An acceptor analogue of β-1,4-galactosyltransferase: Substrate, inhibitor, or both?

Article abstract of DOI:10.1016/j.carres.2017.08.012

Many glycosyltransferase inhibitors in the literature are structurally derived from the donor or acceptor substrate of the respective enzyme. A representative example is 2-naphthyl β-D-GlcNAc, a synthetic GlcNAc glycoside that has been reported as a galactosyltransferase inhibitor. This GlcNAc derivative is attractive as a chemical tool compound for biological and biochemical studies because of its reported potency as an inhibitor, and its short and straightforward synthesis from readily available starting materials. We report that in our hands, 2-naphthyl β-D-GlcNAc behaved, unexpectedly, as an acceptor substrate of the inverting β-1,4-galactosyltransferase (β-1,4-GalT) from bovine milk. This substrate activity has not previously been described. We found that 2-naphthyl β-D-GlcNAc can be an acceptor substrate both for recombinantly expressed β-1,4-GalT, and for a commercial batch of the same enzyme, and both in the presence and absence of bovine serum albumin (BSA). As expected for a full acceptor substrate, this substrate activity was time- and concentration-dependent. Additional experiments show that the observed inhibitor/substrate switch is facilitated by a phosphatase that is an essential component of our enzyme-coupled glycosyltransferase assay. These findings suggest that the behaviour of 2-naphthyl β-D-GlcNAc and related acceptor-based glycosyltransferase inhibitors is strongly dependent on the individual assay conditions. Our results therefore have important implications for the use of 2-naphthyl β-D-GlcNAc and related glycosides as tool compounds in glycobiology and glycobiochemistry.

Full text of DOI:10.1016/j.carres.2017.08.012

Accepted Manuscript
An acceptor analogue of β-1,4-galactosyltransferase: Substrate, inhibitor, or both?
Jingqian Jiang, Gerd K. Wagner
PII:
S0008-6215(17)30466-4
10.1016/j.carres.2017.08.012
CAR 7437
DOI:
Reference:
To appear in: Carbohydrate Research
Received Date: 3 July 2017
Revised Date: 23 August 2017
Accepted Date: 23 August 2017
Please cite this article as: J. Jiang, G.K. Wagner, An acceptor analogue of β-1,4-galactosyltransferase:
Substrate, inhibitor, or both?, Carbohydrate Research (2017), doi: 10.1016/j.carres.2017.08.012.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
Graphical Abstract

ACCEPTED MANUSCRIPT
An acceptor analogue of β-1,4-galactosyltransferase: substrate, inhibitor, or both?
Jingqian Jiang & Gerd K. Wagner*
King’s College London, Department of Chemistry, Faculty of Natural & Mathematical
Sciences, Britannia House, 7 Trinity Street, London, SE1 1DB, UK. Phone: +44 (0)20 7848
1926; e-mail: gerd.wagner@kcl.ac.uk
*
Corresponding author
2
3 August 2017
Keywords: glycosyltransferase, acceptor, inhibitor, substrate
1

ACCEPTED MANUSCRIPT
Abstract
Many glycosyltransferase inhibitors in the literature are structurally derived from the donor or
acceptor substrate of the respective enzyme. A representative example is 2-naphthyl β-
D-
GlcNAc, a synthetic GlcNAc glycoside that has been reported as a galactosyltransferase
inhibitor. This GlcNAc derivative is attractive as a chemical tool compound for biological
and biochemical studies because of its reported potency as an inhibitor, and its short and
straightforward synthesis from readily available starting materials. We report that in our
hands, 2-naphthyl β-
inverting β-1,4-galactosyltransferase (β-1,4-GalT) from bovine milk. This substrate activity
has not previously been described. We found that 2-naphthyl β- -GlcNAc can be an acceptor
D-GlcNAc behaved, unexpectedly, as an acceptor substrate of the
D
substrate both for recombinantly expressed β-1,4-GalT, and for a commercial batch of the
same enzyme, and both in the presence and absence of bovine serum albumin (BSA). As
expected for a full acceptor substrate, this substrate activity was time- and concentration-
dependent. Additional experiments show that the observed inhibitor/substrate switch is
facilitated by a phosphatase that is an essential component of our enzyme-coupled
glycosyltransferase assay. These findings suggest that the behaviour of 2-naphthyl β-D-
GlcNAc and related acceptor-based glycosyltransferase inhibitors is strongly dependent on
the individual assay conditions. Our results therefore have important implications for the use
of 2-naphthyl β-
D
-GlcNAc and related glycosides as tool compounds in glycobiology and
glycobiochemistry.
2

ACCEPTED MANUSCRIPT
Introduction
Glycosyltransferases (GTs) are key enzymes for glycan biosynthesis, which catalyse the
transfer of a sugar from a glycosyl donor, e.g. a sugar-nucleotide, to a suitable acceptor [1].
GTs are involved in many fundamental biological processes contingent on glycan epitopes,
such as cell adhesion and cancer metastasis [2,3]. GT inhibitors are therefore sought after as
chemical tools for glycobiology and as potential lead compounds for drug discovery [4].
An important member of the GT family is the inverting galactosyltransferase β-1,4-GalT,
which catalyses the transfer of D-galactose from a UDP-α-D-galactose (UDP-Gal) donor to N-
acetylglucosamine (GlcNAc)-containing acceptors [5]. The resulting Galβ-1,4-GlcNAc units
are found in many different glycoconjugates, and an isoform of β-1,4-GalT has recently been
identified as a major control point for glycan branching in N-linked glycosylation [6]. β-1,4-
GalT has been used extensively as a model target for the development of GT inhibitors,
including substrate analogues based on either donor [7,8] or acceptor [9]. Amongst the most
potent acceptor-based β-1,4-GalT inhibitors are hydrophobic GlcNAc glycosides such as 2-
naphthyl β-
reported by Wong and co-workers [10]. Due to the presence of the 4-OH group, in principle,
could still behave as a substrate towards β-1,4-GalT. However, two independent studies
D-GlcNAc 1 (Fig. 1), for which low micromolar inhibitory activity has been
1
have described 1 as practically devoid of acceptor substrate activity [10,11]. Similar profiles
towards GalTs have also been reported for other 2-naphthyl glycosides (Fig. 1) [12,13].
For an ongoing programme on the discovery of new GT inhibitor chemotypes, we required an
established β-1,4-GalT inhibitor as a positive control. In the absence of a commercially
available β-1,4-GalT inhibitor, we selected 2-naphthyl β-D-GlcNAc 1 because of its potency
[
10] and simple, two-step synthesis [14]. The only structural difference between 1 and the
natural β-1,4-GalT acceptor substrate GlcNAc is the presence of the naphthyl aglycon. It has
3

ACCEPTED MANUSCRIPT
been reported that the presence of this substituent is sufficient to turn 1 from a β-1,4-GalT
substrate into an inhibitor [11]. The inhibitory activity of 1 and several related GlcNAc
derivatives has been attributed in particular to the lipophilicity of this substituent in the
anomeric position [11].
Herein, we report that in our hands, 1 acted not as an inhibitor but, unexpectedly, as an
acceptor substrate of β-1,4-GalT. These results are in contrast to the previously reported
inhibitory activity of this GlcNAc derivative towards β-1,4-GalT [10,11]. This substrate
activity of 1 was observed both with a batch of bovine β-1,4-GalT recombinantly expressed
in our own laboratory, and a commercial batch of the same enzyme, and both in the presence
and absence of bovine serum albumin (BSA). Further experiments revealed that this
unexpected substrate activity is promoted by a phosphatase, which is an essential component
of our GT assay mixture. Galactosylation of 1 was observed by LC/MS both in the presence
and absence of phosphatase. This profile suggests that this secondary enzyme reveals the
latent acceptor substrate activity of 1 by altering the equilibrium position of several
interconnected enzymatic reactions in the assay. This observation has important implications
for the use of 1 and related glycosides as a tool compounds in glycobiology.
Results
2
-Naphthyl β-D-GlcNAc 1 was synthesized in two steps according to the literature procedure
[
14]. Nucleophilic substitution of commercial 2-acetamido-3,4,6-tri-O-acetyl-2-deoxy-α-D-
glucopyranosyl chloride 2 with 2-naphthol gave the fully acetylated derivative 3, which was
deprotected with 0.05M sodium methoxide to afford the desired 1-(2-naphthyl) 2-acetamido-
2
-deoxy-
β
-D-glucopyranoside 1 (Scheme 1). Results from the analytical characterisation of 1
1
13
by H- and C- NMR and mass spectrometry were consistent with data reported in the
4

ACCEPTED MANUSCRIPT
literature [14]. In particular, the β-configuration at the anomeric position was unambiguously
established based on the coupling constant between H-1 and H-2 (J = 8.4 Hz). This J value is
indicative of the trans orientation of H-1 and H-2, and hence the β-configuration of the
aglycon at the anomeric position of 1.
Although 1 has been reported as a β-1,4-GalT inhibitor, this monosaccharide can, in principle,
also behave as a substrate towards that enzyme. With sufficient quantities of compound 1 in
hand, we therefore decided to first test the potential substrate activity of 1 towards bovine β-
1
,4-GalT. For these experiments, we used a recently developed biochemical
glycosyltransferase assay [15,16]. In this coupled assay, a phosphatase is used to
quantitatively release two equivalents of inorganic phosphate from each molecule of UDP,
the secondary product of the β-1,4-GalT reaction. The free phosphate is then quantified
colorimetrically with a malachite green reagent [15,16]. For the substrate experiments, we
used 1 in place of the standard acceptor substrate GlcNAc. In initial experiments, we
evaluated the effects of acceptor concentration and incubation time on β-1,4-GalT activity
with either 1 or GlcNAc as the acceptor substrate. Experiments with either the standard
acceptor substrate GlcNAc or the putative inhibitor 1 were conducted in parallel on the same
microplate. To account for the potential non-specific hydrolysis of the UDP-Gal donor
substrate, we also included relevant control wells on each microplate (no acceptor, but
otherwise identical conditions). We also carried out separate control experiments to identify
potential interference of 1 with the colorimetric readout of our assay (no enzyme, but
otherwise identical conditions).
Significant β-1,4-GalT activity was observed in both cases, whether GlcNAc or GlcNAc
derivative 1 was present as potential acceptor substrate (Fig. 2A). The observed β-1,4-GalT
activity was dependent on the concentration of the respective acceptor. GlcNAc derivative 1
itself does not interfere with the colorimetric readout of our assay, as evidenced by the lack of
5

ACCEPTED MANUSCRIPT
signal in the control experiment with increasing concentrations of 1 but without enzyme (Fig.
2
C). Interestingly, from these experiments a slightly lower K value can be determined for
m
GlcNAc derivative 1 than for GlcNAc (0.2 ꢀM vs 1 ꢀM), suggesting that 1 may actually be a
better acceptor substrate than GlcNAc. To investigate the time dependency of the β-1,4-
GalTs reaction under these conditions, we next incubated the reaction with a single
concentration of GlcNAc (5 mM) or 1 (1 mM) for 5-40 min. A clear time-dependency of
turnover was observed with both acceptors (Fig. 2D). With increasing incubation time, UDP
formation increased in the presence of both GlcNAc and 1, with a maximal turnover after 30
min. Taken together, these results suggest strongly that, just like its parent compound
GlcNAc, GlcNAc derivative 1 is recognised as an acceptor substrate by bovine β-1,4-GalT.
In order to understand this discrepancy with previous literature reports [10,11], we carried out
additional experiments. In the main literature precedent, β-1,4-GalT from bovine milk was
obtained commercially, and the assay was carried out in the presence of 12.5 mg/mL BSA
[
11]. While BSA is frequently added to assay mixtures to generically reduce protein adhesion
to the reaction container, in the case of GTs, a direct influence of BSA on substrate
specificity, as well as the kinetic parameters of the enzymatic reaction, has also been
observed [17]. To assess the effect of BSA on the behaviour of GlcNAc derivative 1, we
repeated the substrate activity assay in the presence of BSA (12.5 mg/mL). Assay mixtures
were incubated for 20 min and absorbance was recorded at 620 nm. When the formation of
UDP was plotted over the concentration of acceptor (GlcNAc or 1), once again very similar
reaction profiles were observed for both GlcNAc and GlcNAc derivative 1 (Fig. 2B). The
absence of BSA from our initial experiments could therefore be ruled out as a potential
explanation for the observed discrepancy with the previous literature results. Interestingly,
the turnover rates for both GlcNAc and 1 were higher in the presence of BSA than in its
6

ACCEPTED MANUSCRIPT
absence (Fig. 2B). This effect of BSA is in keeping with observations for another
glycosyltransferase [17].
While our initial experiments were all carried out with a batch of bovine β-1,4-GalT
recombinantly expressed in our own laboratory, previously reported results for 1 were
obtained with a commercially available enzyme [11]. The information provided by the
supplier about this β-1,4-GalT is limited, and it was therefore not certain if it was identical to
our own recombinant batch. We therefore decided to repeat our experiments with a batch of
commercial β-1,4-GalT from the same supplier. First, we determined the activity of the
commercial β-1,4-GalT as 12.7 mU/mL (ESI). Then, we repeated the substrate assays with
either GlcNAc or 1 as acceptor, in the absence or presence of BSA. The results were
practically identical to those obtained previously with recombinantly expressed enzyme.
GlcNAc derivative 1 once again acted as a substrate towards β-1,4-GalT, both in the absence
and presence of BSA (Fig. S1). In the presence of BSA, turnover rates increased
approximately 2-fold, for both GlcNAc and 1. The K value of each substrate also increased,
m
by approximately 4-fold (GlcNAc: 0.4 ꢀM without BSA, 1.5 μM with BSA; 1: 0.08 ꢀM
without BSA, 0.3 μM with BSA). Most importantly, both trends (increased turnover, increased
K ) are practically identical for both GlcNAc and GlcNAc derivative 1. This provides further
m
evidence that both monosaccharides can serve, in an identical manner, as acceptor substrates
for β-1,4-GalT.
Our standard assay does not detect the primary product of the GalT reaction, in contrast to the
radiochemical GT assay used in the original reports of 1 as a β-1,4-GalT inhibitor [10,11]. In
order to directly detect the putative galactosylation product of 1, we therefore carried out
LC/MS experiments. 1 (500 µΜ), recombinant β-1,4-GalT (200 µL) and UDP-Gal (500 µΜ)
o
were incubated for 1h at 30 C in HEPES buffer (pH 7.0, 50 mM KCl). Analysis of the
7

ACCEPTED MANUSCRIPT
reaction by LC/MS showed the appearance of a new peak in the chromatogram, at a slightly
shorter retention time than the peak corresponding to 1 (Fig. S2). While the separation of the
two peaks is not complete, the mass spectra confirmed the formation of a species with a
molecular ion of 532, which corresponds to the sodium adduct of mono-galactosylated 1,
Gal-1. This result provides direct evidence that 1 can indeed act as an acceptor substrate for
β-1,4-GalT. The formation of Gal-1 was strongly dependent on the amount of β-1,4-GalT
present, and significant levels of the galactosylated species were only observed at the highest
enzyme concentration tested (Fig. S3, A-C). Interestingly, however, if phosphatase (20 µL)
was present in the reaction mixture, similar levels of Gal-1 were observed even at a lower β-
1
,4-GalT concentration (Fig. S3, D). The conditions used for this last experiment are identical
to the ones used in our standard assay protocol. These results therefore suggest that the
enzymatic profiles which we observed in our biochemical assays were indeed due to the
behaviour of 1 as an acceptor substrate of β-1,4-GalT.
Discussion
Building on the seminal work by Lemieux [18], Palcic has previously shown that GlcNAc
glycosides with a hydrophobic aglycon, such as 8-methoxycarbonyloctyl GlcNAc and 4-
methylumbelliferyl GlcNAc (Fig. 1), are highly effective acceptor substrates for β-1,4-GalT
[
19]. Intriguingly, the closely related GlcNAc derivative 1 has been reported in two separate
studies as an inhibitor, but not a substrate, of β-1,4-GalT [10,11]. In our hands, however, 1
did behave as an acceptor substrate towards this enzyme, with a similar profile as the natural
GlcNAc acceptor. This discrepancy is not due to a case of “mistaken identity” of 1.
Spectroscopic characterisation of our batch of 1 and careful comparison with literature data
8

ACCEPTED MANUSCRIPT
[
14] confirms that the compound used in this study is identical to the one that had been
reported previously.
Our results suggest that the behaviour of 1 towards β-1,4-GalT is strongly dependent on the
assay conditions. In the literature precedent, GlcNAc derivative 1 and its analogues were
tested in a radiochemical GT assay, which directly detects the galactosylated reaction product
[
11]. The enzyme-coupled assay used in the present study requires a phosphatase, in
conjunction with a colorimetric read-out, to measure the formation of UDP, the secondary
product of the β-1,4-GalT reaction [16]. We have eliminated potential false positives that
would specifically affect our assay protocol, e.g. from non-specific hydrolysis of UDP, by
including the requisite control experiments. We have also investigated the role of specific
assay components used in the original assay, such as BSA, and the source of the β-1,4-GalT.
We found that these parameters do not fundamentally alter the outcome of the reaction. These
results demonstrate that despite the different assay formats, the results from the
radiochemical assay used in the previous study, and from our own colorimetric assay are, in
principle, comparable.
Despite the seeming contradiction, our findings are not irreconcilable with the previous
results. We have observed substrate activity for 1 both in the presence and absence of
phosphatase. This suggests that while the presence of the phosphatase in our assay facilitates
the utilisation of 1 as a β-1,4-GalT acceptor substrate, it is not essential. The addition of
phosphatase or NDP-cycling enzymes is a well-established approach in chemo-enzymatic
applications of GTs, to drive the GT reaction to completion and to improve the efficiency of
the overall synthetic process [20]. It is likely that the presence of the phosphatase in our assay
affects the equilibrium between different species along the β-1,4-GalT kinetic pathway in a
similar fashion.
9

ACCEPTED MANUSCRIPT
We propose the following model, which is consistent with both our own data and the
previous findings (Fig. 3): because of its close structural similarity to the natural acceptor
GlcNAc, 1 can indeed be recognised as an acceptor substrate by β-1,4-GalT, which converts
it into the corresponding disaccharide Gal-1. This disaccharide may form a stable, non-
covalent complex with the enzyme, acting effectively as a feedback inhibitor. This would be
consistent with the previous finding that a hydrophobic aglycone such as 2-napthyl is crucial
for strong substrate binding in this series [11]. This inhibitory effect can be overcome either
by increasing the concentration of β-1,4-GalT, or by including phosphatase in the assay
mixture. The irreversible, phosphatase-catalysed hydrolysis of UDP will shift the overall
equilibrium towards the reaction products, thus revealing the latent substrate activity of 1
even at low β-1,4-GalT concentrations (Fig. 3). It has to be emphasised that at present, this
model is hypothetical. Its definitive test will require the synthesis and testing of the Gal-1
disaccharide. While disaccharide product inhibition of glycosyltransferases is perhaps less
common than inhibition by the NDP reaction product, there is precedent for the inhibition of
β-1,4-GalT by disaccharide acceptor analogues [21].
Conclusion
In conclusion, we have found that 2-naphthyl β-D-GlcNAc 1 can behave as an acceptor
substrate towards β-1,4-GalTs from bovine milk. This substrate activity has not previously
been described. Its discovery was facilitated by the presence of a phosphatase in our enzyme-
coupled GT assay, which affects the reaction equilibria along the β-1,4-GalT kinetic pathway.
The finding that 1 can behave as a β-1,4-GalT substrate has implications both for the design
and the application of acceptor-based GT inhibitors: in principle, 1 and related GT inhibitors
are attractive chemical tools to study glycosylation in cells. Our results suggest that such
1
0

ACCEPTED MANUSCRIPT
applications in cells may be complicated by the presence of phosphatases, which will very
likely modulate the inhibitor/substrate behaviour of such acceptor analogues. Following
conversion of 1 into the corresponding disaccharide, it may be this disaccharide that acts as
the actual inhibitory species through a feedback inhibition mechanism. In this case, the
substrate activity of 1 would be a prerequisite for its inhibitory activity. Given the
considerable number of acceptor-based GT inhibitors in the literature, these questions warrant
further investigation. The present study provides a starting point for such investigations.
1
1

ACCEPTED MANUSCRIPT
Experimental section
General. All reagents were obtained commercially and used as received unless stated
otherwise. GlcNAc derivative 1 was synthesized as previously reported and characterised by
1
13
H- and C-NMR spectroscopy and mass spectrometry [14]. Thin-layer chromatography
(
TLC) was performed on pre-coated aluminium plates (Silica Gel 60 F , Merck) and
2
54
compounds were visualised by exposure to UV light (254 and 280nm). Preparative
chromatography was carried out on silica gel 60 (pore size 60 Ǻ, 230-400 mesh, Sigma-
Aldrich) at normal pressure. NMR spectra were recorded at 298K on a Bruker Avance DRX
1
13
4
00 spectrometer (400 MHz for H, 100 MHz for C). Chemical shifts (
δ) are reported in
ppm (parts per million). Coupling constants (J) are reported in Hz.
2
-Naphthyl 3,4,6-tri-O-acetyl 2-acetamido-2-deoxy-β-D-glucopyranoside (3) [14]. To an
aqueous solution of sodium hydroxide (1M, 2 mL), a solution of 3,4,6-tri-O-acetyl-2-deoxy-
α-D-glucopyranosyl chloride 2 (400 mg, 1 equiv.), 2-naphthol (319 mg, 2 equiv.) and tetra-n-
butylammonium bromide (354 mg, 1 equiv.) in methylene chloride (2 mL) was added. The
resulting two-phase system was stirred for 2h at rt. The mixture was diluted with ethyl acetate
and the organic phase washed sequentially with an aqueous solution of 1M sodium hydroxide
and water, then dried. The organic extract was then filtered, and the filtrate concentrated to
yield a crude solid which was purified by chromatography (DCM/MeOH 20:1), to yield 389
1
mg of a white powder (75 %). H-NMR (400 MHz, CDCl ):
δ 7.81 (d, 2H, J = 9.0 Hz, H-4
3
and H-9 of naphthyl), 7.77 (d, 1H, J = 8.0 Hz, H-6 of naphthyl), 7.47 (m, 1H, H-7 of
naphthyl), 7.44 (d, 1H, J = 2.5 Hz, H-1 of naphthyl), 7.38 (m, 2H, H-8, H-5 of naphthyl),
7
.22 (dd, 1H, J = 2.5, 9.0 Hz, H-3 of naphthyl), 5.46 (d, 1H, J = 8.0 Hz, H-1), 5.39 (dd, 1H, J
10.5, 10.5 Hz, H-3), 5.11 (dd, 1H, J = 10.5, 10.5 Hz, H-4), 4.36 (dd, 1H, J = 5.5, 12.5 Hz,
H-6), 4.16 (M, 2H, H-2 and H-6), 4.11 (m, 1H, H-5), 2.07, 2.06, 2.04 (3 × s, 9H, acetyl), 1.96
=
1
2

ACCEPTED MANUSCRIPT
1
3
(
s, 3H, acetamido). C-NMR (100 MHz, CDCl ): δ 20.8, 20.9, 20.9, 23.3, 59.0, 61.9, 71.0,
3
7
4.6, 78.0, 82.3, 121.0, 126.1, 126.5, 127.9, 128.0, 130.3, 137.4, 128.5, 148.1, 156.9, 170.4,
71.4, 171.8, 172.5.
1
1
-(2-Naphthyl) 2-acetamido-2-deoxy-β-D-glucopyranoside (1) [14]. To a solution of 3
(200 mg) in methanol-toluene (1:1) was added a catalytic amount of 0.5M sodium methoxide
in methanol. The reaction mixture was stirred at room temperature for 0.5h and the progress
of the reaction was monitored by TLC (DCM/MeOH 4:1). Upon completion of the reaction,
the organic solution was concentrated, and the residue was purified by chromatography
1
(
DCM/MeOH 4:1), to yield 92 mg of a white powder (88%). H-NMR (400 MHz, MeOD):
δ
7.76 (m, 3H, H-4, H-9, H-6 of naphthyl), 7.42 (m, 2H, H-7, H-1 of naphthyl), 7.34 (m, 1H,
H-8 of naphthyl), 7.18 (dd, 1H, J = 2.4, 9.2 Hz, H-3 of naphthyl), 5.19 (d, 1H, J = 8.4 Hz, H-
1
5
9
7
), 4.00 (dd, 1H, J = 8.4, 10.4Hz, H-2), 3.98 (dd, 1H, J = 2.4, 12Hz, H-6), 3.77 (1H, dd, J =
.6, 12Hz, H-6), 3.63 (dd, 1H, J = 8.4, 10Hz, H-3), 3.58 (m, 1H, H-5), 3.48 (dd, 1H, J = 8.8,
1
3
.6Hz, H-4), 1.99 (s, 3H, NHCH ). C-NMR (100 MHz, MeOD): δ 23.0, 57.4, 62.6, 71.9,
3
5.9, 78.4, 101.0, 111.9, 119.8, 125.3, 127.4, 128.2, 128.6, 130.4, 131.3, 135.9, 156.9, 174.0.
2
+
m/z (ESI) 371.1199 [M+H+Na] , C H NNaO requires 371.1345.
18
22
6
Enzymology. The plasmid for β-1,4-GalT from bovine milk was a generous gift from Dr
Christelle Breton (Grenoble). Bovine β-1,4-GalT was either expressed in our own laboratory
as previously reported [7] or obtained commercially (Sigma). For the biochemical assays, we
used a recently reported colorimetric protocol [16]. All assays were carried out in Nunc clear,
flat-bottom 96-well polystyrene microplates. Assay wells typically contained MnCl , calf-
2
intestinal phosphatase (CIP), chicken egg-white lysozyme (CEL), UDP-Gal donor and either
GlcNAc or 1 as acceptor (for details of the assay protocol see ESI). To quantify the
1
3

ACCEPTED MANUSCRIPT
concentration of inorganic phosphate (P ), malachite green reagents were added, and the
i
absorbance was recorded at 620 nm on a BMG Labtech POLARstar Optima multiplate reader.
Data collection and analysis. A calibration curve (0-12.5 µM UDP, corresponding to 0-25
µM P ) was constructed for each microplate by linear regression. The calibration curve was
i
used to convert absorbance measurements at 620 nm in sample and control wells to [UDP]
(
µM). For each sample and control well, a corresponding background well (containing
identical components but no acceptor) was included, to account for non-specific hydrolysis of
donor. Corrected absorbance values for each well were obtained by subtracting the
corresponding background reading from the absorbance of the respective sample or control
well. The calculated concentration of UDP was plotted against concentration of acceptor (for
substrate assay or control assay) or incubation time (for time-dependent assays). Averages
and standard deviations were calculated in Microsoft Excel.
1
4

ACCEPTED MANUSCRIPT
LC/MS experiments.
Standard assay mixtures for investigating the acceptor properties of compound 1 towards β-
,4-GalT contained UDP-Gal (500 µΜ), compound 1 (500 µΜ), β-1,4-GalT (20, 50 or 200
1
µL, Fig. S3), calf intestinal phosphatase (20 µL, assay D only, Fig. S3), 13 mM HEPES
o
buffer (pH 7.0, 50 mM KCl). Mixtures were incubated for 1h at 30 C in a water bath.
Reactions were stopped by the addition of the same volume of methanol. The mixtures were
centrifuged for 15 min at 1000 rpm. The supernatants were used for LC/MS analysis directly.
LC/MS analysis was carried out on an HPLC reverse phase column (Agilent Eclipse XDB-
C8 4.6×150 mm) with water (0.1 % formic acid) against methanol as the mobile phase. The
gradient is shown in Table 1. The HPLC was coupled to an Advion Compact Mass
Spectrometer (CMS) for mass detection.
1
5

ACCEPTED MANUSCRIPT
Acknowledgment
We thank King’s College London for a PGR studentship (to JJ), the EPSRC National Mass
Spectrometry Facility, Swansea, for the recording of mass spectra, and Clive Page for helpful
discussions. Additional funding from the BBSRC IBCarb Network (grant IBCarb-PoC-0616-
0
40) is gratefully acknowledged. The plasmid for bovine β-1,4-GalT was a generous gift
from Dr Christelle Breton (Grenoble).
1
6

ACCEPTED MANUSCRIPT
References
[
[
[
[
[
[
1]
2]
3]
4]
5]
6]
Hennet T. CMLS, Cell Mol Life Sc 2002; 59: 1081-95.
Holgersson J, Lofling J, Glycobiology 2006; 16: 584-93.
Sperandio M, FEBS J. 2006; 273: 4377-4389.
Schutzbach J, Brockhausen I, Methods Mol Biol 2009; 534: 35-73.
Ramakrishnan B, Shah PS, Qasba PK, J Biol Chem 2001; 276: 37665-71.
McDonald AG, Hayes JM, Bezak T, Gluchowska SA, Cosgrave EFJ, Struwe WB,
Stroop CJM, Kok H, van de Laar T, Rudd PM, Tipton KF, Davey GP, J. Cell Sci. 2014; 127:
014-5026.
5
[
7]
Pesnot T, Jorgensen R, Palcic MM, Wagner GK, Nat Chem Biol 2010; 6: 321-23.
[
8]
Takaya K, Nagahori N, Kurogochi M, Furuike T, Miura N, Monde K, Lee YC,
Nishimura SI, J Med Chem 2005; 48: 6054-65.
[
[
[
9]
Wang S, Vidal S, Carbohydr Chem 2013; 39: 78-101.
10] Chung SJ, Takayama S, Wong CH, Bioorg Med Chem Lett 1998; 8: 3359-3364.
11] Brockhausen I, Benn M, Bhat S, Marone S, Riley JG, Montoya-Peleaz P, Vlahakis JZ,
Paulsen H, Schutzbach JS, Szarek WA, Glycoconj J 2006; 23: 525-41.
[
12] Siegbahn A, Manner S, Persson A, Tykesson E, Holmqvist K, Ochocinska A,
Ronnols J, Sundin A, Mani K, Westergren-Thorsson G, Widmalm G, Ellervik U, Chem. Sci.
014; 5: 3501-3508.
2
[
[
13] Gao Y, Lazar C, Szarek WA, Brockhausen I, Glycoconj. J. 2010; 27: 673-684.
14] Sasaki K, Nishida Y, Kambara M, Uzawa H, Takahashi T, Suzuki T, Suzuki Y,
Kobayashi K. Bioorg Med Chem 2004; 12: 1367-75.
[
[
15] Wu ZL, Ethen CM, Prather B, Machacek M, Jiang W, Glycobiology 2011; 21: 727-33.
16] Tedaldi L, Evitt A, Goos N, Jiang JJ, Wagner GK, MedChemComm 2014; 5: 1193-
2
01.
[
17] (a) Manevski N, Morelo PS, Yli-Kauhaluoma J, Finel M, Drug Metab. Dispos. 2011;
3
9: 2117-2129; (b) Manevski N, Yli-Kauhaluoma J, Finel M, Drug Metab. Dispos. 2012; 40:
2
192-2203.
[
[
[
[
18] Lemieux RU, Bundle DR, Baker DA, J. Am. Chem. Soc.; 1975, 97: 4076-4083.
19] Palcic MM, Heerze LD, Pierce M, Hindsgaul O, Glycoconj. J.; 1988, 5: 49-63.
20] Gijsen HJM, Qiao L, Fitz W, Wong C-H, Chem. Rev.; 1996, 96: 443-473.
21] Brown JR, Yang F, Sinha A, Ramakrishnan B, Tor Y, Qasba PK, Esko JD, J Biol
Chem 2009; 284: 4952-4959.
1
7

ACCEPTED MANUSCRIPT
Fig. 1 GalT acceptor analogues discussed in the text that have been reported as inhibitors (A)
or substrates (B).
1
8

ACCEPTED MANUSCRIPT
a
Fig. 2 Acceptor substrate assays with GlcNAc or 1, and recombinant β-1,4-GalT.
A β-1,4-GalT-catalysed conversion of UDP-Gal donor into UDP at different concentrations
of GlcNAc (left) or 1 (right) as the acceptor, in the absence of BSA.
1
0
5
0
10
5
0
0
2
4
6
0.0
0.5
1.0
Substrate [µM]
Substrate [µM]
B β-1,4-GalT-catalysed conversion of UDP-Gal donor into UDP at different concentrations
of GlcNAc (left) or 1 (right) as the acceptor, in the presence of BSA.
1
0
5
0
10
5
0
0
2
4
6
0.0
0.5
1.0
Substrate [µM]
Substrate [µM]
C Control experiment without β-1,4-GalT.
D Time course.
1
0
5
0
0.0
0.5
1.0
Substrate [µM]
a
Conditions: All experiments were carried out in triplicate. Bars indicate mean values ± S.D.
A β-1,4-GalT, GlcNAc (0-5 mM) or 1 (0-1 mM), UDP-Gal donor (28 µM), MnCl (5 mM),
2
chicken egg-white lysozyme (1 mg/mL), calf-intestinal phosphatase (10 U/mL), DMSO (10%)
o
and buffer (13 mM HEPES, pH 7.0, 50 mM KCl) were incubated in a 96-well plate at 30 C
with shaking for 20 min. The reaction was stopped by the addition of malachite reagents, and
the absorbance was recorded at 620 nm after 30 min. B Conditions as in A but with bovine
1
9

ACCEPTED MANUSCRIPT
serum albumin (1.25 mg/mL). C Conditions as in A with compound 1 (0-1 mM) as the
substrate, but without β-1,4-GalT. D Conditions as in A, with GlcNAc (5 mM) or 1 (1 mM)
as the acceptor, and with incubation times from 5-40 min.
2
0

ACCEPTED MANUSCRIPT
Fig. 3 Hypothetical model for the dual activity of 1 as either acceptor substrate or inhibitor
*
along the β-1,4-GalT kinetic pathway (D: donor sugar-nucleotide, E: enzyme, A: acceptor, A :
*
galactosylated acceptor, D : nucleoside diphosphate).
stable complex of primary
products and GalT:
inhibitory activity
predominates
A
D+E
D E
A D E
A* D* E
irreversible formation of
secondary products:
A*+D*+E
phosphatase
substrate activity
predominates
uridine + Pi
2
1

ACCEPTED MANUSCRIPT
a
Scheme 1
Synthesis of 2-naphthyl β-D-GlcNAc 1 [14].
a
Reagents and conditions: (i) tetrabutylammonium bromide, 1N NaOH, DCM, rt, 2h; (b)
NaOMe, MeOH/toluene (1:1), rt, 0.5h.
2
2

ACCEPTED MANUSCRIPT
Highlights
(1) GlcNAc β1-(2-naphthyl) can behave as an acceptor substrate of β-1,4-
galactosyltransferase from bovine milk
(2) This acceptor substrate activity is unmasked by the presence of a phosphatase in the
assay mixture
(
3) A kinetic model is proposed which reconciles the substrate activity of GlcNAc β1-
2-naphthyl) with its previously reported behaviour as a galactosyltransferase
inhibitor
(