α1,4-Galactosyltransferase-catalyzed glycosylation of sugar and lipid modified Leu-enkephalins

Article abstract of DOI:10.1016/j.molcatb.2013.08.018

Glycosylation of therapeutic peptides has been reported to improve delivery and targeting of various vaccines and drugs to specific cells/tissues. However, chemical synthesis of complex oligosaccharide derivatives via conventional methods can be challengi

Full text of DOI:10.1016/j.molcatb.2013.08.018

Journal of Molecular Catalysis B: Enzymatic 97 (2013) 196–202
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
␣-1,4-Galactosyltransferase-catalyzed glycosylation of sugar and lipid
modified Leu-enkephalins
Pavla Simerskaa,∗, Michelle P. Christiea, Daryn Goodwina, Freda E.-C. Jenb,
Michael P. Jenningsb, Istvan Totha,c
a The University of Queensland, School of Chemistry and Molecular Biosciences, St. Lucia 4072, Queensland, Australia
b Institute for Glycomics, Griffith University, Parklands Drive, Southport, Queensland 4215, Australia
c The University of Queensland, School of Pharmacy, St. Lucia 4072, Queensland, Australia
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 7 June 2013
Received in revised form 23 August 2013
Accepted 26 August 2013
Available online 3 September 2013
Glycosylation of therapeutic peptides has been reported to improve delivery and targeting of various
vaccines and drugs to specific cells/tissues. However, chemical synthesis of complex oligosaccha-
ride derivatives via conventional methods can be challenging due to the need for several orthogonal
hydroxyl group protections. Liposaccharyl galactosyltransferase C, a naturally occurring glycosyltrans-
ferase enzyme from Neisseria meningitidis, was found to have the ability to transfer a galactosyl moiety
to glyco(lipo)peptides. An enzymatic glycosylation of Leu-enkephalin glyco(lipo)peptides was developed
and optimized in this study in order to prepare pain regulating peptides with potentially improved central
nervous system delivery.
Keywords:
Enzymatic synthesis
Glycopeptide
Leu-enkephalin
Galactosyltransferase
Glycosylation
© 2013 Elsevier B.V. All rights reserved.
1. Introduction
water solubility, can be a limiting factor in drug delivery. Lipoamino
acids in combination with carbohydrates have not only improved
Opioid receptors on neuronal cell membranes in the brain and
spinal cord have ligand specificity for endogenous opioid peptides.
These peptides have been shown to possess potent pain-reducing
effects and have the potential to be used as pain regulating thera-
peutics [1,2]. Leu-enkephalin (Enk) was chosen as a model central
nervous system (CNS)-targeted peptide because of its high bind-
ing affinity for ␮- and ␦-type opioid receptors and analgesic effects
in vivo.
Peptides have been used as therapeutics for decades, however
their delivery in vivo remains challenging predominantly because
of poor enzymatic stability and limited bioavailability. Various
strategies involving peptide production, chemical modification and
formulation have been attempted to overcome these issues [3–7].
Glycosylation and lipidation have been studied to increase the
enzymatic stability, membrane permeability, activity and/or affin-
ity of Enk to opioid receptors [8–10].
stability and uptake, but also enhanced water solubility and the
targeting of drug candidates to specific sites [13,14]. Carbohydrates
have useful structural properties for which they have been utilized
as components of peptide drugs [4,6,15–17] and vaccines [18–23].
We describe the chemo-enzymatic glycosylation of Enk with
mono-, di- and trisaccharide sugar moieties terminating with a
galactose residue. It is expected that these modifications would not
only improve biological transport, stability and retention time in
the body but also enable targeting to the asialoglycoprotein recep-
tor. Asialoglycoprotein receptors are not only highly distributed
in hepatic cells but are also expressed in discrete areas of the
brain [24], and have specificity for terminal galactose [25,26]. An
enzymatic strategy [27–29] can circumvent the challenges asso-
ciated with the manipulation of protecting groups on multiple
hydroxyl groups during the chemical synthesis of oligosaccharides
[30,31]. Here we utilized a galactosyltransferase, lipopolysaccharyl
␣-1,4-galactosyltransferase (EC 2.4.1.; LgtC) derived from Neisseria
meningitidis [32,33]. The natural function of LgtC is to transfer an
␣-d-galactosyl moiety from uridine-5ꢀ-diphosphate-␣-d-galactose
(UDP-Gal) donor to the terminal lactose of lipooligosaccharides,
while retaining the anomeric configuration of the donor’s glycosidic
bond, and forming an ␣-1,4 linkage [34,35]. We expressed LgtC in
Escherichia coli and employed it to selectively attach a galactose unit
onto chemically synthesized (lipo)saccharide peptide scaffolds. The
Lipidation was also found to improve transport across biolog-
ical barriers [11] and lipoamino acids [12] have previously been
used for peptide drug delivery [1,8]. While many advantages have
been described, the drawback of the increased lipophilicity, poor
∗
Corresponding author. Tel.: +61 7 3365 4636; fax: +61 7 3365 4273.
E-mail address: p.simerska@uq.edu.au (P. Simerska).
1381-1177/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.molcatb.2013.08.018

P. Simerska et al. / Journal of Molecular Catalysis B: Enzymatic 97 (2013) 196–202
197
combination of chemical and enzymatic synthesis allowed us to
produce compounds, which may permit targeted delivery to the
CNS.
and liquid chromatography mass spectrometry (LC–MS) analyses
were performed on a Perkin-Elmer-Sciex API 3000 using Analyst
1.4 software (Applied Biosystems/MDS Sciex, Toronto, Canada).
High resolution mass spectrometry (HR-MS) was performed using
ABSCIEX 5600 Triple TOF, positive ion mode and approximately
30,000–35,000 mass resolution, 10 ␮L injections. Samples from
the enzymatic reactions were filtered through Vydac® BioSelect
SPE Columns (C4, 1 mL, 214SPE1000 and C18, 1 mL, 218SPE1000)
prior to ES-MS. Columns used for LC–MS were either Vydac® C8
208TP5205 (5 ␮m, 2.1 mm i.d. × 50 mm) or Phenomenex® Luna®
C18 (5 ␮m, 2 mm i.d. × 50 mm). Solvent A used for MS consisted
of 100% Millipore H2O and 0.1% AcOH, while solvent B consisted
of 90% MeOH, 10% purified Millipore H2O, and 0.1% AcOH. Nuclear
magnetic resonance (NMR) spectroscopy was recorded on a Bruker
AM 300 MHz instrument with the chemical shifts given in parts per
million with reference to tetramethylsilane as an internal standard.
2. Experimental
2.1. Materials
Unless otherwise stated, all chemicals used in this project
were of analytical grade or equivalent. N,N-Dimethylformamide
(DMF) was obtained from Emanuel Merck, Darmstadt (EMD, Darm-
stadt, Germany), while O-benzotriazole-N,N,Nꢀ,Nꢀ-tetramethyl-
uronium-hexafluoro-phosphate (HBTU) and Fmoc-l-amino acids
were obtained from Mimotopes (Clayton, VIC, Australia). N,N-
Diisopropylethylamine (DIPEA), dichlormethane (DCM), HCl,
trifluoroacetic acid (TFA), and diethyl ether were supplied by
Merck (Kilsyth, VIC, Australia). Piperidine and triisopropylsaline
(TIPS) scavenger were purchased from Sigma–Aldrich (Castle Hill,
NSW, Australia). Toluene was supplied from Lab-Scan Pty. Ltd.
(Dublin, Ireland). For carbohydrate synthesis, Ac2O was supplied
by Univar (Ingleburn, NSW, Australia), and dimethaminopyri-
dine (DMAP), tetrahydrofuran (THF) and triethylamine (TEA) from
Merck (Kilsyth, VIC, Australia). Succinic anhydride and LiClO4
were purchased from Sigma–Aldrich (Castle Hill, NSW, Australia).
High performance liquid chromatography (HPLC) and mass spec-
trometry (MS) grade acetonitrile (ACN) and MeOH were supplied
by Scharlau (Port Adelaide, SA, Australia), while acetic acid
(AcOH) was supplied from Merck (Kilsyth, VIC, Australia). Thin
layer chromatography (TLC) grade ammonia solution 25% was
purchased from Sigma–Aldrich (Castle Hill, NSW, Australia), iso-
propanol from Lab-Scan Pty. Ltd. (Dublin, Ireland), and CHCl3 from
Merck (Kilsyth, VIC, Australia). For enzymatic experiments, N-
2-hydroxyethylpiperazine-Nꢀ-2-ethanesulfonic acid (HEPES) was
obtained from Grand Island Biological Company (GIBCO, Mul-
grave VIC, Australia), and UDP-Gal disodium salt from CalBioChem
(Darmstadt, Germany).
2.3. Chemical synthesis of substrates
2,3,4,6-Tetra-O-acetyl-␤-d-galactopyranosyl azide (1a) and 2,3,
6-tri-O-acetyl-4-O-(2ꢀ,3ꢀ,4ꢀ,6ꢀ-tetra-O-acetyl-␤-d-galactopyranos-
yl)-␤-d-glucopyranosyl azide (1b) were obtained after the reac-
tion of the bromides with sodium azide in acetone in 67% and
80% yields, respectively (Supporting Information) [36–38]. The
galactose and lactose azides (1a-b) were reduced to amines
through hydrogenation and immediately treated with succinic
anhydride in dry THF (galactose derivative) or pyridine (lactose
derivative) with catalytic amounts of DMAP to form N1-(2,3,4,6-
tetra-O-acetyl-␤-d-galactopyranosyl)succinamic acid 2a (67%
yield) and N1-(2,3,6-tri-O-acetyl-4-O-(2ꢀ,3ꢀ,4ꢀ,6ꢀ-tetra-O-acetyl-␤-
d-galactopyranosyl)-␤-d-glucopyranosyl)succinamic acid 2b (46%
yield) [39,40].
To synthesize the lipid moiety, 1-bromotetradecane and
diethyl acetamidomalonate were reacted under a reflux with HCl
to produce 2-amino-d,l-hexadecanoic acid hydrochloride in 54%
yield. A stirred solution of 5,5-dimethyl-1,3-cyclohexanedione,
TEA and Ac2O yielded (72%) 2-(1-hydroxyethylidene)-5,5-
dimethylcyclohexane-1,3-dione, which was further conjugated to
2-amino-d,l-hexadecanoic acid hydrochloride [41]. The 2-((1-(4,4-
dimethyl-2,6-dioxocyclohexylidene)ethyl)amino)hexadecanoic
acid (Dde-C16), produced with 55% yield, was used as a lipid
building block in the synthesis of glycolipopeptides 3b and 3d.
2.2. Methods
TLC was carried out on Kieselgel 60F254 silica gel coated alu-
minum plates from Merck (Darmstadt, Germany). All TLCs were
developed using eluent CHCl3:MeOH:H2O, 7:5:1, v/v and visu-
alized by 20% H2SO4:EtOH solution or an anisaldehyde reagent
followed by heating, unless stated otherwise. THF was dried over
sodium and benzophenone; DCM over calcium hydride; and CHCl3
over calcium chloride. Analytical reversed phase HPLC (RP-HPLC)
was performed using Shimadzu (Kyoto, Japan) instrumentation
(LabSolutions software, SIL-20AC HT autosampler, LC-20AB pump,
SPD-M10A detector, DGU-20A5 degasser). Analysis was achieved
using a linear 0–100%, 20–35% and/or 30–80% gradient of solvent B
(solvent A: 0.1% TFA in H2O; solvent B: 90% ACN:H2O:0.1% TFA) for
30 min with a 1 mL/min flow rate and detection at 214 nm, unless
stated otherwise.
2.4. Chemical glycosylation and peptide/lipopeptide solid phase
synthesis
All peptides, lipopeptides and glycolipopeptides were synthe-
sized using standard in situ neutralization stepwise solid-phase
synthesis protocol for Fmoc chemistry [42]. Peptides and lipopep-
tides were synthesized using Rink Amide-4-methylbenzhydryl
amine (Rink Amide-MBHA) resin (100–200 mesh, 0.59 mmol/g;
Peptides International, USA). The resin was swollen in 10%
DIPEA/DMF for a minimum of 2 h. Fmoc-l-amino acids (4 eq) were
activated with 4 eq 0.5 M HBTU in DMF, and 5 eq DIPEA, then
reacted with amino groups on Rink Amide-MBHA resin (2× 1 h).
The Fmoc protecting group was removed using 20% piperidine/DMF
(2× 15 min). The Dde protecting group was removed by treating the
resin with 5% hydrazine/DMF (2× 15 min). After each manipulation,
the resin was washed with DMF.
Carbohydrate couplings to the N-terminus of the peptide or
lipopeptide were carried out overnight using 2 eq of glycosyl
(lactose or galactose) succinates activated with 1.9 eq of HBTU
in DMF and 2.5 eq equivalents of DIPEA. The glycopeptides were
de-acetylated using 75% (v/v) hydrazine hydrate in methanol
(2× 20 min). The resin was washed consecutively with DMF,
DCM, and MeOH and left to dry under vacuum overnight. The
Analytical separations were achieved on Grace Vydac®
(Columbia, Maryland, USA) columns (10 ␮m, 4.6 mm i.d. × 250 mm;
Hesperia, CA) – either C8 (208TP104) or C18 (218TP104); or
Alltima® C18 (5 ␮m, 4.6 mm i.d. × 250 mm) depending on the
hydrophobicity of the compounds. Preparative RP-HPLC was per-
formed using a 30–80% gradient of solvent B for 60 min on a
Waters Delta 600 system (Milford, Massachusetts, USA) with a
10–20 mL/min flow rate, and detection at 230 nm using PicoLog
software. Separations were achieved from a Vydac® C8 or Alltima®
C18 preparative column (10 ␮m, 22 mm i.d. × 250 mm). Frac-
tions containing pure compound were pooled and lyophilized
overnight. Electrospray ionization mass spectrometry (ES-MS)

198
P. Simerska et al. / Journal of Molecular Catalysis B: Enzymatic 97 (2013) 196–202
crude glycopeptides were cleaved from the resin using a 3 h
incubation with TFA:TIPS:H2O (95:2.5:2.5) and precipitation
from cold diethyl ether. The precipitated compounds were dis-
solved in ACN:H2O:TFA (for glycolipopeptides; 90:10:0.1) or
H2O:TFA (for glycopeptides; 100:0.1) and lyophilized overnight.
All crude peptides were purified using RP-HPLC as described
above.
2.4.1. N1-Leu-enkephalin-N4-(ˇ-d-galactopyranosyl)
succinamide (GalEnk, 3a)
Purified yield: 70%. ES-MS (C38H53N7O13, 815.4) m/z: 816.4
[M+H]+ (calcd. 816.4); HR-MS (C38H53N7O13
, 815.3701) m/z:
816.3774 [M+H]+ (calcd. 816.3779); RP-HPLC (C18 Vydac®, 20–35%
B) 12.1 min; (C18 Alltima®, 0–70% B) tR = 17.9 min.
2.4.2. N1-Leu-enkephalin-C16-N4-(ˇ-d-galactopyranosyl)
succinamide (GalEnkC16, 3b)
Purified yield: 16%. ES-MS (C54H84N8O14, 1068.6) m/z: 1070.1
[M+H]+ (calcd. 1069.6); HR-MS (C54H84N8O14, 1068.6107) m/z:
1069.6124 [M+H]+ (calcd. 1069.6185); RP-HPLC (C8 Vydac®,
60–100% B) tR = 14.3 and 15.5 min (diastereomers).
Fig. 1. SDS-PAGE gel analysis of the purified LgtC protein stained with Coomassie
Blue. Lane 1: pre-stained protein marker (New England Biolabs); lane 2: 1 ␮g of
purified LgtC protein.
concentration of purified LgtC yield was estimated by BCA (Bicin-
choninic acid) protein assay kit (Pierce). The LgtC protein (1 ␮g)
was analyzed by sodium dodecyl sulfate polyacrylamide gel elec-
trophoresis (SDS-PAGE) using a 4–12% NuPAGE® Novex Bis-Tris
Gels (Invitrogen) stained with Coomassie Blue to confirm its homo-
geneity and molecular mass of 35.7 kDa (Fig. 1). To maintain the
enzyme activity, glycerol (50%, v/v) was added to the LgtC enzyme
mix (2 mg/mL in 20 mM Na2HPO4, pH 7.5 buffer), which was then
stored at −20 ◦C.
2.4.3. N1-Leu-enkephalin-N4-(ˇ-d-galactopyranosyl-(1 → 4)-ˇ-
d-glucopyranosyl)succinamide (LacEnk, 3c)
Purified yield: 34%. ES-MS (C44H63N7O18, 977.4) m/z: 978.5
[M+H]+ (calcd. 978.4); HR-MS (C44H63N7O18
, 977.4230) m/z:
978.4284 [M+H]+ (calcd. 978.4308); RP-HPLC (C18 Vydac®, 20–35%
B) 11.5 min; (C18 Alltima®, 0–70% B) tR = 17.5 min.
2.4.4. N1-Leu-enkephalin-C16-N4-(ˇ-d-galactopyranosyl-
(1 → 4)-ˇ-d-glucopyranosyl)succinamide (LacEnkC16, 3d)
Purified yield: 49%. ES-MS (C60H94N8O19, 1230.7) m/z: 1231.8
[M+H]+ (calcd. 1231.7), 1254.0 [M+Na]+ (calcd. 1253.7); HR-
MS (C60H94N8O19, 1230.6635) m/z: 1231.6675 [M+H]+ (calcd.
1231.6713); RP-HPLC (C8 Vydac®, 60–100% B) tR = 13.7 and
14.8 min (diastereomers).
2.6. Enzymatic glycosylation of glycopeptides and
glycolipopetides
2.6.1. Enzyme activity with lactose as acceptor
Donor UDP-Gal (20 ␮L, 12 mM) and acceptor ␤-d-galactopyra-
nosyl-(1 → 4)-d-glucose (20 ␮L, 10 mM) in 50 mM HEPES buffer,
pH 7.0 containing 10 mM MnCl2, were reacted under catalysis
with LgtC enzyme mixture (10 ␮L, 1 mg/mL, 50% glycerol). The
reaction mixture was incubated at 37 ◦C overnight. The product
(␣-d-Gal-(1 → 4)-␤-d-Gal-(1 → 4)-d-Glc) formation was observed
by TLC (isopropanol:NH4OH:H2O 7:3:2, v/v): RF (GalLac) = 0.16; RF
(Lac) = 0.25 and confirmed by ES-MS (C18H32O11, 424.2) m/z: 425
[M+H]+ (calcd. 425.2).
2.5. Enzyme expression and purification
Recombinant LgtC was expressed and purified according to
Johnstone et al. [27] E. coli AD202 harboring pCWLgtC-25 and con-
taining a His-tag [32,43] was grown to an A600 of 0.4 in 2 L of
shaken 2xYT broth containing 100 ␮g/mL of ampicillin at 37 ◦C.
LgtC expression was induced by addition of 0.5 mM isopropyl 1-
thio-␤-d-galactospyranoside and growth of culture was continued
for 4 h at 37 ◦C. Cells were harvested by centrifugation and re-
suspended in 10 mL buffer A (20 mM Na2HPO4, 0.5 M NaCl, 10 mM
imidazole pH 7.5). Complete EDTA-free Protease Inhibitor Cocktail
(Roche) was added according to product instructions. Cells were
then lysed by 5 passages through a French press at 1000 psi as indi-
cated by gauge pressure. The lysate was centrifuged at 18,000 × g
for 10 min at 4 ◦C. For histidine-tagged protein purification, the
supernatant was collected and applied to an immobilized metal
affinity column containing TALONTM Metal Affinity Resin (BD Bio-
sciences) [27]. After the sample passed through the column, the
column was washed with 10 column volumes of buffer A and 5 col-
umn volumes of buffer A with 50 mM imidazole added. The protein
was eluted by collecting 1 mL fractions over 10 column volumes
of buffer A with 150 mM imidazole. Fractions were analyzed by
4–12% SDS-PAGE (Invitrogen). Fractions containing purified LgtC
were collected and pooled. From 8 g of cells, 8.4 mg of purified LgtC
protein was dialyzed in 10 kDa molecular mass cut-off SnakeskinTM
(Pierce) dialysis tubing against 5 L of 20 mM Na2HPO4, pH 7.5 at 4 ◦C
overnight. The dialyzed sample was concentrated using a 10 kDa
molecular mass cut-off Macrosep® Device (Pall Corporation). The
2.6.2. Analytical enzymatic reactions and optimizations
Donor UDP-Gal (20 ␮L, 6 mM) and GalEnk or LacEnk (20 ␮L,
2 mM) in 50 mM HEPES buffer, pH 7.0 containing 10 mM MnCl2,
were reacted under catalysis with LgtC enzyme mixture (0.5 ␮L,
2 mg/mL, 50% glycerol). For the reactions with GalEnkC16 and
LacEnkC16, 1 ␮L of dimethyl sulfoxide (DMSO) was added to the
reaction mixture. The reactions were incubated at 37 ◦C overnight.
The reaction mixtures were filtered (SPE Vydac® columns: C18 for
glycopeptides or C4 for lipidated glycopeptides) to remove salt
prior to monitoring with RP-HPLC and/or MS (0, 2, 4, 6 and 24 h).
Control reactions were set up as described above replacing LgtC
enzyme with HEPES/MnCl2 buffer.
2.6.3. Semi-preparative enzymatic reactions
UDP-Gal and an acceptor (Table 1) were dissolved in 0.4 mL of
50 mM HEPES buffer, pH 7.0, 10 mM MnCl2. The reaction mixture
used for the synthesis of the lipidated substrates 4b and 4d con-
tained 5% DMSO. LgtC enzyme (2 mg/mL, 50% glycerol) was added
and the reaction mixture was incubated at 37 ◦C. Further UDP-Gal
and LgtC were added to the reaction mixture portion-wise (Table 1).
The reaction mixture was lyophilized and re-dissolved in 120 ␮L of

P. Simerska et al. / Journal of Molecular Catalysis B: Enzymatic 97 (2013) 196–202
199
Table 1
The reaction conditions for the enzymatic synthesis of 4a-d.
Product
UDP-Gal
Acceptor
LgtC (␮L)
Reaction
(mg/␮mol)
(mg/␮mol)
time (days)
4a
4b
4c
4d
3.7/6.0 + 4.3/7.0
3.0/5.0 + 1.2/2.0
6.8/11.0
3a: 1.6/2.0
3b: 1.3/1.2
3c: 3.5/3.6
3d: 2.0/1.5
40 + 50
50 + 90
75
4
7
1
7
2.0/3.0 + 3.0/5.0
25 + 200
solvent A (4a and 4c) or in 40% DMSO in solvent B (4b and 4d)
and purified by RP-HPLC using C18 Alltima® column and 10–35%
gradient of solvent B (4a and 4c) or C8 Vydac® column and 0–70%
gradient of solvent B (4b and 4d).
2.6.3.1. N1-Leu-enkephalin-N4-(˛-d-galactopyranosyl-(1 → 4)-ˇ-d-
galactopyranosyl)succinamide (GalGalEnk) (4a). Purified yield: 45%
(0.9 mg, 0.9 ␮mol); ES-MS (C44H63N7O18, 977.4) m/z: 978.6 [M+H]+
(calcd. 978.4); HR-MS (C44H63N7O18, 977.4230) m/z: 978.4283
[M+H]+ (calcd. 978.4308); RP-HPLC (C18 Alltima®, 0–70% B)
tR = 17.6 min.
Fig. 2. RP-HPLC and ES-MS of the LacEnkC16 (3d) acceptor using C8 column and
60–100% solvent B.
(overnight instead of 2× 1 h) were used to achieve higher yields
with less starting material (data not shown). After de-acetylation
with hydrazine hydrate in MeOH, the crude products were cleaved
from the resin, precipitated from cold diethyl ether, dissolved in
ACN:H2O:TFA and lyophilized overnight.
Following the RP-HPLC purification with gradient of either
0–70% of solvent B using C18 Alltima® column (for glycopeptides
3a and 3c), or 60–100% gradient of solvent B and C8 Vydac® col-
umn (for glycolipopeptides 3b and 3d), GalEnk (3a) was obtained in
70% yield, GalEnkC16 (3b) in 16%, LacEnk (3c) in 34% and LacEnkC16
(3d) in 49% yields. The lipid-containing glycolipopeptides, 3b and
3d, were observed as two peaks on RP-HPLC, as they were diaste-
reomers synthesized from a racemic mixture of d- and l-lipoamino
acid (Fig. 2). Structure elucidation and purity of all products was
confirmed by ES/HR-MS and RP-HPLC, commonly used methods
for peptide containing products [8,9].
2.6.3.2. N1-Leu-enkephalin-C16-N4-(˛-d-galactopyranosyl-(1 → 4)-
ˇ-d-galactopyranosyl)succinamide
(GalGalEnkC16)(4b). Purified
yield: 13% (0.2 mg, 0.16 ␮mol); ES-MS (C60H94N8O19, 1230.6) m/z:
1231.8 [M+H]+ (calcd. 1231.6). RP-HPLC (C8, Vydac® 60–100% B)
tR = 15.2 min (further galactosylated products tR = 11.8, 13.0, 14.0).
LC–MS (C8, Vydac®, 0–100% B), 7.6 min, GalGalEnkC16 1232 m/z:
5.23 min and GalEnkC16 1069 m/z: 5.31 min; UDP-Gal 610 m/z:
1.10 min.
2.6.3.3. N1-Leu-enkephalin-N4-(˛-d-galactopyranosyl-(1 → 4)-
ˇ-d-galactopyranosyl-(1 → 4)-ˇ-d-glucopyranosyl)succinamide
(GalLacEnk) (4c). Purified yield: 71% (2.9 mg, 2.6 ␮mol); ES-MS
(C50H73N7O23, 1139.5) m/z: 1140.7 [M+H]+ (calcd. 1140.5); HR-
MS (C50H73N7O23, 1139.4758) m/z: 1140.4691 [M+H]+ (calcd.
1140.4836); RP-HPLC (C18 Alltima®, 0–70% B) tR = 17.2 min.
3.2. Enzymatic synthesis of Leu-enkephalin glycolipopeptides
2.6.3.4. N1-Leu-enkephalin-C16-N4-(˛-d-galactopyranosyl-(1 → 4)-
ˇ-d-galactopyranosyl-(1 → 4)-ˇ-d-glucopyranosyl)succinamide
(GalLacEnkC16) (4d). Purified yield: 20% (0.4 mg, 0.3 ␮mol); ES-MS
The application of N. meningitidis LgtC ␣-1,4-galactosyl-
transferase [32,45] was previously reported by Antoine et al. and
Zhang et al. [35,46]. In our study, we used LgtC galactosyltrans-
ferase to add a galactose unit from UDP-Gal to glyco(lipo)peptides,
3a-3d, to investigate LgtC specificity and reactivity toward these
substrates.
The LgtC enzyme is known to transfer a galactosyl unit to the
d-lactosyl glycolipid acceptor via an ␣-1,4 linkage [35]. Therefore,
d-lactose was chosen as a first acceptor to test the enzyme activity
in HEPES buffer. Binding interaction with metal ions (e.g. Mn2+) has
been reported as vital for the stability of the enzyme fold and over-
all LgtC activity [32,47] thus, 10 mM MnCl2 was added to all LgtC
catalyzed reactions. The formation of globotriose (␣-d-Gal-(1 → 4)-
␤-d-Gal-(1 → 4)-d-Glc) was observed by TLC: RF of the product
GalLac = 0.16 (RF of Lac = 0.25; isopropanol:NH4OH:H2O 7:3:2, v/v)
and confirmed by MS.
The four glyco(lipo)peptides, 3a-d, and UDP-Gal were dissolved
in HEPES/MnCl2 buffer and reacted under catalysis of LgtC at 37 ◦C
(Scheme 2). The progress of glycosylation of the lipid modified Enk
was monitored by LC–MS, ES-MS and RP-HPLC (with samples fil-
tered through SPE Vydac® columns prior to analysis to remove the
salt). Interestingly, the products were eluted via LC–MS at very sim-
ilar times to the starting materials: GalGalEnkC16 (4b) 5.23 min and
GalEnkC16 5.31 min; GalLacEnkC16 (4d) 5.06 min and LacEnkC16
5.09 min. Therefore, RP-HPLC conditions were optimized for better
separation of the starting materials and products (see below). To
further confirm that the peaks observed by RP-HPLC corresponded
(C66H104N8O24
HR-MS (C66H104N8O24
,
1392.7) m/z: 1394.3 [M+H]+ (calcd. 1393.7);
,
1392.7163) m/z: 1393.7209 [M+H]+
(calcd. 1393.7241); RP-HPLC (C8, Vydac® 60–100% B) tR = 13.0,
14.0 min (diastereomers). LC–MS (C8, Vydac®, 0–100% B, 7.6 min,
GalLacEnkC16 1394 m/z: 5.06 min and LacEnkC16 1232 m/z:
5.09 min; UDP-Gal 610 m/z: 1.10 min.
3. Results and discussion
3.1. Chemical synthesis of Leu-enkephalin and its lipid- and/or
sugar-modified derivatives
In this study, we investigated the effect of N-terminal carbo-
hydrate and C-terminal lipid attachments to the Enk peptide on
N. meningitidis LgtC specificity. Conventional stepwise solid phase
peptide synthetic methodology [42,44] was used for the production
of simple carbohydrate and lipid Enk conjugates (3a-d) (Scheme 1),
which served as substrates of LgtC.
The Dde-C16 was coupled on the Rink Amide-MBHA resin fol-
lowed by couplings of Enk peptide sequence (Tyr-Gly-Gly-Phe-Leu)
using activated Fmoc-l-amino acids with HBTU in DMF and DIPEA.
The carboxylic acid functionalized carbohydrate derivatives, 2a-b,
were coupled to the N-terminus of the Enk peptide or lipopeptide
under similar conditions. However, only 2 equivalents (instead
of 4) of glycosyl succinates, 2a-b, and a prolonged coupling time

200
P. Simerska et al. / Journal of Molecular Catalysis B: Enzymatic 97 (2013) 196–202
Scheme 1. Chemical synthesis of glyco(lipo)peptides 3a-3d. i + ii (for 2a): dry THF, H2, 10% Pd/C, DMAP (cat.), succinic anhydride, RT, overnight; i (for 2b): DCM, H-cube
hydrogenation (H2, Pd/C, 4 h, 40 ◦C, flow rate 2 mL/min); ii (for 2b): dry pyridine, DMAP (cat.), succinic anhydride, 24 h, N2; iii: stepwise solid phase peptide synthesis of
Enk-NH-MBHA resin (for 3a and 3c) or Enk-C16-NH-MBHA resin (for 3b and 3d), DMF, HBTU, DIPEA, overnight; iv: 75% (v/v) hydrazine hydrate in MeOH; v: TFA:TIPS:H2O
(95:2.5:2.5), 3 h, cold diethyl ether.
to products containing Enk, absorbance at 280 nm (for Tyr) and
absorbance at 260 nm (for Phe) were measured and the 260/280
ratio was calculated. The 260/280 ratio was <1, confirming the RP-
HPLC additional peak corresponded to peptide.
LacEnk (3c) was completely converted into GalLacEnk (4c)
after 24 h (Fig. 3). Reaction with the GalEnk (3a) acceptor was not
completed after 24 h, therefore it was reacted for an additional 3
days, when the highest amount of the GalGalEnk (4a) was observed
Scheme 2. Enzymatic glycosylation of LacEnkC16 (3d). The reaction was performed using N. meningitidis LgtC galactosyltransferase, UDP-Gal in HEPES/MnCl2 buffer and 5%
DMSO at 37 ◦C for 7 days to give GalLacEnkC16 (4d) in 20% yield.

P. Simerska et al. / Journal of Molecular Catalysis B: Enzymatic 97 (2013) 196–202
201
of DMSO were tested (0–10%) to increase the solubility of 3b
and 3d. Five percent DMSO was found to be optimal to balance
the substrate solubility (increasing with DMSO additions) and
enzymatic activity (decreasing with DMSO additions).
4. Conclusion
We successfully designed and developed the enzymatic syn-
thesis of modified Enk derivatives. Our results demonstrate the
prospective use of glycosyltransferases for the glycosylation of
glyco(lipo)peptides. The enzymatic reactions catalyzed by LgtC
galactosyltransferase from N. meningitidis were performed on an
analytical scale giving purified yields of between 13% and 71%.
Moreover, the LgtC enzyme had broader substrate specificity than
previously reported (also active toward galactosyl terminated
glycopeptides and glycolipopeptides). The glyco(lipo)peptides
described in this study could potentially improve delivery of Enk
to the CNS and will therefore be tested in the future in vitro and
in vivo experiments.
Fig. 3. Reaction progress of the enzymatic synthesis of GalLacEnk (4c) using LgtC
galactosyltransferase monitored by analytical RP-HPLC on an Alltima® C18 column,
0–70% B.
and by-product formation was detected. The lipid modified Enk
peptides (3b and 3d) were also allowed to react for longer (7 days)
to achieve higher yields (13% and 20%, respectively). UDP-Gal
and LgtC were added to the reaction mixtures portion-wise to
overcome problems with their instability. All peptides 4a-d were
purified by RP-HPLC using C18 Alltima® column and 10–35%
gradient of solvent B (4a and 4c) or 40% DMSO addition, C8
column and 0–70% gradient of solvent B (4b and 4d). Even though
glycopeptide 3c was converted to compound 4c quantitatively
(Fig. 3), the isolated yield was 71%. The pure glycopeptide 4a was
obtained in 45% yield and glycolipopeptides 4b and 4d in 13% and
20% yields, respectively (Table 2). Blank reactions omitting the
enzyme did not yield any products.
Acknowledgements
We would like to acknowledge the Australian Research
Council for their support of this work with the Discovery
Project Grant DP1092829, Professorial Research Fellowship to I.T.
(DP110100212) and an Australian Postdoctoral Fellowship to P.S.
(DP1092829). We would also like to thank to Dr. Alun Jones for his
help with HR-MS.
Appendix A. Supplementary data
Substrate specificity of LgtC was studied previously and showed
667-fold higher relative LgtC activity [35], 500–1200-fold higher
second-order rate [34], or 5-fold higher reaction rate [48] toward
d-lactosyl acceptors when compared with those of d-galactosyl. We
also noticed higher activity of the LgtC enzyme toward d-lactosyl
terminated acceptors, 3c and 3d, although, the difference in activi-
ties (toward d-lactosyl and d-galactosyl acceptors) was not as high
as those published previously. Remarkably, we were not only able
to synthesize GalLacEnk (4c) in 71% yield, but also GalGalEnk (4a)
in 45% yield which is a 9× higher yield than that reported by
Zhang et al. when using similar substrates [35]. The synthesis of 4a
required the use of over 2× higher concentration of UDP-Gal and
LgtC enzyme as well as a longer reaction time (4 days compared to
24 h) than that used for the synthesis of 4c (Table 2). The formation
of several additional galactosylated by-products was also observed
by MS when using GalEnk (3a) and GalEnkC16 (3b) as acceptors in
the enzymatic glycosylation reactions (data not shown).
Supplementary data associated with this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.molcatb.2013.
08.018.
References
[1] P. Varamini, F.M. Mansfeld, J.T. Blanchfield, B.D. Wyse, M.T.I. Smith, I. Toth, PLoS
ONE 7 (8) (2012) e41909, http://dx.doi.org/10.1371/journal.pone.0041909.
[2] A.H. Dickenson, Br. Med. Bull. 47 (1991) 690–702.
[3] L. Pollaro, C. Heinis, MedChemComm 1 (2010) 319–324.
[4] P. Simerska, P.M. Moyle, I. Toth, Med. Res. Rev. 31 (2011) 520–547.
[5] A.T. McCoy, C.C. Benoist, J.W. Wright, L.H. Kawas, J.M. Bule-Ghogare, M.Y. Zhu,
S.M. Appleyard, G.A. Wayman, J.W. Harding, J. Pharmacol. Exp. Ther. 344 (2013)
141–154.
[6] A.S. Abdelrahim, P. Simerska, I. Toth, Int. J. Pharm. 430 (2012) 120–128.
[7] T.O. Price, S.A. Farr, X. Yi, S. Vinogradov, E. Batrakova, W.A. Banks, A.V. Kabanov,
J. Pharmacol. Exp. Ther. 333 (2010) 253–263.
[8] C.D. Cros, I. Toth, J.T. Blanchfield, Bioorg. Med. Chem. 19 (2011) 1528–1534.
[9] R.D. Egleton, S.A. Mitchell, J.D. Huber, M.M. Palian, R. Polt, T.P. Davis, J. Pharma-
col. Exp. Ther. 299 (2001) 967–972.
The natural substrate of LgtC enzyme was reported to be
lipooligosaccharide [34]. Nevertheless, our results showed lower
yields when facilitating reactions with lipidated constructs 3b
and 3d (13% and 20%, respectively) compared with non-lipidated
constructs 3a and 3c (45% and 71%, respectively). This was due
to poor solubility resulting from the long alkyl side chain of the
lipoamino acid. Therefore, it was necessary to include DMSO in
reaction mixtures that contained lipidated peptides (no products
were formed in the reactions without DMSO). Varied amounts
[10] N.O. Elmagbari, R.D. Egleton, M.M. Palian, J.J. Lowery, W.R. Schmid, P.
Davis, E. Navratilova, M. Dhanasekaran, C.M. Keyari, H.I. Yamamura, F. Por-
reca, V.J. Hruby, R. Polt, E.J. Bilsky, J. Pharmacol. Exp. Ther. 311 (2004)
290–297.
[11] J. Wang, D.J. Hogenkamp, M. Tran, W.Y. Li, R.F. Yoshimura, T.B.C. Johnstone,
W.C. Shen, K.W. Gee, J. Drug Target. 14 (2006) 127–136.
[12] W.A. Gibbons, R.A. Hughes, M. Charalambous, M. Christodoulou, A. Szeto, A.E.
Aulabaugh, P. Mascagni, I. Toth, Liebigs Ann. Chem. 12 (1990) 1175–1183.
[13] M. Nomoto, K. Yamada, M. Haga, M. Hayashi, J. Pharm. Sci. 87 (1998) 326–332.
[14] Y. Koda, M. Del Borgo, S.T. Wessling, L.H. Lazarus, Y. Okada, I. Toth, J.T. Blanch-
field, Bioorg. Med. Chem. 16 (2008) 6286–6296.
[15] L. Cipolla, F. Peri, Mini-Rev. Med. Chem. 11 (2011) 39–54.
[16] R.K. Shukla, A. Tiwari, Crit. Rev. Ther. Drug Carrier Syst. 28 (2011) 255–292.
[17] P. Simerska, P.M. Moyle, C. Olive, I. Toth, Curr. Drug Deliv. 6 (2009) 347–358.
[18] R. Hevey, C.C. Ling, Future Med. Chem. 4 (2012) 545–584.
[19] M. Fiore, B. Thomas, V. Dulery, P. Dumy, O. Renaudet, New J. Chem. 37 (2013)
286–289.
Table 2
Substrate specificity of LgtC toward modified Enk acceptors (1 ␮mol).
Acceptor (1 ␮mol)
UDP-Gal (␮mol)
LgtC (␮L)
Purified yield (%)
[20] W. Zhong, M. Skwarczynski, P. Simerska, M.F. Good, I. Toth, Tetrahedron 65
(2009) 3459–3464.
[21] I. Toth, P. Simerska, Y. Fujita, Int. J. Pept. Res. Ther. 14 (2008) 333–340.
[22] B.L. Wilkinson, S. Day, R. Chapman, S. Perrier, V. Apostolopoulos, R.J. Payne,
Chem. Eur. J. 18 (2012) 16540–16548.
GalEnk (3a)
GalEnkC16 (3b)
LacEnk (3c)
6.5
5.8
3.0
5.3
45
117
21
45
13
71
20
LacEnkC16 (3d)
150

202
P. Simerska et al. / Journal of Molecular Catalysis B: Enzymatic 97 (2013) 196–202
[23] P. Simerska, A.B. Abdel-Aal, Y. Fujita, M.R. Batzloff, M.F. Good, I. Toth, Pept. Sci.
90 (2008) 611–616.
[35] J.B. Zhang, P. Kowal, J.W. Fang, P. Andreana, P.G. Wang, Carbohydr. Res. 337
(2002) 969–976.
[24] F. Pacifico, L. Laviola, L. Ulianich, A. Porcellini, C. Ventra, E. Con-
siglio, V.E. Avvedimento, Biochem. Biophys. Res. Commun. 210 (1995)
138–144.
[25] H.A. Harvey, N. Porat, C.A. Campbell, M. Jennings, B.W. Gibson, N.J. Phillips, M.A.
Apicella, M.S. Blake, Mol. Microbiol. 36 (2000) 1059–1070.
[26] H.A. Harvey, M.P. Jennings, C.A. Campbell, R. Williams, M.A. Apicella, Mol.
Microbiol. 42 (2001) 659–672.
[27] K.D. Johnstone, M. Dieckelmann, M.P. Jennings, I. Toth, J.T. Blanchfield, Curr.
Drug Deliv. 2 (2005) 215–222.
[28] Z.Y. Liu, Y.Q. Lu, J.B. Zhang, K. Pardee, P.G. Wang, Appl. Environ. Microbiol. 69
(2003) 2110–2115.
[36] Z. Gyorgydeak, J. Tiem, Adv. Carbohydr. Chem. Biochem. 60 (2006) 103–182.
[37] I. Shin, H.J. Jung, M.R. Lee, Tetrahedron Lett. 42 (2001) 1325–1328.
[38] C.C. Lin, L.C. Huang, P.H. Liang, C.Y. Liu, C.C. Lin, J. Carbohydr. Chem. 25 (2006)
303–313.
[39] P. Varamini, F.M. Mansfeld, J.T. Blanchfield, B.D. Wyse, M.T. Smith, I. Toth, J.
Med. Chem. 55 (2012) 5859–5867.
[40] P.V. Murphy, H. Bradley, M. Tosin, N. Pitt, G.M. Fitzpatrick, W.K. Glass, J. Org.
Chem. 68 (2003) 5692–5704.
[41] B.P. Ross, R.A. Falconer, I. Toth, Molbank 2008 (2008) M566.
[42] P. Alewood, D. Alewood, L. Miranda, S. Love, W. Meutermans, D. Wilson, Meth-
ods Enzymol. 289 (1997) 14–29.
[29] R. Shen, S.A. Wang, X.F. Ma, J.Y. Xian, J. Li, L.W. Zhang, P. Wang, Biochemistry
(Moscow) 75 (2010) 944–950.
[43] W.W. Wakarchuk, M. Gilbert, A. Martin, Y.Y. Wu, J.R. Brisson, P. Thibault, J.C.
Richards, Eur. J. Biochem. 254 (1998) 626–633.
[30] S.G. Gouin, J. Kovensky, Tetrahedron Lett. 48 (2007) 2875–2879.
[31] A.V. Gudmundsdottir, M. Nitz, Org. Lett. 10 (2008) 3461–3463.
[32] K. Persson, H.D. Ly, M. Dieckelmann, W.W. Wakarchuk, S.G. Withers, N.C.J.
Strynadka, Nat. Struct. Biol. 8 (2001) 166–175.
[44] M. Schnolzer, P. Alewood, A. Jones, D. Alewood, S.B.H. Kent, Int. J. Pept. Res.
Ther. 13 (2007) 31–44.
[45] M.P. Jennings, Y.N. Srikhanta, E.R. Moxon, M. Kramer, J.T. Poolman, B. Kuipers,
P. van der Ley, Microbiology 145 (1999) 3013–3021.
[33] A.W. Berrington, Y.C. Tan, Y. Srikhanta, B. Kuipers, P. van der Ley,
I.R.A. Peak, M.P. Jennings, FEMS Immunol. Med. Microbiol. 34 (2002)
267–275.
[46] T. Antoine, C. Bosso, A. Heyraud, E. Samain, Biochimie 87 (2005) 197–203.
[47] L.L. Lairson, B. Henrissat, G.J. Davies, S.G. Withers, Annu. Rev. Biochem. 77
(2008) 521–555.
[34] H.D. Ly, B. Lougheed, W.W. Wakarchuk, S.G. Withers, Biochemistry 41 (2002)
6572.
[48] W.W. Wakarchuk, A. Cunningham, D.C. Watson, N.M. Young, Protein Eng. 11
(1998) 295–302.