)

4

5

46

47

48

2

R = H

R = Bz

R = H

4

Entry

TBAF (eq.)

R = Bz

R = H

R = Bz

R = H

1

2

3

9.5

4.5

1.1

—

19

75

—

16

—

23

—

nd

—

fluoro-ribosides was isolated (Table 1, entry 2). Use of only 1.1

equiv. of TBAF led to desired 2,4-di-O-benzoylated riboside 45

in 75% yield as the major product, with only traces of the

mono-protected derivatives. 3F-RiboMU 48 was then obtained

in 78% yield after deprotection.

Finally, radical deoxygenation of thionocarbonate 49, pre-

pared from 8 in 40% yield, aﬀorded 3-deoxy-xyloside 50 in

8

3

5% yield, which was subsequently deprotected to give Scheme 11 Retrosynthesis of 2,4-modiﬁed xylosides 52 and 53.

H-XylMU 51 in 85% yield (Scheme 10).

In conclusion, 3F-XylMU 44 and 3F-RiboMU 48, as well as

3

H-XylMU 51, were successfully prepared from xyloside 8.

Regarding the conformation of these β-glycosides, NMR data

4

support that xyloside 44 (J1,2 = 7.3 Hz) is in a

mation, while riboside 48 (J₁_,₂, = 4.1 Hz) and xyloside 51 (J₁_,₂

Hz) are not. It should be noted that xyloside 8 (J1,2 4.4 Hz)

1

C confor-

<

1

was already in a twisted conformation, probably due to the

presence of the benzoyl groups at C-2 and C-4, and that was

the case for all the intermediates during the syntheses of the

three analogs. However, while the final deprotection restored

the normal chair conformation for xyloside 44, it surprisingly

did not for riboside 48 and xyloside 51.

2

.1.5 Synthesis of analogs modified at two positions.

While synthesizing analogs modified at C-2, C-3 or C-4, we

wanted to prepare an example of xyloside modified at two posi-

tions, to further study how it would aﬀect the enzymatic

aﬃnity. We thereby report the synthesis of 2,4-dideoxy-4-

fluoro-xyloside 52 and 2,4-dideoxy-xyloside 53.

Scheme 12 Reagents and conditions: (a) PhO(CS)Cl (3.2 eq.), CH

pyridine, rt, 1.5 h; (b) 95% TFA, CH Cl , rt, 10 min; (c) BzCl (3 eq.), pyri-

dine, rt, 12 h; (d) Bu SnH (1.5 eq.), ACBN, benzene, 80 °C, 3 h; (e)

2 2

Cl ,

Scheme 10 Reagents and conditions: (a) PhO(CS)Cl (3.2 eq.), CH

2

Cl

2

,

2

pyridine, rt, 1.5 h; (b) Bu SnH (1.5 eq.), ACBN, benzene, 80 °C, 3 h; (c)

3

MeONa, MeOH, rt, 30 min.

MeONa, MeOH, rt, 30 min; (f) DAST (3 eq.), diglyme, 100 °C, 7 min.

Org. Biomol. Chem.

This journal is © The Royal Society of Chemistry 2017

View Article Online

Organic & Biomolecular Chemistry

Paper

2

1

tion of 55 aﬀorded the 2-deoxy-arabinoside 58 in 44% yield to the corresponding O-xylosides. For example, Jacobsson

after isolation from a complex mixture, while benzoylation of et al. investigated a series of 1-thio-β-xylopyranosides that

2

5

5 followed by deoxygenation and deprotection aﬀorded 58 in showed strong priming of relatively short GAG chains.

better 78% yield over three steps.

Siegbahn et al. showed that sulfur analog 65, as well as

The synthesis of intermediate 59, precursor of our targeted sulfone analog 66, were galactosylated by β4GalT7 more

8

a

analogs, required the selective benzoylation of the C3-OH. eﬃciently than the corresponding O-xyloside 64 (Fig. 2).

According to the NMR data, diol 58 was in a normal chair con- Furthermore, 66 was able to eﬃciently prime GAG chains in

formation, which should allow to diﬀerentiate the equatorial cell studies. Therefore, in order to investigate the eﬀect of

C3-OH over the axial C4-OH. Various methods were thus replacing the anomeric oxygen with sulfur, we decided to

tested, as shown in Table 2. While benzoylation of 58 in pyri- prepare analogs 67–74 (Fig. 2). As the first results of the bio-

dine at −35 °C did not work (entry 1), reaction in presence of logical tests on C-2, C-3 and C-4 modified O-xylosides were

Me SnCl led to a mixture of mono-protected products in 3 : 1 suggesting that the best inhibition was obtained when repla-

2 2

ratio in favour of 59, but with a low conversion (entry 2). A cing the C4-OH by a fluorine atom, we decided to synthesize

better conversion was obtained using dibutyltin-mediated ben- sulfur analogs 69 and 70, as well as sulfone analogs 73 and 74.

zoylation (entry 4), although the selectivity was only 2 : 1. Upon

purification, 59 could be isolated in 51% yield. Having isolated of the corresponding peracetylated bromides 75 and 76 with

the C4-OH, 59 was either fluorinated or deoxygenated. 7-mercapto-4-methylcoumarin, followed by transesterification

XylSMU 67 and AraSMU 68 were prepared by glycosylation

Fluorination with DAST in diglyme aﬀorded 2,4-dideoxy-4- in the Zemplén conditions (Scheme 13). Furthermore, oxi-

fluoro-xyloside 61 in 51% yield. However, final deprotection dation of the sulfur prior to the transesterification aﬀorded

led to 2H,4F-XylMU 52 in only 30% yield, due to the sensitivity sulfone analogs 79 and 80 which gave, after deprotection

of the compound upon neutralization of the sodium metho- XylSO

xide with acidic resin. 4F-XylSMU 69 and 4F-XylSO

2

MU 71 and AraSO

2

MU 72.

MU 73 were synthesized from

2

Deoxygenation of thionocarbonate 62, prepared from 59 in AraSMU 68 using the same strategy as described for the corres-

8% yield, aﬀorded 2,4-dideoxy-derivative 63 in 42% yield. ponding O-xylosides (Scheme 14). The C4-OH was first isolated

5

Final deprotection led to 2,4H-XylMU 53 in only 24% yield, the in four steps to give arabinoside 84 in 44%, which was then

compound being even more sensitive than 52 to traces of acid. fluorinated with DAST using the best conditions identified in

In conclusion, starting with arabinoside 14 as a common the O-series to give xyloside 85 in 51% yield. 85 was then either

precursor, we were able to prepare two analogs modified at C-2 directly deprotected to aﬀord 4F-XylSMU 69 in 63% yield, or

and C-4. 2H,4F-XylMU 52 was obtained in 4% over nine steps, oxidized and then deprotected to give 4F-XylSO MU 73.

2

while 2,4H-XylMU 53 was obtained in 2% yield over ten steps. However, while oxidation of 85 successfully aﬀorded sulfone

Moreover, it should be noted that, according to the NMR, analog 86 in 87% yield, subsequent treatment with sodium

β-glycosides 52 (J₁_,₂= 5.0 and 3.7 Hz) and 53 (J₁_,₂= 3.6 and <1 methoxide led to the formation of a significant side product

Hz) seem to be in a twisted chair conformation. In addition to identified as glycal 87, which competed with the deprotection

the chemical modifications, this might have a critical impact of 86, lowering the yield of 73 to 36% yield. This could be

on the recognition of these molecules by β4GalT7, and there- explained by the high acidity of the anomeric proton neigh-

fore on the enzymatic activity.

bouring the sulfone group.

2

.1.6 Synthesis of analogs modified at the anomeric

center. It has already been reported that xylosides with

S-glycosidic bonds generally initiate GAG priming comparable

Table 2 Selective benzoylation of position 3 of 2-deoxy derivative 58

a

Yield (%)

Entry

1

Conditions

BzCl, pyr

59

60

—

−

35 °C, 4 h

BzCl, Me SnCl

(1) Bu SnO, CHCl

0 °C, 4 h

2) BzCl, CH

2

3

2

DIPEA, THF, rt, 1 h

/MeOH 1 : 1

33

51

10

25

2

3

6

(

Fig. 2 Modiﬁcations at the anomeric center: reported analogs 64–66,

XylMU 1, synthesized analogs 13–22 and our targeted sulfur and sulfone

analogs 67–74.

2 2

Cl

a

Determined by NMR.

This journal is © The Royal Society of Chemistry 2017

Org. Biomol. Chem.

View Article Online

Paper

Organic & Biomolecular Chemistry

Scheme 15 Reagents and conditions: (a) (i) Bu

reﬂux, 3 h, (ii) ClCH C(O)Cl (1 eq.), CH Cl , 0 °C, 1 h; (b) BzCl (3 eq.),

CH Cl , pyridine, 0 °C → rt, 12 h; (c) thiourea (5 eq.), MeOH, reﬂux, 6 h;

d) (i) Tf O (1.5 eq.), CH Cl , pyridine, 0 °C, 10 min, (ii) TBAF (1.1 eq.),

MeCN, 75 °C, 20 min; (e) MeONa, MeOH, rt, 30 min; (f) mCPBA (4 eq.),

CH Cl , 0 °C, 2 h.

2

SnO (1.5 eq.), iPrOH,

Scheme 13 Reagents and conditions: (a) 7-mercapto-4-methyl-

coumarin (1.5 eq.), K

2 3 2 2

CO , acetone, rt, 12 h; (b) mCPBA (4 eq.), CH Cl ,

2

0

°C, 2 h; (c) MeONa, MeOH, rt, 30 min.

2

(

2

92 in 46% yield (two steps). 92 was then either directly depro-

tected to aﬀord 70 in 95% yield, or oxidized and then depro-

tected to give 74. Oxidation of 92 gave sulfone analog 93 in

89% yield, which was then treated with sodium methoxide.

Again, we observed the competitive formation of glycal 94, and

74 was obtained in only 49% yield.

In conclusion, sulfur analogs 67–70 and sulfone analogs

1–74 were synthesized in good yields. On a conformational

7

aspect, all these molecules are in normal chair conformation.

In summary, we prepared a library of 19 analogs of XylMU 1

(

Fig. 3), which were tested as substrates and/or inhibitors for

β4GalT7.

2.2 Biochemical and biological screening of modified

xylosides

In order to identify lead compounds with biological eﬀects

from the series of analogs obtained by chemical synthesis, a

two-step screening process was set up. This procedure

Scheme 14 Reagents and conditions: (a) 2-methoxypropene (4 eq.),

ACS, pyridine, rt, 2 h; (b) BzCl (3 eq.), pyridine, CH Cl , rt, 12 h; (c) 95%

TFA, rt, 30 min; (d) BzCl (1.1 eq.), CH Cl , pyridine, −15 °C, 2 h; (e) DAST

2

(

3 eq.), diglyme, 100 °C, 7 min; (f) MeONa, MeOH, rt, 30 min; (g) mCPBA included: (1) an in vitro assay to assess the potency of the com-

(

4 eq.), CH

2

Cl

2

, 0 °C, 2 h.

pounds to act as substrates, inhibitors or activators of the puri-

fied recombinant human β4GalT7 using XylMU 1 as a refer-

ence acceptor substrate, and (2) an in cellulo screening test to

4

F-AraSMU 70 and 4F-AraSO

2

MU 74 were synthesized from identify compounds capable of modulating (initiating or inhi-

XylSMU 67 using the same strategy as described in the O-series biting) galactosyltransferase activity in cultured HeLa cells,

(Scheme 15). The C4-OH was first isolated using a chloro- taken as model cells expressing β4GalT7 endogenously. In the

acetate to give 90 in 32% yield over four steps. Synthesis of the second step, the inhibitory potency of the eﬀector molecule

triflate followed by fluorination with TBAF aﬀorded arabinoside was monitored through its capacity to reduce the formation of

Org. Biomol. Chem.

This journal is © The Royal Society of Chemistry 2017

View Article Online

Organic & Biomolecular Chemistry

Paper

Fig. 4 (a) Eﬀects of thio- and sulfonyl xylosides 67 and 71 on β4GalT7

activity. Assays using puriﬁed recombinant 6His-GST-hβ4GalT7ΔNt60

were performed as described in the experimental data section. Kinetic

curves of puriﬁed recombinant enzyme towards XylSMU 67 and

2

XylSO MU 71 were determined using nonlinear least-squares regression

analysis of the data ﬁtted to substrate inhibition rate equation. The

results are the mean values ± S.E. of three independent measurements

on assays performed in duplicate. (b) Eﬀects of XylSMU 67 and

2

XylSO MU 71 on decorin core protein glycosylation in HeLa cells. HeLa

cells are stably expressing the human recombinant decorin core protein.

The decorin glycosylation level was monitored by western blotting, as

described in the experimental data section, and then quantiﬁed using

ImageJ software. Immunoblot analyses of the decorin core protein

Fig. 3 Summary of the synthesized analogs of XylMU 1.

GAG chains onto the core protein of decorin, chosen as a glycanation level are shown as inserts. The results are the mean values ±

model PG.

The relative rates of enzymatic activation/inhibition and

β4GalT7 kinetic parameters were determined in vitro as

described in the experimental data section and are all reported

in Table 3 and Fig. 4. Immunoblots from the in cellulo screen-

ing step are shown in Fig. 4 and in ESI.†

S.E. of three independent measurements. Statistical analysis was carried

out by a Student’s t-test with **p < 0.01 and ***p < 0.001 versus decorin

glycosylation in the presence of DMSO.

2

3

pharmacological properties,

developing such derivatives

2

.2.1 Modifications at the anomeric position. Since modifi- could be an attractive option to treat pathological situations

cations at the anomeric center have been shown to signifi- involving a GAG defect.

cantly influence the capacity of xylosides to act as primers of

Thio and sulfonyl derivatives of 1 acted as better substrates

8

a,21d

GAG synthesis,

thio- and sulfonyl derivatives of XylMU 1 than the O-xyloside acceptor. Indeed, XylSMU 67 and

were tested as substrates for the recombinant β4GalT7. Since it XylSO MU 71 respectively displayed enzymatic eﬃciencies

2

was previously shown that thioxylosides exhibit interesting about 2.5 and 2.2 times greater than that obtained with 1,

mainly due to apparent lower K

m

values (3 and 2 fold, respect-

ively) (Table 3). These results are in agreement with those

8

a

Table 3 Kinetic parameters of β4GalT7 towards C2-modiﬁed analogs. obtained with their corresponding naphthyl counterparts.

Kinetic parameters of puriﬁed recombinant 6His-GST-hβ4GalT7ΔNt60 Interestingly, detailed kinetic analyses showed that these com-

towards modiﬁed analogs were determined as described in the experi-

mental data section. The results are the mean values of three indepen-

inhibition kinetics (Fig. 4a), which could be attributed to an

dent measurements ± S.D. on assays performed in duplicate. ND indi-

pounds did not display Michaelis–Menten profile but substrate

allosteric control mechanism. Such regulation mechanism has

cates that no kinetic constant could be determined using an excess of

glycoside

been previously observed for UDP-glucuronosyltransferases

2

4

involved in drug metabolism, but to our knowledge, has not

been described for glycosyltransferases responsible for GAG

synthesis. Importantly, in cellulo tests indicated that both 67

and 71 eﬃciently competed with the GAG-substitution of the

decorin core protein, inducing a reduction in band intensity

and a shift towards lower molecular weights as their concen-

tration increased (Fig. 3b). In this respect, 67 was particularly

cat (min⁻

1

)

Eﬃciency (min⁻¹mM⁻¹

)

Analogs

K

m

(mM)

k

XylMU 1

0.34 ± 0.04

>10

195.4 ± 14.9 570.0

2

F-XylMU 33

F-LyxoMU 34 >10

H-XylMU 37

ND

3.06 ± 0.23

133.2 ± 20.9 44.3

0.11 ± 0.007 155.6 ± 17.7 1426.8

0.15 ± 0.041 194.0 ± 41.1 1253.3

XylSMU 67

XylSO MU 71

2

This journal is © The Royal Society of Chemistry 2017

Org. Biomol. Chem.

View Article Online

Paper

Organic & Biomolecular Chemistry

eﬃcient since a concentration of 10 µM was suﬃcient to 67 and XylSO

2

MU 71 act as “good” substrates and competitors

produce a band shift comparable to that produced by 1 at of β4GalT7, both 4F-XylSMU 69 and 4F-XylSO MU 73 deriva-

2

1

50 µM (Fig. 3b).

tives were only moderate eﬀectors of the enzyme.

In conclusion, substituting the O-linker of 1 by either a

It has to be noted that the deoxy analog 4H-XylMU 27 is

sulfur atom or a sulfone group provides a new interesting class also a weak eﬀector of the enzyme (Fig. 5b).

of xylosidic substrates that eﬃciently compete with the GAG-

2.2.3 Arabinosides. In order to assess whether

substitution of decorin in cellulo. The biochemical/biological L-arabinosides are also able to modulate the β4GalT7 activity,

properties of these molecules deserve to be further investi- AraMU 13, AraSMU 68 and AraSO MU 72 were tested as poten-

2

gated in a therapeutic context.

.2.2 Modifications at the C-4 position. Since it has been

tial acceptor substrates or inhibitors for β4GalT7.

8

b

2

In contrast to its corresponding naphthyl-derivative,

reported by us and others that 4F-derivatives display interest- AraMU 13 moderately enhanced the β4GalT7 activity, with an

7

a,9,25

ing inhibitory properties,

we decided to test the capacity increase of up to 20% of the galactosylation rate of XylMU 1 in

of the C4-fluorinated analogs of the thio- and sulfonyl-xylo- the presence of 5 mM of 13 (Fig. 5c). AraSMU 68 led to a

sides to modulate β4GalT7 activity. 4F-XylSMU 69, and to a modest increase (about 13%) of the relative enzymatic β4GalT7

lesser extent 4F-XylSO

XylMU 1 catalyzed by the purified β4GalT7 by about 20% compound precipitated at concentrations higher than 2.5 mM

Fig. 5d) and was thus a weaker inhibitor compared to and was thus not suitable for further in vitro assays. AraSO MU

2

MU 73, reduced the galactosylation of activity when used at 2.5 mM (Fig. 5c). However, the latter

(

2

9

4

F-XylMU 18. Comparable results were observed in cellulo, as 72 operated as a versatile but weak eﬀector depending on its

no significant change of the glycosylation level of the decorin concentration: 2.5 mM of the molecule slightly enhanced the

core protein occurred in the presence of 69 (see the ESI†). β4GalT7 activity (about 7%), while 5 mM resulted in a decrease

Altogether, these results show that, under our experimental of about 18% of the galactosyltransferase activity (Fig. 5c).

conditions, although their unsubstituted counterparts XylSMU However, the molecular mechanism by which these arabino-

sides, and especially 72, modulate the β4GalT7 activity has not

yet been characterized. It can be suggested that replacing the

naphthyl group by 4-methylumbelliferyl as an aglycon pro-

duces subtle conformational changes in the structure of the

overall conjugate, thus aﬀecting its capacity to interact with

the enzyme. This conjecture should be ascertained through

advanced enzymatic assays coupled with structural analyses of

the compounds in complex with the protein. Furthermore, it

should be noted that none of these arabinosides were able to

compete with the glycosylation process of the decorin core

protein in cells (see the ESI†).

The ability of C4-fluorinated L-arabinosides to activate or

inhibit the galactosylation of XylMU 1 by β4GalT7 was also

investigated. 4F-AraMU 22 displayed significant inhibitory pro-

perties. About 18% relative inhibition was reached at 2.5 and

5

mM of 22 (Fig. 5b). Unlike 22, 4F-AraSO MU 74 displayed

2

activator properties in vitro. About 13% and 22% relative

increases of enzymatic activity were observed for the latter

compound at 2.5 and 5 mM, respectively (Fig. 4d). It is reason-

able to assume that activation from exogenous eﬀectors arises

as a consequence of a positive allosteric control mechanism,

which has not yet been described for human glycosyltrans-

ferases involved in GAG synthesis. Such a hypothesis is corro-

borated by the substrate inhibition kinetics described above

for the thio- and sulfonyl derivatives of 1. Structural investi-

gations are currently in progress to discover the β4GalT7 allo-

2

steric binding site(s). However, 4F-AraSO MU 74 presented no

activating eﬀect on the glycosylation of the decorin core

Fig. 5 Eﬀects of modiﬁed xylosides on β4GalT7 activity. Assays using

puriﬁed recombinant 6His-GST-hβ4GalT7ΔNt60 were performed in the protein in cellulo, possibly because this compound is not lipo-

presence of ﬁxed 1 (0.3 mM) and UDP-Gal (1 mM) and two concen-

trations of modiﬁed xylosides (2.5 mM and 5 mM). Activities are pre-

philic enough to modulate the β4GalT7 activity in live cells

(

see the ESI†). 4F-AraSMU 70 showed no eﬀect on β4GalT7

sented compared to activity of XylMU 1. The results are the mean values

S.E. of three independent measurements on assays performed in

activity (Fig. 5d), indicating that the C4-fluorinated thio- and

sulfonyl-L-arabinosides did not show better inhibitory potency

than their O-counterpart. Further advanced enzymatic studies

±

duplicate. ND indicates that no eﬀect could be detected using an excess

of modiﬁed xylosides.

Org. Biomol. Chem.

This journal is © The Royal Society of Chemistry 2017

View Article Online

Organic & Biomolecular Chemistry

Paper

would be useful to determine whether the inhibitions 48 aﬀected the β4GalT7 activity in vitro (Fig. 5a) or were able to

observed for the C4-fluorinated L-arabinosides 22 and 74 compete with the decorin core protein glycosylation in cellulo

follow a competitive inhibition model, taking into account the (see the ESI†). These results somewhat diﬀer from those

2

6

high flexibility of the β4GalT7 structure. Furthermore, none reported for 3-deoxy-3-fluoro-naphthyl xyloside, which was

8

b

of these C4-fluorinated derivatives were able to aﬀect the glyco- shown to inhibit β4GalT7, but may result from the use of

sylation process of the decorin core protein in cells (see the 4-methylumbelliferone as an aglycon. Advanced structural

ESI†). Altogether, these results indicate that epimerisation of investigations are in progress to decipher the precise role of

the hydroxyl group at the C4 position and replacement by a the aglycon in the potency of xylosides to aﬀect in vitro or

fluorine mostly prevent the xyloside analog from binding to in cellulo β4GalT7 activity, which has been highlighted by many

8a,21d,28

the active site and acting as substrates or inhibitors of previous studies.

β4GalT7. eﬀector. Regarding the in vitro enzyme activity (Fig. 5a), 51 dis-

.2.4 Modifications at the C-2 position. The role and played around 5% relative activation rate for 2.5 mM and less

3H-XylMU 51 acted as a weak β4GalT7

2

eﬀects of xylose 2-O-modifications represent a major biological than 5% relative inhibition rate for 5 mM, whereas no eﬀect of

issue in the GAG biology field since phosphorylation at this this compound could be observed in cellulo (see the ESI†).

position is a key regulatory mechanism of linker formation

Altogether, these results clearly show that C3-modified

2

7

and GAG chain synthesis initiation. We have previously analogs of XylMU 1 should not be considered as precursors for

reported that C2-phosphorylation of xylose critically aﬀects the β4GalT7 eﬀector candidates, as substituting the equatorial C3

capacity of a xyloside analog to be a substrate of the recombi- OH-group mainly generated compounds that did not aﬀect the

2

7b

nant β4GalT7.

Thus, in this study, we have examined the in vitro galactosyltransferase activity or the in cellulo decorin

glycosylation. These results are in good agreement with the

eﬀects of various 2-O-modifications on the β4GalT7 activity.

It was clearly seen that 2-fluoro compounds were not suit- impact of the C3-substituent on the conformational equili-

able substrates for the enzyme, as both 2F-XylMU 33 and brium of the xylopyranoside moiety and therefore on the

2

F-LyxoMU 34 showed very low in vitro galactosyltransferase potency of the conjugates to bind and aﬀect the enzyme

2

9

activity, with a K

m

value higher than 10 mM (Table 3). In activity.

addition, 2H-XylMU 37 was a weak substrate, with an enzy-

matic eﬃciency about 13 times lower than that obtained for

XylMU 1 (Table 3). These results were confirmed in cellulo, as

none of these C2-modified glycosides eﬃciently competed

3

Conclusions

with the glycosylation process of the decorin core-protein (see Based on our previous work that demonstrated that 4-methyl-

the ESI†). These results corroborate our previous findings, umbelliferone (MU) as an aglycon fits well into the β4GalT7

9

indicating that the C2-phosphorylated analog of 7-methoxy-2- active site, we designed a rational approach to produce

naphthyl xyloside was not a substrate for the recombinant analogs of XylMU 1 targeting this enzyme to be used as enzy-

2

7b

β4GalT7.

They also emphasize that the integrity of the C2 matic modulators and/or cellular eﬀectors for GAG initiation.

position is essential for a xyloside to bind the active site and to To this end, we prepared a large chemolibrary of glycosides

act as a substrate of β4GalT7.

modified at C-2, C-3 and C-4 by epimerisation, deoxygenation

Finally, it was shown that, similarly to 4F-XylSO

2

MU 74, and fluorination. We also designed an original two-step

2

H,4F-XylMU 52 exhibited activator properties at 2.5 mM con- approach to test these compounds: (i) in vitro as substrates

centration, with an increase of about 17% with respect to the and/or inhibitors of the purified recombinant human β4GalT7

control activity (Fig. 5b). Unfortunately, 52 precipitated at con- and (ii) in cellulo as modulators of GAG synthesis.

centrations higher than 2.5 mM under aqueous conditions,

Several key points have emerged from this study.

rendering it unsuitable for further in vitro assays. Nonetheless, Importantly, by combining a sulfur atom or a sulfone group at

because of its small molecular mass and its hydrophobic pro- the anomeric center and MU as an aglycon, we designed an

perties, it was worth checking its activating properties interesting new class of glycosides that acted as substrates in

in cellulo. Indeed, 52 enhanced the glycosylation level of the vitro and as eﬃcient competitors of GAG synthesis in cellulo.

decorin core protein, since a marked band shift towards These β4GalT7 modulators open up avenues towards the devel-

higher molecular mass could be observed as the concentration opment of GAG-based drugs for pathological situations invol-

of 52 was increased (see the ESI†). Thus, unlike either ving GAG disruption of this metabolic pathway.

4

F-XylSMU 69 or 4F-AraSO

2

MU 74, 2H,4F-XylMU 52 may

We have expounded, for the first time described here for a

deserve to be further considered for its biological activity, GAG synthesizing glycosyltransferase, that human β4GalT7

being one of the few β4GalT7 activators reported to date with responds to allosteric control mechanisms. Indeed,

potential eﬃcacy in cells.

2H,4F-XylMU 52 moderately, but significantly, activated

2

.2.5 Modifications at the C-3 position. We finally deter- β4GalT7 in vitro and in cellulo. It is, to our knowledge, the first

mined whether substitutions onto the C3 position of XylMU 1 described molecule targeting β4GalT7 that is capable of poten-

would give rise to novel eﬀectors of β4GalT7. To this end, tiating GAG initiation synthesis in cellulo. On the other hand,

3

F-XylMU 44, 3F-RiboMU 48 and 3H-XylMU 51 were syn- we show that the thio and sulfonyl analogs of XylMU 1 exhibi-

thesized. However, none of the fluorinated derivatives 44 and ted typical substrate inhibition kinetics. These results point to

This journal is © The Royal Society of Chemistry 2017

Org. Biomol. Chem.

View Article Online

Paper

Organic & Biomolecular Chemistry

an allosteric regulation of human β4GalT7, either positive or in cellulo, and new allosteric control mechanisms in the regu-

negative depending on the structure of the xyloside. Structural lation of β4GalT7. These findings open up new avenues

investigations of human β4GalT7 in complex with these xylo- towards chemical biology tools to manipulate GAG synthesis

side analogs are currently in progress to identify such possible and to intervene in pathological situations involving GAG

allosteric binding site(s).

alterations.

A precise picture of the consequences of modifying the

hydroxyl group at diﬀerent positions of the xylose moiety was

also drawn. The 2-O-modified glycosides are very weak accep-

tor substrates, mainly due to a reduced apparent aﬃnity for

4

Experimental data

β4GalT7. This observation is likely to be attributed to the loss 4.1 Synthesis

of hydrogen bonding between the hydroxyl group at this posi-

tion and the neighboring carboxylic acid Asp228, in agreement

Solvents were dried by standard methods, and molecular

sieves were activated prior to use by heating for 4 h at 500 °C.

Melting points were determined in capillary tubes with a

Büchi apparatus and are uncorrected. Optical rotations were

2

6b

with the structure of human β4GalT7

and our model of the

9

XylMU 1 binding site. This is a major result from a biological

point of view. It confirms the assumption that the phosphory-

lated form of xylose, known to occur in vivo, is unlikely to be a 19

1

measured at 20–25 °C with a PerkinElmer 341 polarimeter. H,

1

13

F ( H-decoupled) and C NMR spectra were recorded at

substrate of human β4GalT7, in striking contrast to the follow-

2

5 °C using a Bruker NanoBay 400 instrument, with Me Si as

4

ing glycosyltransferase in the pathway β3GalT6, which displays

an internal standard, unless otherwise stated. High-resolution

mass spectra (ESI HRMS) were recorded on a Bruker maXis

mass spectrometer by the “Fédération de Recherche ICOA/

CBM” (FR 2708) platform. Flash-silica chromatography was

performed on Silica gel 60 (0.040–0.063 mm, Merck,

Darmstadt). The reactions were monitored by TLC on coated

aluminium sheets (Silica gel 60 GF254, Merck), and spots were

detected under UV light and by charring with a 95 : 5 mixture

of ethanol and sulfuric acid.

2

7a

a strong preference towards C2 phosphorylated xylose.

This

result reinforces the crucial role of xylose C2 substitution as a

key regulation mechanism of β4GalT7 activity and GAG syn-

thesis initiation.

The 3-O-modified glycosides cannot be further considered

as modulators of β4GalT7 as very weak activatory or inhibitory

eﬀects have been shown under our experimental conditions.

The importance of this position can be attributed either to the

loss of a hydrogen bond with Asp228 or to the presence of

alternatives, such as screw conformers. Indeed, the impor-

tance of conformational eﬀects in modified analogs, in par-

ticular at the C3 position, has been emphasized in previous

For the NMR data, 4-methylumbelliferyl aglycon group

atoms are numbered as indicated below:

8

c

studies.

The C4 position emerges as the most suitable position to

target in order to modulate (activate or inhibit) β4GalT7

activity and GAG priming in cells, as the highest percentage of

4.1.1 Typical procedure for the preparation of triflates.

activation or inhibition was shown with 2H,4F-XylMU 52 and Alcohol (1 mmol) was dissolved in freshly distilled CH Cl

2

4

F-AraSO

2

MU 74 as positive eﬀectors of β4GalT7, and (2 mL) and pyridine (0.48 mL, 6 mmol), and then cooled to

4

F-XylSMU 69 and 4F-AraMU 22 as β4GalT7 inhibitors. These 0 °C while stirring under argon. Trifluoromethylsulfonic an-

9

7a,b,8a,b

results corroborate previous studies by us and others

indicating that replacing the C4-OH group by a fluorine is a was continued for 10 minutes. The reaction mixture was then

suitable strategy for the design of inhibitors of β4GalT7. diluted with CH Cl , washed with cold 1 N HCl and sat.

However, it is important to highlight that the activatory and NaHCO , dried over MgSO , filtered and concentrated. The tri-

hydride (0.13 mL, 1.5 mmol) was added dropwise, and stirring

2

3

4

inhibitory potencies of these C4-modified analogs remain rela- flate was either purified by flash chromatography (BDA-pro-

tively modest. This supports the idea of a narrow active site tected products) or directly engaged in subsequent reactions

comprising a network of hydrogen bonds between the hydroxyl (Bz-protected products).

groups of the xylose moiety and carboxylic acid amino acid

residues that does not tolerate well changes at any position. TBANO

4.1.2 Typical procedure for the epimerization with

. Triflate (1 mmol) was dissolved in dry toluene (2 mL)

2

We thus also contributed to resolving the question whether and solid tetrabutylammonium nitrite (1.44 g, 5 mmol) was

GAG priming is possible from analogs with xylose modifi- added. After stirring for 20 min at rt, the reaction mixture was

2 2

cations, as reported in earlier investigations on modified evaporated. The residue was dissolved in CH Cl , washed with

3

0

benzyl-D-xylopyranosides, or whether xylose modifications brine, dried over MgSO₄, filtered and concentrated.

alter the capacity of the analogs to bind and aﬀect the Epimerized alcohol was obtained after flash chromatography.

8

c

β4GalT7, as more recently postulated. Our results support

4.1.3 Typical procedure for the fluorination with DAST.

the second hypothesis.

Method A: A stirred solution of alcohol (1 mmol) in freshly dis-

2 2

Altogether, we have designed and exhibited good substrates tilled CH Cl (2 mL) was cooled to −40 °C under Ar. DAST

and competitors of GAG synthesis in cellulo, activator xyloside (0.6 mL, 4.5 mmol) was then added, and the solution was left

analogs of β4GalT7 capable of stimulating GAG synthesis to stir for 4 h, warming to rt. The reaction was then quenched

Org. Biomol. Chem.

This journal is © The Royal Society of Chemistry 2017

View Article Online

Organic & Biomolecular Chemistry

Paper

with MeOH and diluted with CH

2

Cl

2

. The crude mixture was (100 MHz, CDCl

3

) 161.1, 159.6, 154.8, 152.3 (4 C, C-1 MU, C-3

washed with sat. NaHCO , dried over MgSO , filtered and con- MU, C-6 MU, C-8 MU), 125.5 (1 C, C-5 MU), 115.1 (1 C, C-0

3

4

centrated. The desired fluoro product was obtained after flash MU), 113.7 (1 C, C-4 MU), 112.8 (1 C, C-7 MU), 104.5 (1 C, C-2

chromatography. Method B: To a solution of alcohol (1 mmol) MU), 99.8, 99.6 (2 C, Cquat BDA), 98.6 (1 C, C-1), 72.7, 68.7 (2 C,

in diglyme (2.5 mL) was added DAST (0.36 mL, 3 mmol). The C-2, C-3), 67.3 (1 C, C-4), 66.5 (1 C, C-5), 48.1, 48.0 (2 C,

+

mixture was then stirred at 100 °C for 7 min, cooled to 0 °C, OMe BDA), 18.7 (1 C, Me MU), 17.6 (2 C, Me BDA); ESI HRMS

+

and quenched with MeOH. The crude mixture was diluted [M + H] m/z 423.1649 calcd for C21

H

27

O

9

, found: 423.1647.

2,3-O-(2′,3′-dimethoxybutane-

) was co-evaporated with water (2×), 2′,3′-diyl)-α-L-arabinopyranoside (12). Alcohol 5b (0.15 g,

Cl . The residue was then purified by 0.35 mmol) was subjected to the general procedures for the

flash chromatography to aﬀord the desired fluoro product. preparation of triflate. Flash chromatography (CH Cl /acetone

.1.4 Typical procedure for the fluorination with TBAF. 98 : 2 + 0.1% Et

with EtOAc and washed with sat. NaHCO and water (5×). The

4.2.2 4-Methylumbelliferyl

3

dried solution (MgSO

toluene (2×) and CH

4

2

4

3

N) aﬀorded triflate 10 (171 mg, 88%) as a

−145.5 (c 0.96 in CHCl ); δ (400 MHz,

Triflate (1 mmol) was dissolved in freshly distilled acetonitrile white foam. [α]

D

3

H

(

9 mL), and to this tetrabutylammonium fluoride (1 M solu- CDCl ) 7.54 (1 H, d, J 8.4, H-5 MU), 6.98 (2 H, m, H-2 MU,

3 4,5

tion in THF, 9.5 mL, 9.5 mmol) was added. The reaction H-4 MU), 6.21 (1 H, d, J7,Me 1.5, H-7 MU), 5.33 (1 H, d, J1,2 7.1,

mixture was heated at reflux for 1 h, cooled to rt and concen- H-1), 5.02 (1 H, m, H-4), 4.32 (1 H, dd, J5ax,5eq 13.0, J4,5ax/eq 5.1,

trated. The residue was dissolved in CH Cl , washed with sat. H-5ax/eq), 4.13 (1 H, dd, J₂_,₃10.7, J₃_,₄8.4, H-3), 3.94 (1 H, dd,

2

NaHCO

trated. The desired fluoro product was obtained after flash OMe BDA), 2.43 (3 H, d, Me MU), 1.37, 1.35 (6 H, 2 s, Me BDA);

chromatography. δ_C(100 MHz, CDCl ) 160.9, 159.2, 154.8, 152.1 (4 C, C-1 MU,

.1.5 Typical procedure for the preparation of thionocarbo- C-3 MU, C-6 MU, C-8 MU), 125.6 (1 C, C-5 MU), 115.4 (1 C, C-0

nates. To a solution of alcohol (1 mmol), pyridine (1.6 mL, MU), 113.5 (1 C, C-4 MU), 113.1 (1 C, C-7 MU), 104.7 (1 C, C-2

0 mmol) and DMAP (124 mg, 1 mmol) in dry CH Cl (6 mL) MU), 100.1, 100.0 (3 C, C_q_u_a_tBDA, CF ), 98.4 (1 C, C-1), 82.4

3 4

and brine, dried over MgSO , filtered and concen- H-2), 3.82 (1 H, dd, J4,5ax/eq 6.1, H-5ax/eq), 3.37, 3.32 (6 H, 2 s,

3

4

2

3

at 0 °C under Ar was added O-phenyl chlorothionoformate (1 C, C-4), 68.6, 68.5 (2 C, C-2, C-3), 64.0 (1 C, C-5), 48.2, 48.1

(

1

0.44 mL, 3.2 mmol), and the solution was stirred at rt for (2 C, OMe BDA), 18.7 (1 C, Me MU), 17.4 (2 C, Me BDA); δ

F

+

.5 h. The reaction mixture was then diluted with CH Cl , (376 MHz, CDCl ) −74.4 (s); ESI HRMS [M + H] m/z 555.1142

2

3

washed with water, 1 M HCl (only Bz-protected products), sat. calcd for C22

NaHCO and water, dried over MgSO , filtered and concen-

trated. The crude was then purified by flash chromatography general procedure for the epimerization with TBAN. Flash

to aﬀord the desired thionocarbonate. chromatography (CH Cl /acetone 12 : 1 + 0.1% Et N) aﬀorded

.1.6 Typical procedure for the radical deoxygenation. 12 (109 mg, 84%) as a white foam. M.p. 180 °C (from EP/

Argon was bubbled for 1 h at rt into a solution of the thiono- EtOAc); [α] −86.5 (c 0.99 in CHCl ); δ (250 MHz, CDCl ) 7.45

H

26

F

3

O

11S, found: 555.1146.

3

4

Triflate 10 (171 mg, 0.31 mmol) was then subjected to the

2

3

4

D

3

H

3

carbonate (0.33 mmol), Bu

3

SnH (145 μL, 0.5 mmol) and ACBN (1 H, d, J4,5 8.2, H-5 MU), 6.94 (2 H, m, H-2 MU, H-4 MU), 6.13

(50 mg) in benzene (5 mL). The reaction mixture was stirred at (1 H, d, J7,Me 1.4, H-7 MU), 5.09 (1 H, d, J1,2 7.9, H-1), 4.18 (1 H,

8

0 °C for 3 h, cooled to rt and concentrated. The residue was dd, J₂_,₃10.3, H-2), 4.08 (1 H, dd, J₅_a_x_,₅_e_q12.9, J₄_,₅_a_x_/_e_q2.0,

then purified by flash chromatography to aﬀord the desired H-5ax/eq), 3.97 (1 H, m, H-4), 3.84 (1 H, dd, J3,4 3.1, H-3), 3.72

deoxygenated product. (1 H, dd, J4,5ax/eq 1.7, H-5ax/eq), 3.31, 3.26 (6 H, 2 s, OMe BDA),

.36 (3 H, d, Me MU), 1.32, 1.328 (6 H, 2 s, Me BDA); δ_C

100 MHz, CDCl ) 161.1, 159.7, 154.8, 152.4 (4 C, C-1 MU, C-3

2,3-O-(2′,3′-dimethoxybutane- MU, C-6 MU, C-8 MU), 125.4 (1 C, C-5 MU), 114.9 (1 C, C-0

2

(

4

.2 Selected experimental data

.2.1 4-Methylumbelliferyl

3

4

2

′,3′-diyl)-β-D-xylopyranoside (5b). A suspension of 1 (300 mg, MU), 114.0 (1 C, C-4 MU), 112.7 (1 C, C-7 MU), 104.3 (1 C, C-2

0

.97 mmol) in a solution of 2,2,3,3-tetramethoxybutane MU), 100.4, 99.9 (2 C, Cquat BDA), 98.6 (1 C, C-1), 69.5 (1 C,

(0.215 mL, 1.94 mmol) and trimethyl orthoformate (0.43 mL, C-3), 67.2 (1 C, C-5), 67.0 (1 C, C-4), 66.4 (1 C, C-2), 48.1, 47.9

3

.88 mmol) in methanol (5 mL) was treated with camphor- (2 C, OMe BDA), 18.7 (1 C, Me MU), 17.7, 17.6 (2 C, Me BDA);

+

27 9

sulfonic acid (12 mg, 0.05 mmol), and the mixture was refluxed ESI HRMS [M + H] m/z 423.1649 calcd for C21H O , found:

overnight. The cooled reaction mixture was then treated with 423.1648.

powdered NaHCO₃(0.5 g), filtered and concentrated. Flash

chromatography (CH Cl /acetone 12 : 1 + 0.1% Et N) aﬀorded dimethoxybutane-2′,3′-diyl)-β-D-xylopyranoside (17). Alcohol 12

the protected carbohydrates 5a (241 mg, 59%) and 5b (152 mg, (50 mg, 0.12 mmol) was subjected to the general procedure for

7%) as pale yellow foams. 5b: [α]_D+128.6 (c 0.73 in CHCl₃); the fluorination with DAST (method A). Flash chromatography

(400 MHz, CDCl ) 7.51 (1 H, d, J4,5 8.6, H-5 MU), 6.98 (2 H, (toluene/EtOAc 4 : 1 + 0.1% Et N) aﬀorded 17 (25 mg, 49%) as

m, H-2 MU, H-4 MU), 6.18 (1 H, d, J7,Me 1.2, H-7 MU), 5.21 a white foam. [α] −163.1 (c 1 in CHCl ); δ (400 MHz, CDCl

1 H, d, J₁_,₂7.1, H-1), 4.13 (1 H, m, H-5eq), 3.98 (1 H, m, H-4), 7.53 (1 H, d, J₄_,₅8.6, H-5 MU), 7.00 (2 H, m, H-2 MU, H-4 MU),

4.2.3 4-Methylumbelliferyl

4-deoxy-4-fluoro-2,3-O-(2′,3′-

2

3

δ

H

3

D

3

H

3

)

(

3

.80 (2 H, m, H-2, H-3), 3.49 (2 H, dd, J5ax,5eq 11.7, J4,5ax 9.0, 6.19 (1 H, d, J7,Me 1.3, H-7 MU), 5.39 (1 H, d, J1,2 6.8, H-1), 4.77

H-5ax), 3.37, 3.34 (6 H, 2 s, OMe BDA), 2.53 (1 H, d, J4,OH 3.2, (1 H, dm, J4,F 52.2, H-4), 4.16 (2 H, m, H-3, H-5ax/eq), 3.91

-OH), 2.41 (3 H, d, Me MU), 1.38, 1.35 (6 H, 2 s, Me BDA); δ_C(1 H, ddd, J₂_,₃11.1, J₂_,_F1.0, H-2), 3.79 (1 H, m, H-5ax/eq), 3.37,

4

Org. Biomol. Chem.

View Article Online

Paper

Organic & Biomolecular Chemistry

3

.36 (6 H, 2 s, OMe BDA), 2.42 (3 H, d, Me MU), 1.38, 1.35

4.2.6 4-Methylumbelliferyl 4-deoxy-4-fluoro-α-L-arabinopyr-

(

6 H, 2 s, Me BDA); δ_C(100 MHz, CDCl ) 160.9, 159.5, 154.8, anoside (22). 20 (47 mg, 0.1 mmol) was subjected to the same

3

1

52.2 (4 C, C-1 MU, C-3 MU, C-6 MU, C-8 MU), 125.5 (1 C, C-5 procedure as that described for the preparation of 18 using

MU), 115.1 (1 C, C-0 MU), 113.5 (1 C, C-4 MU), 112.9 (1 C, C-7 95% TFA. Flash chromatography aﬀorded 22 (21 mg, 67%).

MU), 104.7 (1 C, C-2 MU), 99.7, 99.5 (2 C, C_q_u_a_tBDA), 98.6 (1 C, [α]_D+1.0 (c 1.02 in CHCl /MeOH 1 : 1); δ_H(400 MHz, CDCl₃/

3

C-1), 83.6 (1 C, d, J4,F 183.1, C-4), 69.9 (1 C, d, J3,F 22.3, C-3), MeOD 1 : 1) 7.60 (1 H, d, J4,5 8.8, H-5 MU), 7.06 (1 H, dd,

6

8.0 (1 C, d, J2,F 9.7, C-2), 64.4 (1 C, d, J5,F 27.2, C-5), 48.1, (2 C,

J

4,5 8.8, J2,4 2.4, H-4 MU), 7.02 (1 H, d, H-2 MU), 6.17 (1 H, d,

J₇_,_M_e1.5, H-7 MU), 4.98 (1 H, d, J₁_,₂7.3, H-1), 4.79 (1 H, dm,

4,F 48.9, H-4), 4.19 (1 H, ddd, J5ax,5eq 13.3, J5ax,F 12.2, J4,5ax 2.5,

H-5ax), 3.81 (3 H, m, H-2, H-3, H-5eq), 2.42 (3 H, d, Me MU);

.2.4 4-Methylumbelliferyl 4-deoxy-4-fluoro-β-D-xylopyrano- δ_C(100 MHz, CDCl /MeOD 1 : 1) 162.1, 160.2, 154.6, 153.8

OMe BDA), 18.7 (1 C, Me MU), 17.6, 17.5 (2 C, Me BDA); δ

F

+

(376 MHz, CDCl

3

) −190.4; ESI HRMS [M + H] m/z 425.1606

8

26FO , found: 425.1603.

J

calcd for C21

H

4

3

side (18). A mixture of 17 (80 mg, 0.19 mmol) in 95% TFA (4 C, C-1 MU, C-3 MU, C-6 MU, C-8 MU), 125.8 (1 C, C-5 MU),

1 mL) was stirred at rt for 2 h. The reaction mixture was then 114.9 (1 C, C-0 MU), 113.9 (1 C, C-4 MU), 111.9 (1 C, C-7 MU),

co-evaporated with water (3×), toluene (2×) and CH Cl . Flash 103.9 (1 C, C-2 MU), 100.7 (1 C, C-1), 88.7 (1 C, d, J₄_,_F179.2,

(

2

chromatography (CH

3%) as a white foam. M.p. 208 °C (from MeOH); [α]

c 1.42 in CHCl /MeOH 1 : 1); δ_H(400 MHz, CDCl /MeOD 1 : 1) MeOD 1 : 1) −206.2; ESI HRMS [M + H] m/z 311.0925 calcd

2

Cl

2

/MeOH 9 : 1) aﬀorded 18 (31 mg, C-4), 71.6 (1 C, d, J3,F 18.3, C-3), 70.7 (1 C, d, J2,F 1.3, C-2), 64.3

5

(

D

−62.9 (1 C, d, J5,F 20.3, C-5), 18.1 (1 C, Me MU); δ (376 MHz, CDCl /

F

3

+

3

7

.64 (1 H, d, J4,5 8.9, H-5 MU), 7.07 (1 H, dd, J2,4 2.4, H-4 MU), for C15

6

H16FO , found: 311.0927.

7

.03 (1 H, d, H-2 MU), 6.20 (1 H, d, J7,Me 1.7, H-7 MU), 5.09 (1 4.2.7 4-Methylumbelliferyl

2,3-O-(2′,3′-dimethoxybutane-

H, d, J₁_,₂6.9, H-1), 4.50 (1 H, dm, J₄_,_F49.7, H-4), 4.16 (1 H, 2′,3′-diyl)-4-O-(phenylthio)thionocarbonyl β-D-xylopyranoside

ddd, J5ax,5eq 12.3, J5eq,F 7.5, J4,5eq 5.2, H-5eq), 3.77 (1 H, ddd, J3, (24). Alcohol 5b (0.468 g, 1.1 mmol) was subjected to the

F

16.5, J2,3 = J3,4 8.5, H-3), 3.66 (1 H, ddd, J5ax,F 8.4, J4,5ax 6.5, general procedure for the preparation of thionocarbonate.

H-5ax), 3.60 (1 H, dd, H-2), 2.45 (3 H, d, Me MU); δ (100 MHz, Flash chromatography (petroleum ether/EtOAc 2 : 1 + 0.1%

C

CDCl

3

/MeOD 1 : 1) 162.1, 160.1, 154.6, 153.9 (4 C, C-1 MU, C-3 Et

3

N) aﬀorded 24 (466 mg, 76%) as a white foam. [α]

D

−174.9

MU, C-6 MU, C-8 MU), 125.9 (1 C, C-5 MU), 115.0 (1 C, C-0 (c 1.01 in CHCl

3

); δ (400 MHz, CDCl ) 7.53 (1 H, d, J4,5 8.6,

H

3

MU), 113.8 (1 C, C-4 MU), 111.9 (1 C, C-7 MU), 104.0 (1 C, C-2 H-5 MU), 7.44 (2 H, m, o-Ph), 7.32 (1 H, m, p-Ph), 7.15 (2 H, m,

MU), 100.8 (1 C, C-1), 89.3 (1 C, d, J4,F 180.7, C-4), 73.7 (1 C, d, m-Ph), 7.01 (2 H, m, H-2 MU, H-4 MU), 6.20 (1 H, d, J7,Me 1.7,

J3,F 19.0, C-3), 72.3 (1 C, d, J2,F 8.6, C-2), 62.7 (1 C, d, J5,F 28.2, H-7 MU), 5.49 (1 H, m, H-4), 5.42 (1 H, d, J1,2 6.7, H-1), 4.42

C-5), 18.0 (1 C, Me MU); δ (376 MHz, CDCl /MeOD 1 : 1) (1 H, dd, J₅_a_x_,₅_e_q13.4, J₄_,₅_a_x_/_e_q4.7, H-5ax/eq), 4.23 (1 H, dd,

F

3

+

−

C

198.8; ESI HRMS [M

+

H] m/z 311.0925 calcd for

J

2,3 10.8, J3,4 8.3, H-3), 4.03 (1 H, dd, H-2), 3.85 (1 H, dd, J4,5ax/eq

4.9, H-5ax/eq), 3.39, 3.33 (6 H, 2 s, OMe BDA), 2.43 (3 H, d, Me

4-deoxy-4-fluoro-2,3-O-(2′,3′- MU), 1.39, 1.36 (6 H, 2 s, Me BDA); δ (100 MHz, CDCl ) 194.1

15

H

16FO

6

, found: 311.0924.

4

.2.5 4-Methylumbelliferyl

C

3

dimethoxybutane-2′,3′-diyl)-α-L-arabinopyranoside (20). The (1 C, CvS), 161.1, 159.5, 154.8, 153.3, 152.3 (5 C, Cquat Ph, C-1

compound was prepared from either 5b or 10. From 5b: MU, C-3 MU, C-6 MU, C-8 MU), 129.6 (2 C, o-Ph), 126.7 (1 C,

Alcohol 5b (40 mg, 0.09 mmol) was subjected to the general p-Ph), 125.5 (1 C, C-5 MU), 121.8 (1 C, m-Ph), 115.1 (1 C, C-0

procedure for the fluorination with DAST (method B). Flash MU), 113.6 (1 C, C-4 MU), 112.9 (1 C, C-7 MU), 104.6 (1 C, C-2

3

chromatography (toluene/EtOAc 4 : 1 + 0.1% Et N) aﬀorded 20 MU), 99.7 (2 C, Cquat BDA), 98.7 (1 C, C-1), 79.6 (1 C, C-4), 68.7

(

16 mg, 41%) and elimination product 21 (22%). From 10: (1 C, C-2), 68.3 (1 C, C-3), 63.4 (1 C, C-5), 48.2, 48.1 (2 C, OMe

+

Triflate 10 (0.114 g, 0.21 mmol) was subjected to the general BDA), 18.7 (1 C, Me MU), 17.6, 17.5 (2 C, Me BDA); ESI HRMS

+

procedure for the fluorination with TBAF. 20 (32 mg, 36%) was [M + H] m/z 559.1632 calcd for C28

isolated along with 21 (15%). 20: [α]_D−59.9 (c 0.98 in CHCl₃); 4.2.8 4-Methylumbelliferyl 4-deoxy-2,3-O-(2′,3′-dimethoxy-

(400 MHz, CDCl ) 7.52 (1 H, d, J4,5 8.7, H-5 MU), 7.01 (2 H, butane-2′,3′-diyl)-β-D-xylopyranoside (26). 24 (0.46 g,

31

H O10S, found: 559.1629.

δ

H

3

m, H-2 MU, H-4 MU), 6.20 (1 H, d, J7,Me 1.3, H-7 MU), 5.18 (1 H, 0.82 mmol) was subjected to the general procedure for the

d, J₁_,₂7.9, H-1), 4.81 (1 H, dm, J₄_,_F52.2, H-4), 4.30 (1 H, dd, radical deoxygenation. Flash chromatography (petroleum

J

5ax,5eq = J5eq,F 12.1, J4,5eq 1.7, H-5eq), 4.25 (1 H, dd, J2,3 10.6, ether/EtOAc 2 : 1 + 0.1% Et

H-2), 3.94 (1 H, ddd, J3,F 29.2, J3,4 2.5, H-3), 3.81 (1 H, dd, J5ax,F white foam. [α] −86.9 (c 1.08 in CHCl

4.6, J₄_,₅_a_x< 1, H-5ax), 3.39, 3.33 (6 H, 2 s, OMe BDA), 2.42 (3 H, 7.50 (1 H, d, J₄_,₅8.6, H-5 MU), 6.98 (2 H, m, H-2 MU, H-4 MU),

d, Me MU), 1.42, 1.35 (6 H, 2 s, Me BDA); δ (100 MHz, CDCl 6.17 (1 H, s, H-7 MU), 5.10 (1 H, d, J1,2 7.8, H-1), 4.12 (1 H, m,

61.0, 159.6, 154.8, 152.3 (4 C, C-1 MU, C-3 MU, C-6 MU, C-8 H-5ax/eq), 3.91 (1 H, m, H-3), 3.75 (1 H, dd, J2,3 9.6, H-2), 3.66

MU), 125.4 (1 C, C-5 MU), 115.1 (1 C, C-0 MU), 114.1 (1 C, C-4 (1 H, m, H-5ax/eq), 3.38, 3.31 (6 H, 2 s, OMe BDA), 2.41 (3 H, s,

MU), 112.9 (1 C, C-7 MU), 104.2 (1 C, C-2 MU), 100.5, 99.9 (2 C, Me MU), 1.89 (2 H, m, H-4), 1.34 (6 H, s, Me BDA); δ

quat BDA), 98.6 (1 C, C-1), 86.5 (1 C, d, J4,F 184.8, C-4), 68.4 (100 MHz, CDCl ) 161.1, 159.8, 154.8, 152.4 (4 C, C-1 MU, C-3

1 C, d, J₃_,_F17.2, C-3), 66.3 (1 C, d, J₂_,_F1.6, C-2), 65.8 (1 C, d, J₅_,_FMU, C-6 MU, C-8 MU), 125.4 (1 C, C-5 MU), 114.9 (1 C, C-0

0.4, C-5), 48.2, 48.0 (2 C, OMe BDA), 18.7 (1 C, Me MU), 17.6 MU), 113.9 (1 C, C-4 MU), 112.7 (1 C, C-7 MU), 104.2 (1 C, C-2

3

N) aﬀorded 26 (206 mg, 62%) as a

D

3

); δ (400 MHz, CDCl

H

3

)

3

C

3

)

1

C

(

2

3

+

(2 C, Me BDA); δ

F

(376 MHz, CDCl

3

) −204.1; ESI HRMS [M + H]

MU), 99.9, 99.8 (2 C, Cquat BDA), 98.6 (1 C, C-1), 71.4 (1 C, C-2),

67.4 (1 C, C-3), 62.9 (1 C, C-5), 48.1, 47.9 (2 C, OMe BDA), 29.6

m/z 425.1606 calcd for C H FO , found: 425.1607.

21

26

8

Org. Biomol. Chem.

View Article Online

Organic & Biomolecular Chemistry

Paper

+

(

1 C, C-4), 18.7 (1 C, Me MU), 17.7, 17.6 (2 C, Me BDA); ESI

6.9, 300 mM NaCl, 10 mM DTT, 20 µM MnCl

2

, 0.5% Triton,

+

HRMS [M + H] m/z 407.1400 calcd for C H O , found: 1 mM EDTA, and 5% (v/v) glycerol). The suspended cells were

2

1

27 8

4

07.1698.

.2.9 4-Methylumbelliferyl 4-deoxy-β-D-xylopyranoside (27). collected from the supernatant after centrifugation for 30 min

6 (50 mg, 0.12 mmol) was subjected to the same procedure as at 12 000g. The supernatant was supplemented with endonu-

lysed using a French press at 25 kpsi. Soluble proteins were

4

2

that described for the preparation of 18. Flash chromatography clease (Pierce Universal Nuclease, Thermo Fisher) and 2 mM

(

CH

11 mg, 33%). M.p. 185 °C (from EP/EA); [α]_D−41.0 (c 1, clarified by filtration (0.2 μm Supor® Membrane; PALL-Life

3 : 1); δ (400 MHz, MeOD/CDCl 3 : 1) 7.63 (1 H, Science) and applied onto a 5 mL glutathione Sepharose

2 2 2

Cl /MeOH 9 : 1) aﬀorded the deprotected compound 27 MnCl . After incubation (1 hour, 4 °C), the supernatant was

MeOH/CHCl

3

H

3

d, J4,5 8.7, H-5 MU), 7.07 (1 H, dd, J2,4 2.4, H-4 MU), 7.02 (1 H, column (GSTrap 4B; GE Healthcare) connected to an AKTA

d, H-2 MU), 6.19 (1 H, d, J₇_,_M_e1.3, H-7 MU), 4.94 (1 H, d, J₁_,₂Prime Plus instrument (GE Healthcare). Proteins were eluted

7

.2, H-1), 4.04 (1 H, ddd, J5ax,5eq 12.0, J4ax,5eq 4.9, J4eq,5eq 2.4, as 2 mL fractions using an elution buﬀer (50 mM Tris-base,

H-5eq), 3.73 (1 H, ddd, J3,4ax 10.8, J2,3 8.6, J3,4eq 5.0, H-3), 3.67 pH 8.0, 10 mM reduced glutathione and 150 mM NaCl). The

1 H, ddd, J₄_a_x_,₅_a_x11.8, J₄_e_q_,₅_a_x2.4, H-5ax), 2.45 (3 H, d, Me protein purity of the eluted fractions was evaluated by 12%

MU), 2.03 (1 H, dddd, J4ax,4eq 13.3, H-4eq), 1.75 (1 H, dddd, (w/v) SDS-PAGE analysis, followed by staining with Coomassie

H-4ax); δ (100 MHz, MeOD/CDCl 3 : 1) 162.2, 160.4, 154.6, Brilliant Blue. Fractions containing the pure protein were con-

53.9 (4 C, C-1 MU, C-3 MU, C-6 MU, C-8 MU), 125.8 (1 C, C-5 centrated and used to determine the kinetic parameters of the

MU), 114.8 (1 C, C-0 MU), 113.8 (1 C, C-4 MU), 111.8 (1 C, C-7 enzyme. The protein concentration was measured using the

(

C

3

1

3

2

MU), 103.8 (1 C, C-2 MU), 101.0 (1 C, C-1), 74.6 (1 C, C-2), 70.6 Bradford technique.

(

1 C, C-3), 61.6 (1 C, C-5), 32.3 (1 C, C-4), 18.1 (1 C, Me MU);

4.3.3 Determination of the in vitro kinetic parameters of

+

ESI HRMS [M + H] m/z 293.1019 calcd for C15

H

17

O

6

, found: hβ4GalT7ΔNt60. Kinetic constants, i.e. the Michaelis constant

2

93.1018.

m

(K ) and the catalytic constant (kcat), were determined towards

modified xylosides by incubating the purified 6His-GST-

hβ4GalT7ΔNt60 (100 ng) with increasing concentrations of

4

.3 Biological assays

4

.3.1 Expression vector construction. The human β4GalT7 xylosides (0–5 mM) for 30 min at 37 °C, in the presence of a

sequence (GenBank® nucleotide sequence accession number fixed concentration of UDP-Gal (1 mM) and 10 mM MnCl , in

2

NM_007255) was cloned by PCR amplification from a placenta a Bis-Tris buﬀer (50 mM, pH 6.5). The incubation mixture was

2

6b

cDNA library (Clontech), as described previously.

For bac- then centrifuged at 10 000g for 1 min at 4 °C. The samples

terial expression, a truncated form of β4GalT7 was expressed were analyzed by high performance liquid chromatography

as a fusion protein of a 6His tag and glutathione S-transferase (HPLC) with a reverse phase C18 analytical column (xBridge,

(6His-GST-hβ4GalT7). A sequence lacking the codons of the 4.6 × 150 mm, 5 μm, Waters) using Waters equipment

first 60 N-terminal amino acids was amplified from the full- (Alliance Waters e2695) coupled to a UV detector (Shimadzu

length cDNA and subcloned into NcoI and NotI sites of SPD-10A). Kinetic parameters were determined by nonlinear

pETM30 to produce the plasmid pETM30-hβ4GalT7ΔNt60. For least-squares regression analysis of the data fitted to the

the in cellulo analysis of decorin core protein glycosylation, the Michaelis–Menten rate equation or the substrate inhibition

human decorin cDNA sequence (GenBank® accession number rate equation using the curve-fitter program of GraphPad

NM_001920.3) was cloned by PCR amplification from a pla- Prism 5 (GraphPad Software, USA).

centa cDNA library and the full-length cDNA sequence was

inserted into pcDNA3.1(+) to produce pcDNA-decorinHis.

4.3.4 In vitro competition assays of hβ4GalT7ΔNt60

activity by modified xylosides. The in vitro inhibitory activities

3

1

4

.3.2 Bacterial expression and purification of the soluble of modified xylosides were evaluated using 100 ng of purified

form of hβ4GalT7ΔNt60. To express 6His-GST- protein incubated for 30 min at 37 °C in a Bis-Tris buﬀer

hβ4GalT7ΔNt60, E. coli Rosetta2 (DE3) bacteria were trans- (50 mM, pH 6.5), 10 mM MnCl and 1 mM UDP-Gal in the

2

formed with the pETM30-hβ4GalT7ΔNt60 plasmid. The recom- presence of a fixed concentration of the XylMU acceptor sub-

binant bacteria were then selected based on kanamycin resis- strate (0.3 mM) and various concentrations of the xyloside

tance (50 μg mL⁻). A single recombinant colony was inocu- analogs (0, 2.5 and 5 mM). The amount of the reaction

lated into fresh Luria–Bertani (LB) medium and cultured over- product (Gal-XylMU) was measured by HPLC, as described

night at 37 °C on an orbital shaker (160 rpm). The overnight above. Inhibition or activation percentages were obtained from

culture was transferred into fresh LB medium, supplemented enzymatic activity determined in the presence or absence of

with 50 μg mL⁻kanamycin and incubated at 37 °C until the modified xylosides.

1

OD₆₀₀value reached 0.6. The expression of hβ4GalT7DNt60

4.3.5 In cellulo analysis of decorin core protein glycosyla-

was induced by the addition of 0.5 mM isopropyl-β-D-thiogalac- tion. For the eukaryotic expression of human decorin, HeLa

topyranoside (IPTG, Sigma-Aldrich) to the bacterial suspension cells (American Type Culture Collection, ATCC) were cultured

that was then incubated overnight at 25 °C under continuous in Dulbecco’s modified Eagle’s medium (DMEM) sup-

shaking (160 rpm). The bacterial cells were then harvested by plemented with 10% (v/v) fetal bovine serum (FBS, Invitrogen),

−

1

centrifugation at 6000g for 20 min at 4 °C. The pellets were penicillin (100 units per ml)–streptomycin (100 mg ml ), and

resuspended in a lysis buﬀer (50 mM sodium phosphate, pH 1 mM glutamine. The pcDNA-decorinHis plasmid was trans-

Org. Biomol. Chem.

View Article Online

Paper

Organic & Biomolecular Chemistry

fected into HeLa cells grown to 80% confluency, using

PolyEthylenImine (PEI, Polysciences, Inc.) according to the

6 (a) N. B. Schwartz, L. Calligan, P.-L. Ho and A. Dorfman,

Proc. Natl. Acad. Sci. U. S. A., 1974, 71, 4047;

(b) H. C. Robinson, M. J. Brett, P. Tralaggan, D. A. Lowther

and M. Okayama, Biochem. J., 1975, 148, 25.

manufacturer’s recommendations, and selected with geneticin

1

(

G-418, 1 mg ml⁻, Sigma). Recombinant cells were cultured in

a 6 well-plate in DMEM complete medium, starved for 24 h,

and supplemented with modified xylosides (0–300 µM). After

7 (a) Y. Tsuzuki, T. K. N. Nguyen, D. R. Garud, B. Kuberan

and M. Koketsu, Bioorg. Med. Chem. Lett., 2010, 20, 7269;

(b) D. R. Garud, V. M. Tran, X. V. Victor, M. Koketsu and

B. Kuberan, J. Biol. Chem., 2008, 283, 28881.

7

2 h, the cell medium was collected, concentrated by centrifu-

gation at 4 °C for 45 min at 7000g using an Amicon Ultra4 3

MWCO concentrating system (Merck, Millipore, Germany) and

analyzed by SDS-PAGE (10 μg of protein per lane). The glycosy-

lation level of the decorin core protein was monitored by

immunoblotting using a 1 : 500 dilution of a primary polyclo-

nal anti-human decorin antibody (R&D Systems) and a 1 : 5000

dilution of a secondary anti-mouse antibody coupled to horse-

radish peroxidase (Cell Signaling). The expression level of gly-

cosylated decorin in cells treated with 0.6% (v/v) DMSO (xylo-

side diluent) was used as a control. The expression level of gly-

cosylated decorin in cells treated with 300 µM XylMU was used

as a positive competition control.

8 (a) A. Siegbahn, K. Thorsheim, J. Stahle, S. Manner,

C. Hamark, A. Persson, E. Tykesson, K. Mani,

G. Westergren-Thorsson, G. Widmalm and U. Ellervik, Org.

Biomol. Chem., 2015, 13, 3351; (b) A. Siegbahn, S. Manner,

A. Persson, E. Tykesson, K. Holmqvist, A. Ochocinska,

J. Roennols, A. Sundin, K. Mani, G. Westergren-Thorsson,

G. Widmalm and U. Ellervik, Chem. Sci., 2014, 5, 3501;

(c) A. Siegbahn, U. Aili, A. Ochocinska, M. Olofsson,

J. Roennols, K. Mani, G. Widmalm and U. Ellervik, Bioorg.

Med. Chem., 2011, 19, 4114.

9 M. Saliba, N. Ramalanjaona, S. Gulberti, I. Bertin-Jung,

A. Thomas, S. Dahbi, C. Lopin-Bon, J.-C. Jacquinet,

C. Breton, M. Ouzzine and S. Fournel-Gigleux, J. Biol.

Chem., 2015, 290, 7658.

Conﬂicts of interest

10 S. Kaneko, M. Kitaoka, A. Kuno and K. Hayashi, Biosci.,

Biotechnol., Biochem., 2000, 64, 741.

There are no conflicts to declare.

1

1 E. V. Eneyskaya, D. R. Ivanen, K. A. Shabalin,

A. A. Kulminskaya, L. V. Backinowsky, H. Brumer III and

K. N. Neustroev, Org. Biomol. Chem., 2005, 3, 146.

Acknowledgements

1

2 (a) S. Oscarson and P. Svahnberg, J. Chem. Soc., Faraday

Trans. 1, 2001, 873; (b) D. J. Jenkins and B. V. L. Potter,

J. Chem. Soc., Faraday Trans. 1, 1998, 41;

This work was supported by the Agence Nationale de la

Recherche (ANR Meca-GT ANR-13-BSV8-0011-01), grants from

the Région Lorraine and FEDER, University of Lorraine to

N. Ramalanjaona, Labex SynOrg Grant No. ANR-11-LABX-0029,

CNRS in the framework of GDR GAG (GDR 3739), and grants

from the International Associated Laboratory (SFGEN) between

the CNRS-UL and the University of Dundee.

(

c) L. S. Khasanova, F. A. Gimalova, S. A. Torosyan,

A. A. Fatykhov and M. S. Miftakhov, Russ. J. Org. Chem.,

011, 47, 1125; (d) B. Kongkathip, N. Kongkathip and

2

J. Rujirawanich, Synth. Commun., 2014, 44, 2248.

3 J. L. Montchamp, F. Tian, M. E. Hart and J. W. Frost, J. Org.

Chem., 1996, 61, 3897.

4 Y. Demizu, Y. Kubo, H. Miyoshi, T. Maki, Y. Matsumara,

N. Moriyama and O. Onomura, Org. Lett., 2008, 10,

5075.

1

Notes and references

1

(a) M. Bernfield, M. Götte, P. W. Park, O. Reizes,

M. L. Fitzgerald, J. Lincecum and M. Zako, Annu. Rev. 15 (a) T. Liang, C. N. Neumann and T. Ritter, Angew. Chem.,

Biochem., 1999, 68, 729; (b) J. S. Clark, P. N. Bishop and

A. J. Day, Front. Immunol., 2013, 4, 412.

Int. Ed., 2013, 52, 8214; (b) K. Dax, M. Albert, J. Ortner and

B. J. Paul, Carbohydr. Res., 2000, 327, 47.

2

3

4

M. M. Fuster and J. D. Esko, Nat. Rev. Cancer., 2005, 5, 526.

J. Halper, Adv. Exp. Med. Biol., 2014, 802, 49.

(a) M. Götte, D. Spillmann, G. W. Yip, E. Versteeg, 17 K. Plé, M. Chwalek and L. Voutquenne-Nazabadioko,

F. G. Echtermeyer, T. H. van Kuppevelt and L. Kiesel, Hum. Eur. J. Org. Chem., 2004, 1588.

16 R. Albert, K. Dax, W. R. Link and A. E. Stütz, Carbohydr.

Res., 1983, 118, C5–C6.

Mol. Genet., 2008, 17, 996; (b) T. Okajima, K. Yoshida, 18 J. A. Wright and J. J. Fox, Carbohydr. Res., 1970, 13, 297.

T. Kondo and K. Furukawa, J. Biol. Chem., 1999, 274, 22915; 19 M. Ueki, K. Kai, M. Amemiya, H. Horino and H. Oyamada,

(

2

c) K. Furukawa and T. Oakjima, Biochim. Biophys. Acta,

002, 1573, 377; (d) M. Götte and H. Kresse, Biochem. 20 S. Manner and U. Ellervik, Synlett, 2014, 1271.

Genet., 2005, 43, 65; (e) C. Bui, I. Talhaoui, M. Chabel, 21 (a) H. C. Robinson, M. J. Brett, P. J. Tralaggan,

J. Chem. Soc., Chem. Commun., 1988, 415.

G. Mulliert, M. W. Coughtrie, M. Ouzzine and S. Fournel-

Gigleux, FEBS Lett., 2010, 584, 3962.

R. Almeida, S. B. Levery, U. Mandel, H. Kresse,

T. Schwientek, E. P. Bennet and H. Clausen, J. Biol. Chem.,

D. A. Lowther and M. Okayama, Biochem. J., 1975, 148, 25;

(b) J. A. Robinson and H. C. Robinson, Biochem. J., 1981,

250, 5200; (c) S. O. Kolset, K. Sakurai, I. Ivhed, A. Oevervatn

and S. Suzuki, Biochem. J., 1990, 265, 637; (d) M. Sobue,

H. Habuchi, K. Ito, H. Yonekura, K. Oguri, K. Sakurai,

5

1

999, 274, 26165.

Org. Biomol. Chem.

View Article Online

Organic & Biomolecular Chemistry

Paper

S. Kamohara, Y. Ueno, R. Noyori and S. Suzuki, Biochem. J., 27 (a) J. Wen, J. Xiao, M. Rahdar, B. P. Choudhury,

1

987, 241, 591.

2 M. Jacobsson, K. Mani and U. Ellervik, Bioorg. Med. Chem.,

007, 15, 5283.

J. Cui, G. S. Taylor, J. D. Esko and J. E. Dixona, Proc. Natl.

Acad. Sci. U. S. A., 2014, 111, 15723; (b) S. Gulberti,

V. Lattard, M. Fondeur, J. C. Jacquinet, G. Mulliert,

P. Netter, J. Magdalou, M. Ouzzine and S. Fournel-Gigleux,

J. Biol. Chem., 2005, 280, 1417.

2

3 (a) A. L. Myers, V. V. Upreti, M. Khurana and N. D. Eddington,

J. Clin. Pharmacol., 2008, 48, 1158; (b) D. Pless, J. N. Gouze,

C. Senay, R. Herber, P. Leroy, V. Barberousse, S. Fournel- 28 J. F. García-García, G. Corrales, J. Casas, A. Fernández-

Gigleux and J. Magdalou, Drug Metab. Dispos., 1999, 27, 588.

Mayoralas and E. García-Junceda, Mol. Biosyst., 2011, 7,

2

4 Y. Nishimura, S. Maeda, S. Ikushiro, P.L. Mackenzie,

1312.

Y. Ishii and H. Yamada, Biochem. Biophys. Res. Commun., 29 J. Ronnols, S. Manner, A. Siegbahn, U. Ellervik and

011, 404, 86. G. Widmalm, Org. Biomol. Chem., 2013, 11, 5465.

5 K. Raman, M. Ninomiya, T. K. Nguyen, Y. Tsuzuki, 30 F. N. Lugemwa, A. K. Sarkar and J. D. Esko, J. Biol. Chem.,

2

M. Koketsu and B. Kuberan, Biochem. Biophys. Res.

Commun., 2011, 404, 86.

6 (a) B. Ramakrishnan and P. K. Qasba, J. Biol. Chem., 2010,

1996, 271, 19159.

31 I. Talhaoui, C. Bui, R. Oriol, G. Mulliert, S. Gulberti,

P. Netter, M. W. Coughtrie, M. Ouzzine and S. Fournel-

Gigleux, J. Biol. Chem., 2010, 285, 37342.

2

85, 15619; (b) Y. Tsutsui, B. Ramakrishnan and

P. K. Qasba, J. Biol. Chem., 2013, 288, 31963.

32 M. M. Bradford, Anal. Biochem., 1976, 72, 248.

Org. Biomol. Chem.

Article Doi

Synthesis of a library of variously modified 4-methylumbelliferyl xylosides and a structure-activity study of human β4GalT7

DOI: 10.1039/c7ob02530k

Source and publish data:

Authors:

Article abstract of DOI:10.1039/c7ob02530k

Full text of DOI:10.1039/c7ob02530k

Products guided by the article

R&D Labs maybe for 145370-04-3

Relevant to this article

Hot Product