Multiple glycosylation of de novo designed α-helical coiled coil peptides

Article abstract of DOI:10.1016/j.bmc.2010.03.061

The aim of this study was to investigate the influence of multiple O-glycosylation in α-helical coiled coil peptides on the folding and stability. For this purpose we systematically incorporated one to six β-galactose residues into the solvent exposed positions of a 26 amino acid long coiled coil helix. Surprisingly, circular dichroism spectroscopy showed no unfolding of the coiled coil structure for all glycopeptides. Thermally induced denaturations reveal a successive but relative low destabilization of the coiled coil structure upon introduction of β-galactose residues. These first results indicate that O-glycosylation of the glycosylated variants is easily tolerated by this structural motif and pave the way for further functional studies.

Full text of DOI:10.1016/j.bmc.2010.03.061

Bioorganic & Medicinal Chemistry 18 (2010) 3703–3706
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Multiple glycosylation of de novo designed a-helical coiled coil peptides
*
Jessica A. Falenski, Ulla I. M. Gerling, Beate Koksch
Institute of Chemistry and Biochemistry, Organic Chemistry, Freie Universität Berlin, 14195 Berlin, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
The aim of this study was to investigate the influence of multiple O-glycosylation in a-helical coiled coil
Received 18 January 2010
Revised 18 March 2010
Accepted 25 March 2010
Available online 27 March 2010
peptides on the folding and stability. For this purpose we systematically incorporated one to six b-galact-
ose residues into the solvent exposed positions of a 26 amino acid long coiled coil helix. Surprisingly, cir-
cular dichroism spectroscopy showed no unfolding of the coiled coil structure for all glycopeptides.
Thermally induced denaturations reveal a successive but relative low destabilization of the coiled coil
structure upon introduction of b-galactose residues. These first results indicate that O-glycosylation of
the glycosylated variants is easily tolerated by this structural motif and pave the way for further func-
tional studies.
Keywords:
Glycopeptide
Post-translational modification
Helix
Ó 2010 Elsevier Ltd. All rights reserved.
Solid phase synthesis
1. Introduction
structure is characterized by a periodicity of seven residues, the so-
called heptad repeat, which is commonly denoted (a-b-c-d-e-f-g)n
The
a
-helical coiled coil structure is an ubiquitous protein fold-
(Fig. 1). Positions a and d are typically occupied by nonpolar resi-
dues that form the first recognition domain by hydrophobic core
packing. Charged amino acids in positions e and g form the second
recognition motif by interhelical electrostatic interactions. Polar
residues are often found in the remaining heptad repeat positions
b, c and f, which are solvent exposed. These positions are not di-
rectly involved in helix oligomerization and thus well suited for
modifications such as glycosylation.6 b-Galactose was introduced
by glycosylation of serine’s hydroxyl side chains in positions b, c,
and f. Here, we used b-galactose as an easily accessible residue that
serves as a model to study the impact of O-linked sugars on coiled
coil peptide folding. All peptides were synthesized by conventional
solid phase peptide synthesis (SPPS) following the Fmoc-protecting
strategy and a building block approach to introduce the O-linked b-
galactose.
ing motif with diverse biological functions. It is represented in mo-
tor proteins as myosin and kinesin, in cytoskeletal proteins such as
cortexillin and keratin, in DNA-binding proteins as GCN4, and it
plays a key role during HIV fusion.1,2
Most proteins can be post-translationally modified by glycosyla-
tion. This reversible, covalent omnipresent modification can have
great impact on the structure and function of proteins, ranging from
changes in solubility, oligomerization and conformation, subcellular
localization or binding, to interactions with other proteins.3,4 Glyco-
sylation is essential for a number of fusion proteins to fold into a tri-
meric state, which is again necessary for the approximation of the
pathogen to the cell surface of the host cell.2 However, the prediction
of these changes is still challenging. De novo designed coiled coils
have been proven to mimic protein structures and are therefore use-
ful tools to elucidate protein structure and function. This study ad-
dresses the influence of multiple O-glycosylation on the
Table 1 shows the sequences of the investigated peptides. Basis
peptide B contains two additional serine residues (in positions 13
and 14) in comparison to peptide A to make possible an extension
of glycosylated sites up to six galactose residues per 26mer of pep-
tide. The conformation and stability of all peptides were investi-
gated by circular dichroism spectroscopy.
conformation and stability of
to understand the complex mechanisms involved in protein folding.
a-helical coiled coil peptides in order
Coiled coils consist of two to seven
a-helices wrapped around
each other in a slight left-handed superhelical twist.5 The primary
Abbreviations: Abz, o-aminobenzoic acid; CD, circular dichroism; DBU, diaza-
bicyclo[5,4,0]-undec-7-ene; DIC, diisopropylcarbodiimide; DIEA, N,N-diisopropyl-
ethylamine; DMF, dimethylformamide; ESI-MS, electospray ionization mass
2. Results and discussion
spectrometry; Fmoc, 9-fluorenylmethoxycarbonyl; (Gal-)S, b-D-galactopyranosyl-
2.1. Oligomerization of unglycosylated and highly glycosylated
peptide B
L
-serine; HOAt, 1-hydroxy-7-azabenzotriazole; HOBt, 1-hydroxybenzotriazole;
SPPS, solid phase peptide synthesis; TBTU, O-(benzotriazol-1-yl)-N,N,N0,N0-tetram-
ethyluronium tetrafluoroborate; TFA, trifluoroacetic acid; TIPS, triisopropylsilane.
* Corresponding author. Tel.: +49 30 838 55344.
Analytical ultracentrifugation experiments were performed to
exclude that differences in stability arise from a change in the olig-
E-mail address: koksch@chemie.fu-berlin.de (B. Koksch).
0968-0896/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2010.03.061

3704
J. A. Falenski et al. / Bioorg. Med. Chem. 18 (2010) 3703–3706
Figure 1. Helical wheel representation of the basis peptide B coiled coil dimer with the indicated glycosylation sites at positions 3, 10, 13, 14, 17, and 24.
Table 1
Table 2
Sequences of the coiled coil-based model peptides
Molecular mass distribution from sedimentation velocity experiments and respective
surface areas in percent
Peptide
Sequence
a
a
Peptide
Mw peak 1
Mw peak 2
A
Abz–LESKLKELESKLKELESKLKELESKL-OH
A-10
A-17
A-10/17
B
Abz–LESKLKELE(Gal-)SKLKELESKLKELESKL-OH
Abz–LESKLKELESKLKELE(Gal-)SKLKELESKL-OH
Abz–LESKLKELE(Gal-)SKLKELE(Gal-)SKLKELESKL-OH
Abz-LESKLKELESKLSSLESKLKELESKL-OH
B
3510 (11%)
—
8206 (82%)
8905 (96%)
B-3/10/13/14/17/24
a
In g/mol.
B-17
B-10/13/14
Abz-LESKLKELESKLSSLE(Gal-)SKLKELESKL-OH
Abz-LESKLKELE(Gal-)SKL(Gal-)S(Gal-)SLESKLKELESKL-OH
B-10/13/14/17 Abz-LESKLKELE(Gal-)SKL(Gal-)S(Gal-)SLE(Gal-)SKLKELESKL-
3109 g/mol), in which the dimer fraction is much bigger. For pep-
tide B-3/10/13/14/17/24 (M: 4081 g/mol) sedimentation velocity
experiments reveal an exclusive dimeric structure. As both pep-
tides show a dimeric structure as major compound, glycosylation
has a negligible influence on the oligomerization state of the
peptide.
OH
B-3/10/13/14/ Abz-LE(Gal-)SKLKELE(Gal-)SKL(Gal-)S(Gal-)SLE
17/24
(Gal-)SKLKELE(Gal-)SKL-OH
omerization state due to the glycosylation. Peptide oligomerization
was determined for the unglycosylated control peptide B and for
the extensively glycosylated peptide B-3/10/13/14/17/24 which
contains six galactose residues. Sedimentation molar mass distri-
butions of both peptides were compared. The molecular weights
of these peptides were determined from integration of the diffu-
sion corrected molar mass distribution from sedimentation veloc-
ity experiments (Fig. 2). Table 2 shows that the observed molecular
weights match the values of monomer and dimer for peptide B (M:
2.2. Conformation and stability of the glycosylated coiled coil
peptides
The impact of introduced O-linked b-galactose residues on the
folding and stability of the coiled coil peptides was investigated
by circular dichroism spectroscopy. Surprisingly, characteristic
a-
helical coiled coil CD-spectra with two minima at 208 and
222 nm were observed for all variants (Fig. 3). Temperature in-
duced denaturation showed a successive but unexpectedly low im-
pact on the coiled coil structure upon introduction of up to twelve
b-galactose residues per coiled coil dimer (Fig. 3). Decreased melt-
ing temperature of peptide B compared to peptide A can be as-
signed to the replacement of glutamate and lysine at positions
13 and 14 by serine, which has a lower helix propensity.7
Most likely the destabilizing effect of glycosylation can be
attributed to the sterical demand of the sugar moiety. Calculated
volumes8 reveal that single glycosylation increases the serine side
chain volume dramatically from 26.1 to 168.7 Å. Furthermore, car-
bohydrate hydroxyl groups can be involved in hydrogen bonding
to the backbone amides. These hydrogen bonds compete with
the stabilizing intramolecular hydrogen bonding pattern in a-heli-
ces. However, hydrogen bonding has only a high impact on struc-
tural stability in hydrophobic environment and plays a minor role
in solvated positions.9 Since the galactose residues are incorpo-
rated into the solvent exposed face of the coiled coil helices, hydro-
gen bonding to the peptide backbone will have a negligible effect
on peptide conformation in our study.
Up to date, only few studies dealing with the incorporation of
carbohydrates in coiled coil or helix-loop-helix motifs have been
undertaken and they have only assigned the aspect of single sub-
Figure 2. Molar mass distribution c(M) from sedimentation velocity analysis.
Samples containing 100
lM peptide B (black line) and 100 lM peptide B-3/10/13/
14/17/24 (red line) in 100 mM phosphate buffer pH 7.4, respectively.

J. A. Falenski et al. / Bioorg. Med. Chem. 18 (2010) 3703–3706
3705
Figure 3. CD-spectra of peptides and corresponding glycopeptides A (a) and B (b). Thermal denaturation curves of peptides and corresponding glycopeptides A (c) and B (d).
All peptides were dissolved in 100 mM phosphate buffer pH 7.4 containing 0.025% NaN3.
stitution.10–13 However, the investigation of multiple glycosylation
is of major importance, since most protein-carbohydrate interac-
tions—including the recognition of many bacteria and viruses by
their host—are of multivalent nature.2 Mehta et al have shown that
the introduction of a single monosaccharide into position d of the
hydrophobic core already led to an expected unfolding of the na-
tive coiled coil structure by perturbing the close interhelical side
chain packing. However, herein presented results show that multi-
ple O-glycosylations with the monosaccharide b-galactose in sol-
vent exposed positions are well tolerated in coiled coil
structures. This finding correlates well with the observation that
N-glycosylation sites NXS/T are concentrated in the positions b, c,
reaction mixture was diluted with DCM and washed with 1 M HCl
(3 ꢁ 50 mL) and H2O (2 ꢁ 50 mL). The organic phase was dried with
MgSO4 and the solvents were evaporated. The crude product was
purified by preparative reversed phase HPLC equipped with a Luna™
C8 column (10
l, 250 ꢁ 21.20 mm, Phenomenex). Eluation solvents
were ACN/0.1% TFA and water Millipore/0.1% TFA. The flow rate was
20 mL minꢂ1 and a gradient of 5% to 70% ACN in 30 min was used. In
each run 250 mg raw product was dissolved in 1 mL ACN and in-
jected. ACN and H2O of the combined fractions were evaporated.
The product was dissolved in DCM, which was evaporated immedi-
ately to give a white powder (2.18 g, 47% yield). ESI-MS (M+H)+ calcd
658.2136, obsd 658.2140.
and f of the a-helical coiled coil domain of the glycoprotein lami-
nin.14 Based on these first results further studies are being pre-
formed to elucidate the impact of more complex biologically
relevant carbohydrate moieties.
3.2. Peptide synthesis, purification, and characterization
Peptides were synthesized using a SyroXP-I peptide synthesizer
(Multi-SynTech GmbH) on a 0.05 mM scale according to standard
Fmoc/tBu chemistry. Preloaded Fmoc-Leu–Wang residue
(0.64 mmol/g; NovabiochemÒ-Merck) was used for all peptides.
For standard couplings a fourfold excess of amino acid and coupling
reagents TBTU/HOBt as well as an eightfold excess of DIEA relative to
resin loading was used. All couplings were preformed as double cou-
plings (30 min). The coupling mixture contained 0.23 M NaClO4 to
prevent on-resin aggregation. Glycosylated amino acid was intro-
3. Materials and methods
a
3.1. Synthesis of N -(9-fluorenylmethoxycarbonyl)-3-O-(2,3,4,6-
tetra-O-acetyl-b-D-galactopyranosyl)-L-serine
The synthesis of the building block for SPPS followed the instruc-
tions of Kihlberg and co-workers (1).15 Galactosepentaacetate
duced manually. Fmoc-Ser(b-D-Gal(Ac)4)-OH was activated by
(2.73 g, 7 mM) and Fmoc-L-Ser-OH (2.89 g, 8.37 mM) were dissolved
means of DIC/HOAt without the addition of base to prevent racemi-
sation and deprotection of the sugar protecting groups. The molar
in70 mlACNandBF3ꢀEt2O(2.64 ml, 21 mM)wasadded slowlyunder
nitrogen atmosphere. After 1 h stirring at room temperature the

3706
J. A. Falenski et al. / Bioorg. Med. Chem. 18 (2010) 3703–3706
excess of Fmoc-Ser(b-
D
-Gal(Ac)4)-OH and coupling reagents was re-
tion (AUC) was performed on a XL-I (Beckman-Coulter, Palo Alto,
CA) ultracentrifuge at 25 °C applying the UV–vis absorption op-
tics at 318 nm and using standard 12 mm double sector center
pieces. Sedimentation velocity experiments were performed at
duced to 1.5-fold for the first and 0.5-fold for the second coupling
and reaction time was extended to three hours. A mixture of DBU
and piperidine (2% each) in DMF was used for Fmoc deprotection
(4 ꢁ 5 min). Peptides were cleaved from resin by treatment with
2 mL TFA/TIPS/H2O (90:9:1) for three hours. The resin was washed
twice with TFA (1 mL) and DCM (dry, 1 mL) and excess solvent was
removed by evaporation. The peptides were precipitated with cool
diethyl ether, centrifuged, and dried in vacuum. Glycopeptides were
deacetylated by the treatment with sodium methanolate (20 mM in
methanol, pH 11) over night. Thereafter the pH was adjusted to 4 by
addition of concentrated acetic acid and the solvent was evaporated.
Purification of all peptides was carried out by preparative reversed
60,000 rpm and overall peptide concentrations of 100 lM. The
samples were dissolved in 100 mM phosphate buffer at pH 7.4.
The partial specific volume of the samples was determined in
a density oscillation tube (DMA 5000, Anton Paar, Graz) to be
0.611 ml gꢂ1 for peptide B and 0.715 ml gꢂ1 for peptide B-3/10/
13/14/17/24. The buffer density (
viscosity ( = 0.009243 P), both at 25 °C, were calculated using
the freeware program SEDNTERP
q = 1.010934 g/ml) and buffer
g
16,17
.
The sedimentation velocity
18
data were evaluated using the program SEDFIT yielding the sed-
imentation molar mass distributions shown in Figure 2.
phase HPLC equipped with a Luna™ C8 (10
l, 250 ꢁ 21.20 mm, Phe-
nomenex) column. Eluation solvents were ACN/0.1% TFA and water
Millipore/0.1% TFA. The flow rate was 20 mL minꢂ1. Purified pep-
tides were characterized by analytical HPLC and ESI-MS.
Acknowledgment
We thank the DFG (SFB 765, ‘Multivalency as Chemical Organi-
zation and Action Principle: New Architectures, Functions, and
Applications’) for financial support.
3.3. Concentration determination
Peptide concentrations were estimated by UV spectroscopy
using the absorption maximum at 320 nm of o-aminobenzoic acid
(Abz), which was attached to each N-terminus. A calibration curve
was recorded using different concentrations of H2N-Abz-Gly-
OHꢀHCl. After concentration determination peptide solutions were
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmc.2010.03.061.
diluted to yield 100 lM peptide.
3.4. Circular dichroism
References and notes
Peptides were dissolved in phosphate buffer (100 mM, pH 7.4,
0.025% NaN3). CD-spectra were recorded on a Jasco-715 spectropo-
larimeter at 20 °C using 0.1 cm Quartz SuprasilÒ cuvettes (Hellma).
Thermal induced denaturations were preformed by recording the
CD-signal at 222 nm upon heating from 10 to 100 °C (3 °C/min).
Elipticity was normalized to concentration (c/mol Lꢂ1), number of
residues (n = 27, including the N-terminal label Abz) and path
length (l/cm) using Eq. 1 where hobs is the measured ellipticity in
millidegrees and [h] the mean residue ellipticity in 103 deg cm2 d-
molꢂ1 residueꢂ1. All measurements were repeated three times and
1. Parry, D. A. D.; Bruce Fraser, R. D.; Squire, J. M. J. Struct. Biol. 2008, 163, 258.
2. Dutch, R. E.; Jardetzky, T. S.; Lamb, R. A. Biosci. Rep. 2000, 20, 597.
3. Varki, A. Glycobiology 1993, 3, 97.
4. Dwek, A. D. Chem. Rev. 1996, 96, 683.
5. Woolfson, D. N. Adv. Protein Chem. 2005, 70, 79.
6. Pagel, K.; Koksch, B. Curr. Opin. Chem. Biol. 2008, 12, 730.
7. Chou, P. Y.; Fasman, G. D. Biochemistry 1974, 13, 222.
8. Zhao, Y. H.; Abraham, M. H.; Zissimos, A. M. J. Org. Chem. 2003, 68, 7368.
9. Scheike, J. A.; Baldauf, C.; Spengler, J.; Albericio, F.; Pisabarro, M. T.; Koksch, B.
Angew. Chem., Int. Ed. 2007, 46, 7766.
10. Mehta, S.; Meldal, M.; Ferro, V.; Duus, J. Ø.; Bock, K. J. Chem. Soc., Perkin Trans. 1
1997, 9, 1365.
11. Vijayalekshmi, S.; George, S. K.; Andersson, L. K.; Kihlberg, J.; Baltzer, L. Org.
Biomol. Chem. 2003, 1, 2455.
the average was calculated.
hobs
½hꢃ ¼
ð1Þ
12. Andersson, L. K.; Dolphin, G. T.; Kihlberg, J.; Baltzer, L. J. Chem. Soc., Perkin Trans.
2 2000, 3, 459.
10; 000 ꢀ l ꢀ c ꢀ n
13. Andersson, L.; Stenhagen, G.; Baltzer, L. J. Org. Chem. 1998, 63, 1366.
14. Beck, K.; Hunter, I.; Engel, J. FASEB J. 1990, 4, 148.
15. Kihlberg, J.; Elofsson, M.; Salvador, L. A. Methods Enzymol. 1997, 289, 221.
16. http://www.rasmb.bbri.org.
3.5. Analytical ultracentrifugation
17. Lebowitz, J.; Lewis, M. S.; Schuck, P. Protein Sci. 2002, 11, 2067.
18. Schuck, P. Biophys. J. 2000, 78, 1606.
Prior to AUC analysis, the proteins were extensively dialyzed
against phosphate buffer (100 mM, pH 7.4) using a SpectraPorÒ-
Membran (MWCO; Carl Roth GmbH). Analytical ultracentrifuga-