Selective Inhibition of Aggregation and Toxicity of a Tau-Derived Peptide using Its Glycosylated Analogues

Article abstract of DOI:10.1002/chem.201504950

Protein glycosylation is a ubiquitous post-translational modification that regulates the folding and function of many proteins. Misfolding of protein monomers and their toxic aggregation are the hallmark of many prevalent diseases. Thus, understanding the role of glycans in protein aggregation is highly important and could contribute both to unraveling the pathology of protein misfolding diseases as well as providing a means for modifying their course for therapeutic purposes. Using β-O-linked glycosylated variants of the highly studied Tau-derived hexapeptide motif VQIVYK, which served as a simplified amyloid model, we demonstrate that amyloid formation and toxicity can be strongly attenuated by a glycan unit, depending on the nature of the glycan itself. Importantly, we show for the first time that not only do glycans hinder self-aggregation, but the glycosylated peptides are capable of inhibiting aggregation of the non-modified corresponding amyloid scaffold. Amyloid attenuation: Using β-O-linked glycosylated variants of the Tau-derived hexapeptide motif VQIVYK, which served as a simplified amyloid model scaffold, it is demonstrated that amyloid formation and toxicity can be strongly attenuated by a glycan unit, depending on the nature of the glycan itself. Moreover, the glycosylated peptides inhibit aggregation of the non-modified corresponding amyloid scaffold.

Full text of DOI:10.1002/chem.201504950

DOI: 10.1002/chem.201504950
Full Paper
&
Chemical Biology |Hot Paper|
Selective Inhibition of Aggregation and Toxicity of a Tau-Derived
Peptide using Its Glycosylated Analogues
[
a]
[b]
[b]
[a]
Moran Frenkel-Pinter, Michal Richman, Anna Belostozky, Amjaad Abu-Mokh,
[a]
[b]
[a]
Ehud Gazit,* Shai Rahimipour,* and Daniel Segal*
Abstract: Protein glycosylation is a ubiquitous post-transla-
tional modification that regulates the folding and function
of many proteins. Misfolding of protein monomers and their
toxic aggregation are the hallmark of many prevalent diseas-
es. Thus, understanding the role of glycans in protein aggre-
gation is highly important and could contribute both to un-
raveling the pathology of protein misfolding diseases as well
as providing a means for modifying their course for thera-
peutic purposes. Using b-O-linked glycosylated variants of
the highly studied Tau-derived hexapeptide motif VQIVYK,
which served as a simplified amyloid model, we demonstrate
that amyloid formation and toxicity can be strongly attenu-
ated by a glycan unit, depending on the nature of the
glycan itself. Importantly, we show for the first time that not
only do glycans hinder self-aggregation, but the glycosylat-
ed peptides are capable of inhibiting aggregation of the
non-modified corresponding amyloid scaffold.
Introduction
correct folding and maintaining the structural integrity, thus
preventing protein aggregation. Protein aggregation is the
hallmark of protein misfolding diseases, which are character-
ized by the self-assembly of monomers of certain proteins into
toxic oligomers and fibrils composed of b-sheet structures,
Protein glycosylation, the abundant enzyme-directed site-spe-
cific process that attaches glycans to proteins, regulates the
folding and function of many proteins. Two main types of pro-
tein glycosylation are known in the secretory pathway, classi-
fied according to the nature of the linkage between the core
region of the glycan and the modified residue in the protein:
N-glycosylation, which occurs on asparagine residue, and
O-glycosylation, which occurs on a serine or threonine resi-
[
5]
termed amyloids. These conditions, include amyloid plaques
of Ab peptides and neurofibrillary tangles (NFTs)—aggregates
of hyperphosphorylated Tau protein in Alzheimer’s disease
(AD) and amyloid aggregates composed of the prion protein
[
5b,6]
in Creutzfeldt–Jakob disease (CJD).
Notably, various amy-
[
1]
due. Many monosaccharides, including N-acetylgalactosamine
GalNAc), N-acetylglucosamine (GlcNAc) and galactose are in-
loid forming proteins including Tau and the prion protein are
glycosylated. Thus, understanding the effect of glycans on pro-
tein self-assembly, which appears to be an initial key step in
the pathology of these amyloidogenic diseases, can both con-
tribute to unraveling the pathology of these diseases as well
as provide means for modifying the course of these diseases
for therapeutic purposes.
(
[2]
volved in the O-glycosidic bond formation in nature. In addi-
tion, many proteins within the nucleus, cytoplasm and mito-
chondria are dynamically glycosylated with GlcNAc on their
serine or threonine residues through a b-glycosidic linkage.
This process, termed O-GlcNAcylation, is highly competitive
[
3]
with phosphorylation on the same or adjacent amino acids.
Various studies have shown that glycosylation has many im-
Glycan chains attached to proteins often hinder their aggre-
gation rate by modulating the conformational properties of
[4]
[7]
plications on protein folding. Glycan chains attached to nas-
cent proteins are believed to be important for promoting their
the protein involved. Specifically, several studies have sug-
gested that glycosylated peptides favor conformations in
which the peptide backbone bends away from the bulky gly-
[
8]
[
a] M. Frenkel-Pinter, A. Abu-Mokh, Prof. E. Gazit, Prof. D. Segal
Department Molecular Microbiology and Biotechnology, and
the Interdisciplinary Sagol School of Neurosciences
George S. Wise Faculty of Life Sciences, Tel-Aviv University
Tel Aviv 69978 (Israel)
cosylation site. For example, the Tau protein in the brain of
AD patients is less O-GlcNAcylated and more hyperphosphory-
lated than in healthy individuals. Hyperphosphorylation of Tau
has been causally linked to its propensity to form toxic amy-
loid aggregates, while O-GlcNAcylation of Tau has been shown
E-mail: dsegal@post.tau.ac.il
[
7c,m]
ehudg@post.tau.ac.il
to reduce its aggregation rate.
Indeed, inhibiting the
[
b] Dr. M. Richman, A. Belostozky, Prof. S. Rahimipour
Department of Chemistry, Bar-Ilan University
Ramat-Gan 5290002 (Israel)
enzyme b-N-acetylglucosaminidase (OGA) that removes
O-GlcNAc from proteins, increased Tau O-GlcNAcylation in tau-
opathy model mice, decreased formation of Tau aggregates
E-mail: rahimis@biu.ac.il
[
7c]
and reduced neuronal cell loss. Despite these findings which
imply a role for glycans in reducing protein aggregation, a sys-
Supporting information for this article can be found under http://
dx.doi.org/10.1002/chem.201504950.
Chem. Eur. J. 2016, 22, 5945 – 5952
5945
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
tematic study comparing the effect of different glycans on
amyloid formation has not been performed, except for few re-
ployed in this study and their abbreviations are shown in
Table 1. All peptides were synthesized using the solid-phase
peptide synthesis.
[7k,l,n]
ports utilizing a prion-derived amyloid scaffold.
Importantly, in nature both glycosylated and non-glycosylat-
ed variants of a glycoprotein coexist in the cellular environ-
ment. Thus, they may interact with each other and impact
their respective aggregation. The studies mentioned above
have all addressed only the effect of the glycan on the self-as-
sembly of the peptide/protein, yet none explored a possible
effect of a glycopeptide on the corresponding non-glycosylat-
ed amyloid scaffold as likely occurs in the cellular milieu. Nota-
bly and interestingly, as opposed to well controlled protein
glycosylation, the non-enzymatic process of protein glycation,
which occurs mainly on lysine residues, was reported to often
To examine the ability of glycosylated PHF6-derived peptides
to form amyloid structures, the peptides were incubated in the
presence of heparin and their aggregation kinetics was fol-
[
11]
lowed by the thioflavin S (ThS) fluorescence assay. The re-
sults suggested that while PHF6 and the non-modified 7aa
peptide aggregate rapidly within less than 30 min (Figure 2
and Figure S1 in the Supporting Information), glycosylation of
the 7aa scaffold with either Gal, GlcN and GlcNAc drastically re-
duced its propensity to aggregate (Figure 2).
[9]
accelerate amyloid fibril formation.
Table 1. Sequences of the PHF6-based model peptides used in this
study.
Most glycosylated proteins harbor large N- or O-linked
glycan trees, which are very difficult to synthesize and conju-
gate to a full-length protein of interest. To simplify this system,
we have studied the effect of various monosaccharides on the
aggregation propensity of a Tau-derived peptide, Ac-SVQIVYK-
Peptide
PHF6
Sequence
2
[Ac-VQIVYK-NH ]
7aa
2
[Ac-SVQIVYK-NH ]
7aa–Gal
[Ac-S(bGal)VQIVYK-NH2]
2
[Ac-S(bGlcN)VQIVYK-NH ]
[Ac-S(bGlcNAc)VQIVYK-NH ]
2
NH (corresponding to residues 305–311 in the full-length, 441
7aa–GlcN
2
7aa–GlcNAc
amino acids long Tau protein), as an amyloid scaffold. This 7aa
peptide is based on the short fragment VQIVYK (also known as
PHF6), which was shown to be critical for Tau aggregation into
[10]
oligomers and formation of NFTs. We employed Ac-SVQIVYK-
NH as a simplified amyloidogenic model peptide, bearing in
2
mind that it is not known to carry glycans in vivo. We decorat-
ed this scaffold with various glycans, including b-linked galac-
tose (Gal), glucosamine (GlcN) and N-acetylglucosamine
(
GlcNAc), on a single glycosylation site on the serine residue.
We examined in vitro the effect of the various glycans on the
self-assembly propensity of the modified PHF6 and on the in-
hibitory properties of the corresponding glycosylated peptides
towards amyloidogenic aggregation of non-modified native
PHF6.
Our results indicate that glycosylation dramatically de-
creased self-fibrillogenesis of Ac-SVQIVYK-NH , and in addition
2
rendered it remarkably efficient in inhibiting the aggregation
of the non-modified PHF6. Importantly, the effect of glycosyla-
tion appears to be strongly dependent on the nature of the
glycan itself.
Results and Discussion
Figure 1. Chemical structure of PHF6-derived peptides used in the study.
Seven amino acid long, Ac-SVQIVYK-NH , extended versions of PHF6, were
synthesized, and glycosylated on their serine residues with various glycan
ThS analysis reveals that glycosylation of the PHF6-derived
peptide inhibit its amyloidogenic aggregation
2
moieties, including galactose (Gal), glucosamine (GlcN) and N-acetyl glucose-
amine (GlcNAc).
To explore whether glycosylation of PHF6 scaffold modulates
its aggregation propensity, an extended version of it was gly-
cosylated on an upstream Ser with different glycans, including
Gal, GlcN and GlcNAc (Table 1 and Figure 1). All the glycans
had the same b-glycosidic bond conformation to exclude
effect of anomerization of the C1 carbon. GlcN and GlcNAc
differ only in their charge due to the free amino group of GlcN
at the C2 position, whereas Gal lacks the amino group at C2
position and differs in the orientation of a single hydroxyl
group at the C4 position. The sequences of all peptides em-
The various glycans have distinct effects on the secondary
structure of the amyloid scaffold
In order to study the effect of the different glycans on the sec-
ondary structure of PHF6-derived peptide, circular dichroism
(CD) spectroscopy was used. In the absence of heparin, all
PHF6-derived peptides exhibited a large negative peak near
Chem. Eur. J. 2016, 22, 5945 – 5952
www.chemeurj.org
5946
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 2. ThS analysis of self-assembly of PHF6-derived peptides. Represen-
tative normalized end-point ThS fluorescence obtained after 60 min incuba-
tion of 100 mm of the 7aa-derived peptides in the presence of 1 mm heparin
at 258C. The ThS fluorescence value observed from PHF6 signal is consid-
ered as 100% aggregation. The Student’s t-test analysis showed *, P<0.05,
**, P<0.01.
[
12]
2
00 nm, indicating a random coil conformation (Figure 3A).
Addition of heparin to the non-glycosylated PHF6-derived pep-
tides caused secondary structural changes over time from
a random coil to a b-sheet conformation, characterized by
a strong positive peak around 195 nm and a negative peak
around 220 nm (Figure 3B). Glycosylation of the 7aa peptide
with GlcN drastically inhibited the structural transformation
from random coil to b-sheet structure, while conjugation with
Gal only partially inhibited this transformation. In contrast, gly-
cosylation of the 7aa peptide with GlcNAc had no effect on
the secondary structure of the peptide (Figure 3B).
Figure 3. CD spectra of the various peptides in the presence or absence of
heparin. The secondary structure of 7aa-derived peptides incubated for 3 h
at 258C in the: A) absence, or B) presence of heparin.
its weak ThS signal, since thioflavin-based fluorophores favor
TEM analysis of PHF6-derived peptides reveals that glycans
alter the morphology of the resulting amyloid aggregates
[
11,13]
binding to specific amyloid structures.
Collectively, these results demonstrate that various glycans,
linked by a b-glycosidic bond to the Tau-derived PHF6 amylo-
idogenic scaffold, reduce the self-aggregation propensity of
the glycosylated PHF6-derived peptides and alter their secon-
In order to study the effect of the glycans on the morphology
of the PHF6-derived peptide, TEM analysis was performed on
the samples of the various modified and non-modified pep-
tides. Figure 4 shows the TEM images of the peptides after
being incubated with heparin for 25 min. PHF6 and the non-
modified 7aa peptide formed clustered large fibrils, which
were helically twisted around their central axis. PHF6 fibrils
were all straight, as opposed to the fibrils formed by the non-
modified 7aa peptide, which were a mixture of straight and
moderately curved fibrils. In contrast, only very few fibrils were
observed for the 7aa–GlcN peptide (Figure 4), in agreement
with the results of the ThS and CD analyses. The 7aa–GlcNAc
peptide, which exhibited a b-sheet conformation in the CD
analysis (Figure 3B), also formed fibrillar structures similar to
those formed by PHF6, although fewer fibrils were detected
and various amorphous structures were observed. These TEM
results are in good agreement with the ThS results, indicating
the lower amyloidogenic propensity of 7aa–GlcNAc to self-ag-
gregate (Figure 2). The 7aa–Gal peptide, which exhibited a mix-
ture of random coil and b-sheet structures in the CD spectra
(
Figure 3B), formed large clustered fibrils that were much
Figure 4. TEM images of self-assembled 7aa-based peptides. TEM images
were taken after the peptide analogues (100 mm) were allowed to aggregate
at 258C for 25 min following initiation of aggregation by the addition of
heparin (1 mm).
more curved in their morphology than those formed by PHF6
and the non-modified 7aa peptide (Figure 4). The highly
curved fibril morphology of 7aa–Gal peptide may account for
Chem. Eur. J. 2016, 22, 5945 – 5952
www.chemeurj.org
5947
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
dary conformation and aggregate morphology (Figure 3,
Figure 4). Our findings indicate that the effect of glycosylation
on self-aggregation of the amyloid scaffolds is strongly depen-
dent on the nature of the glycan itself, as the different glycans
used in our study (b-linked Gal, GlcN and GlcNAc) demonstrat-
ed different effects on amyloidogenic self-aggregation propen-
sity of the Tau-derived peptide. Comparable results have been
reported for a prion-derived amyloid scaffold using a-man-
nose, a-galactose, a-GalNAc, a-GalNAc, a-GlcNAc, b-GlcNAc, or
These results correlate well with the CD and TEM results,
where only 7aa–GlcN did not generate b-sheet-rich fibrillar
structures (Figure 3B and Figure 4).
Our results may shed light on the effect of the various gly-
cans on PHF6-indecuced cytotoxicity. We observed a positive
strong correlation between aggregation propensity and cyto-
toxicity. The non-aggregative 7aa–GlcN demonstrated very low
toxicity, whereas the non-modified 7aa and the two remaining
7aa–derived glycopeptides, which were more aggregative,
were also more cytotoxic.
[
7k,l,n]
b-GalNAc.
Various elements within the glycan ring, such as the orienta-
tion of a single hydroxyl group at the C4 of the glycan and the
anomeric position of the glycan, have been shown to affect
[7k,l,n]
the self-aggregation propensity of glycopeptides.
In our
study we found that 7aa–GlcN is not aggregative, as opposed
to 7aa–GlcNAc and 7aa–Gal. This is most probably due to the
presence of the charged amino group on C2 of the glycan
ring, which likely induces a charge repulsion between the
PHF6 scaffold and thus decreases its self-assembly propensity.
Notably, two of the glycans used in this study that hindered
self-aggregation of the PHF6-scaffold, namely b-Gal and b-
GlcNAc, were previously reported to slow down in vitro fibrillo-
genesis of other amyloid model peptides, including the prion-
derived peptide, the 26 amino acids long model a-helical hair-
pin peptide and the a-helical hairpin peptide (a-helix/turn/a-
[
7a,k–n]
helix).
To the best of our knowledge, our study is the first
Figure 5. The effect of PHF6-derived peptides on the viability of PC12 cells.
PC12 cells were treated for 24 h with 10 mm 7aa-derived peptides and cell vi-
ability was measured using the MTT assay. The Student’s t-test analysis
showed ***, P<0.001.
to explore the impact of charged b-GlcN on amyloidogenic ag-
gregation propensity, and to demonstrate its high inhibitory
effect.
Glycosylation of the PHF6-derived peptide attenuates its cy-
totoxicity
Table 2 summarizes the effect of the different glycans on
both the structure and toxicity of the PHF6-derived peptides.
Collectively, these results suggest that effect of glycosylation
on toxicity could be best predicted by the CD and TEM analy-
ses and not by the ThS binding capability.
One of the most important basic mechanisms of neuronal dys-
function in amyloid neurodegeneration is the cytotoxic interac-
[
14]
tion of amyloid aggregates with the cell plasma membrane.
We have recently demonstrated that Ac-PHF6 permeates PC12
[15]
cells and dose-dependently reduces their viability. In order
to determine the effect of PHF6-derived peptides on cell viabil-
ity, PC12 cell were incubated for 24 h with the peptides
PHF6-derived glycosylated peptides inhibit the aggregation
rate of native PHF6
(
10 mm) and their survival was determined by the MTT assay.
To explore a possible inhibitory effect of the glycosylated pep-
tides on PHF6 aggregation, we compared the aggregation of
PHF6 in the presence or absence of the peptides listed in
Table 1, at a 1:1 or 10:1 molar ratio, in favor of PHF6 (Figure 6).
Co-incubation of 100 mm of PHF6 with 100 mm of the non-
modified 7aa peptide (Figure 6A) caused a significant increase
Figure 5 shows that AcPHF6 and its 7aa non-glycosylated ana-
logue induced a significant toxicity towards PC12 cells. Glyco-
sylation of the PHF6-derived peptide (7aa) with Gal and
GlcNAc did not alter its toxicity, while its conjugation with
GlcN significantly (P<0.001) reduced the toxicity (Figure 5).
Table 2. Effect of the different glycans on the self-assembly and toxicity of PHF6-derived peptides.
[
a]
[b]
Peptide
PHF6
Secondary structure
Aggregate abundance and morphology
many straight, helically twisted fibrils
Toxicity [%]
b-sheet
b-sheet
31.5Æ1.7
38.8Æ0.8
39.5Æ1.4
9.1Æ0.8
7
7
7
7
aa
many straight and few moderately curved, helically twisted fibrils
many curved fibrils
aa–Gal
aa–GlcN
aa–GlcNAc
mixture of b-sheet and random coil
random coil
b-sheet
very few short fibrils
few curved fibrils, few amorphous structures
43.4Æ0.9
[
a] Aggregation following addition of heparin. [b] According to the TEM analysis.
Chem. Eur. J. 2016, 22, 5945 – 5952
www.chemeurj.org
5948
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
of the total ThS fluorescence signal, as expected. Incubation of
PHF6 (100 mm) with the lower concentration of the non-modi-
fied 7aa peptide (10 mm, Figure 6A) resulted in similar ThS fluo-
rescence signal as PHF6 alone, suggesting that it had no inhib-
itory effect on the aggregation of PHF6. In contrast, incubation
of PHF6 with various 7aa-glycosylated peptides (Figure 6B–D),
markedly decreased the aggregation rate of PHF6 (Table 3). No-
tably, for all glycopeptides, an effect of their dose on inhibition
of PHF6 aggregation was observed (Figure 6B–D). Despite
their inhibitory effect on the aggregation rate of PHF6, the gly-
cosylated peptides mostly did not lead to a decrease in the
ThS fluorescence signal at the end of the experiment (Fig-
ure 6E). As a control, we verified that supplementing PHF6
with PHF6 molecule itself increased its aggregation as mea-
sured by ThS Fluorescence (Figure S2 in the Supporting Infor-
mation).
Table 3. Aggregation rate following co-incubation of PHF6 with the vari-
ous peptides at a 1:1 molar ratio.
Peptide
Average aggregation rate
(normalized to PHF6)
PHF6
7aa
7aa–Gal
1
1.3Æ0.18
0.41Æ0.03
0.41Æ0.07
0.59Æ0.06
7
7
aa–GlcN
aa–GlcNAc
Figure 6. The effect of the 7aa-derived peptides on PHF6 aggregation. Kinetic ThS fluorescence measurements of PHF6 (at 100 mm, 1 mm heparin) in the pres-
ence or absence of: A) non-modified 7aa peptide, B) 7aa–Gal, C) 7aa–GlcN, and D) 7aa–GlcNAc, at 10:1 and 1:1 molar ratio (PHF6/inhibitory peptide). End-
point ThS fluorescence recordings of (A–D) are shown in (E). The end-point ThS fluorescence signal was normalized to PHF6 signal. The Student’s t-test analy-
sis showed *, P<0.05, **, P<0.01.
Chem. Eur. J. 2016, 22, 5945 – 5952
www.chemeurj.org
5949
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
The effect of the parent 7aa peptide and its glycosylated an-
alogues on PHF6 aggregation was also studied by TEM (Fig-
ure S3 in the Supporting Information). The results confirmed
that incubation of non-modified 7aa peptide and its glycosylat-
ed analogues with PHF6 had little or no effect on the amount
of generated fibrils. The results also demonstrate that in-
creased aggregation occurs upon co-incubation of PHF6 with
the higher concentration of the non-modified 7aa peptide (Fig-
ure S3), in accordance with ThS fluorescence results (Fig-
ure 6A). Yet, the morphology of some of the fibrils, resulting in
these co-incubation experiments, appears to be different from
that of PHF6 alone. Cell viability experiments revealed that the
nontoxic 7aa–GlcN does not alter PHF6-induced toxicity to-
wards the PC12 cell line (Figure S4 in the Supporting Informa-
tion).
ing interactions by glycopeptides may have bearing on drug
design.
Conclusions
Glycosylation of proteins play a key role in their structure,
function, and stability, and has been implicated in various dis-
eases involving toxic aggregation of the misfolded proteins.
Our results demonstrate that amyloid formation can be attenu-
ated by a single glycan unit. Nevertheless, the effect of glyco-
sylation appears to be also strongly dependent on the nature
of the glycan itself. We demonstrated that in addition to hin-
drance of self-aggregation by glycans, glycopeptides are capa-
ble of inhibiting the aggregation of non-modified correspond-
ing amyloid scaffolds.
While hindrance of self-aggregation of amyloid scaffolds by
glycans has been previously reported, to the best of our
knowledge, this study demonstrates for the first time that gly-
cosylated peptides are capable of inhibiting the aggregation
of a non-modified corresponding amyloid scaffold. Importantly,
this inhibitory effect appears to be strongly dependent on the
nature of the glycan itself. We note that inhibition of aggrega-
tion by glycopeptides should be also examined in other experi-
We believe that our results may shed new light on the com-
plex mechanism of protein aggregation in vivo, where several
differently modified glycopeptides coexist simultaneously
inside the cell environment. Our results provide valuable infor-
mation for the design of glycopeptide mimetics as possible
therapeutic agents for protein misfolding diseases.
[
16]
mental setups, such as the fast flow microfluidic system. We Experimental Section
hypothesize that the glycopeptides were able to compete for
Peptide synthesis
interaction with native PHF6 monomers and intercalate with
All chemicals and reagents were of analytical grade. Fluorenylme-
them, leading to less interaction between the PHF6 monomers
themselves. This inhibitory effect is of special importance since
in cellular environment both glycosylated and non-glycosylat-
ed variants of a given protein coexist and may interact with
each other, thus impacting their respective aggregation and
toxicity. We speculate that rational conjugation of glycans to
peptides induces steric hindrance that could be used as
a novel way for robust and specific inhibition of amyloidogenic
aggregation, thus expanding the biotechnological applications
thoxycarbonyl (Fmoc)-protected amino acid derivatives, and all
other reagents for solid-phase peptide synthesis were purchased
either from Novabiochem (San Diego CA, USA) or GL Bio-chem
(
Shanghai, China). Unless otherwise stated, all other chemicals
were obtained from Sigma–Aldrich (Rehovot, Israel). Peptides were
purified to homogeneity (>98% purity) by RP-HPLC and analyzed
by mass spectrometry. Fmoc-Ser(b-Ac GlcNHBoc)-OH and Fmoc-l-
3
[
22]
Ser(b-Ac
Gal)-OH were synthesized as previously described.
4
[17]
of glycopeptides for therapeutic purposes.
Fmoc-Ser(b-Ac GlcNHAc)-OH
3
[
7l]
Ho et al. proposed that glycosylation interferes with amy-
loid formation during the nucleation step. This was based on
the observation that Prion-derived glycopeptides aggregated
once seeded with the wild-type prion-derived peptide fibrils.
When considering the self-interactions of the glycosylated pep-
tides or their interactions with PHF6 it is worthwhile to note
that glycans can form a variety of noncovalent bonds with pro-
Fmoc-Ser(b-Ac GlcNHBoc)-OH was treated with 50% TFA in DCM
3
for 2 h. The solvents were concentrated in vacuo, and the crude
product was then dissolved in CH Cl and acetylated using 4 equiv
2
2
of acetic anhydride and 2 equiv of DIPEA in DCM for 2 h. The sol-
vents were evaporated by reduced pressure and the crude material
was then purified by preparative RP-HPLC.
[
18]
tein residues.
The hydroxyl groups of carbohydrates can
Synthesis of glycosylated peptides
form hydrogen bonds with amino acids such as lysine, argi-
nine, histidine, aspartic acid and glutamic acid. On the other
hand, the hexose ring can adopt a conformation in which sev-
eral of its carbon atoms are in a cluster that can form energeti-
cally favorable CH–p stacking interactions with parallel aligned
aromatic rings of, for example, tyrosine, phenylalanine or tryp-
VQIVYK-NH2 was synthesized automatically (Vantage, AAPPTec,
Louisville, KY) on Rink-Amide resin using solid-phase peptide syn-
thesis and employing the common Fmoc strategy. The peptide-
[23]
bound resin was either acetylated following Fmoc deprotection,
using
a solution of acetic anhydride (10 equiv) and DIPEA
(20 equiv) in NMP for 1 h, or coupled overnight with either Fmoc-
[
18,19]
Ser(b-Ac GlcNHBoc)-OH, Fmoc-l-Ser(b-Ac Gal)-OH, Fmoc-Ser(b-
tophan.
It is thus possible that the glycans in our glyco-
3
4
Ac GlcNHAc)-OH (3 equiv) or Fmoc-l-Ser(OtBU)-OH in the presence
3
scaffolds interact with the lysine or with the tyrosine, located
in the PHF6 core itself, disrupting the tight packing interac-
tions that mediate protein self-assembly into fibrils. The aro-
matic tyrosine residue in PHF6 was previously shown to be
of PyBOP (3 equiv), HOAt (3 equiv) and N-methylmorpholine (NMM,
9
equiv) in N-methylpyrrolidone (NMP). Following incorporation of
the amino acids, the terminal Ser was acetylated following Fmoc
deprotection, using a solution of acetic anhydride (10 equiv) and
DIPEA (20 equiv) in NMP for 1 h. Finally, the O-acetyl groups of the
glycosylated Ser were removed with 20% hydrazine hydrate in
[20]
crucial for its aggregation. Given the crucial role of aromatic
[21]
residues in amyloid formation, interference with (p–p) stack-
Chem. Eur. J. 2016, 22, 5945 – 5952
www.chemeurj.org
5950
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
MeOH (2”2.5 h), and the final products were released from the
resin and totally deprotected using a mixture of TFA/triethylsilane/
(10 mL) were placed for 2 min on 400-mesh copper grids covered
with carbon-stabilized Formvar film (Electron Microscopy Sciences
(EMS), Hatfield, PA). Excess fluid was removed, and the grids were
negatively stained with 2% uranyl acetate solution (10 mL) for
2 min. Finally, excess fluid was removed and the samples were
viewed by a JEM-1400 TEM (JEOL) or Tecnai G2 TEM (FEI), operated
at 80 kV.
H O (95:2.5:2.5 v/v) for 2 h. All the glycosylated peptides were
2
then precipitated with cold diethyl ether, purified by preparative
HPLC and analyzed by HRMS. HRMS m/z: calcd for C H N O (Ac-
3
8
62
8
10
VQIVYK-NH ): 790.9465, found 791.4660. HRMS m/z: calcd for
2
C H N O (Ac-SVQIVYK-NH ): 877.0390, found 877.5140. HRMS
41
68 10 11
2
m/z: calcd for C H N O (Ac-S(bGal)VQIVYK-NH ): 1039.1796,
4
7
78 10 16
2
found 1039.5674. HRMS m/z: calcd for C H N O
(Ac-
S(bGlcN)VQIVYK-NH ): 1038.1949, found 1038.5832. HRMS m/z:
4
7
79 11 15
Cell culture experiments
2
calcd for C H N O
found 1080.5936.
(Ac-S(bGlcNAc)VQIVYK-NH2): 1080.2315,
4
9
81 11 16
Pheochromocytoma (PC12) cells were grown in low-glucose Dul-
becco’s modified Eagle medium (DMEM) supplemented with 10%
horse serum and 5% fetal bovine serum (FBS), l-glutamine, penicil-
ThS fluorescence aggregation assay
lin, and streptomycin in a 5% CO , humidified atmosphere, at
2
3
78C. To study the effect of the various peptides on the viability of
In order to ensure the monomeric state of the peptides, the
lyophilized peptides were pretreated for 10 min with 1,1,1,3,3,3-
hexafluoroisopropanol (HFIP) at 378C. Following the dissolution,
the solvent was evaporated using a SpeedVac or a stream of nitro-
gen. Immediately prior to the experiment, the peptides were dis-
solved in water and sonicated for 5 min. The concentration of each
4
PC12 cell cultures, cells were seeded at a cell density of 1”10
cells per 100 mL medium, into each well of a 96-well plate. Fresh
medium containing the peptides (10 mm) was then added to the
wells and incubation was continued for an additional 24 h. Cell via-
bility was then determined by the 3-(4,5-dimethylthiazol-2-yl)-2,5-
diphenyltetrazolium bromide (MTT) assay. For studying a possible
inhibitory effect of the various peptides on PHF6-induced toxicity,
fresh medium containing PHF6 (10 mm) together with increasing
concentrations of the inhibitory peptide (molar ratio of 10:1 and
peptide was then determined (calculated according to e
of
80
2
À1
À1
1
490m cm ) and adjusted to 1 mm concentration as the stock
solution. A stock solution of Thioflavin S (ThS, 2 mm, Sigma–Al-
drich, Rehovot, Israel) and heparin (1 mm, Sigma–Aldrich, Rehovot,
Israel) were prepared in 20 mm MOPS (pH 7.2). For self-assembly
experiments, stock solutions were diluted in 100 mL wells of a 96-
well black plate so that the final mixture contained 100 mm of the
peptide/glycopeptide and 100 mm ThS in 20 mm MOPS. For experi-
ments involving inhibition of PHF6 aggregation, PHF6 (100 mm)
was incubated in the wells of a 96-well black plate with or without
the inhibitory peptides (10 or 100 mm) in MOPS buffer (20 mm) in
the presence of ThS (100 mm). Immediately prior to experiment,
heparin (1 mm) was added to initiate peptide aggregation, as previ-
1
:1, PHF6/inhibitory peptide) were added to the cells. Control wells
received the equivalent amount of water and MOPS (20 mm) were
considered as 100% cell viability. Cells were incubated for an addi-
tional 24 h, and cell viability was then determined by the MTT
assay. Each experiment was performed in quadruplicate and re-
peated at least three times.
Statistics
[10a,24]
Student‘s t-test was performed for evaluating statistical significance
of the observed differences (p-value <0.05 was considered to be
statistically different).
ously described.
As a control, the PHF6 peptide was supple-
mented instead of the inhibitory peptide. Kinetic fluorescence data
were collected at 258C in triplicate or quadruplicate using Infinite
M200 microplate fluorescence reader (Tecan, Switzerland), with
measurements taken at 1 min intervals for 60 min. Excitation and
emission wavelengths were 440 and 490 nm, respectively. All of Acknowledgements
the experiments were repeated 3–4 times to ensure reproducibility.
For calculation of average aggregation rate (Table 3), kinetic data
from the first five minutes of the experiment were used and the
corresponding aggregation curves were fitted to a linear trend line
We are grateful to the members of the D.S., E.G. and S.R. re-
search groups for fruitful discussions, and especially to Dr.
Marina Chemerovski from S.R.’s group.
(
linear regression), from which the slope was calculated.
Keywords: amyloid formation · glycopeptides · glycosylation ·
PHF6 · protein and peptide aggregation
Circular dichroism (CD) analysis
To analyze the secondary structure of the various peptides, each
peptide (100 mm) was incubated in the absence or presence of
[
1] A. Varki, J. D. Esko, K. J. Colley, in Essentials of Glycobiology, 2nd ed.
Eds.: A. Varki, R. D. Cummings, J. D. Esko, H. H. Freeze, P. Stanley, C. R.
1
mm heparin for 3 h at 258C. Prior to measurements, samples
(
were diluted 2.5 times and placed in a 0.1 mm cuvette. CD spectra
were then recorded on a Chirascan spectrometer in the range of
Bertozzi, G. W. Hart, M. E. Etzler), Cold Spring Harbor, 2009.
2] R. S. Haltiwanger, J. B. Lowe, Annu. Rev. Biochem. 2004, 73, 491–537.
[
1
85–260 nm, and the background was subtracted from the CD
[3] G. W. Hart, Y. Akimoto, in Essentials of Glycobiology, 2nd ed. (Eds.: A.
Varki, R. D. Cummings, J. D. Esko, H. H. Freeze, P. Stanley, C. R. Bertozzi,
G. W. Hart, M. E. Etzler), Cold Spring Harbor, 2009.
spectra.
[
4] a) D. Shental-Bechor, Y. Levy, Proc. Natl. Acad. Sci. USA 2008, 105, 8256–
Transmission electron microscopy (TEM)
8
5
261; b) D. Shental-Bechor, Y. Levy, Curr. Opin. Struct. Biol. 2009, 19,
24–533; c) J. L. Price, D. Shental-Bechor, A. Dhar, M. J. Turner, E. T.
For self-assembly experiments, each peptide (100 mm) was allowed
to aggregate for 25 min at 258C in the presence of heparin (1 mm).
For PHF6 inhibition experiments, PHF6 (100 mm) was incubated
with heparin under the same conditions in the presence or ab-
sence of increasing concentrations of the inhibitory peptides
Powers, M. Gruebele, Y. Levy, J. W. Kelly, J. Am. Chem. Soc. 2010, 132,
5359–15367; d) J. L. Price, E. K. Culyba, W. Chen, A. N. Murray, S. R.
Hanson, C. H. Wong, E. T. Powers, J. W. Kelly, Biopolymers 2012, 98, 195–
11; e) J. J. Caramelo, A. J. Parodi, FEBS Lett. 2015, 589, 3379–3387; f) A.
1
2
Tannous, G. B. Pisoni, D. N. Hebert, M. Molinari, Semin. Cell Dev. Biol.
2015, 41, 79–89.
(
molar ratio of 10:1 and 1:1, PHF6/inhibitory peptide). Samples
Chem. Eur. J. 2016, 22, 5945 – 5952
www.chemeurj.org 5951
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[
[
[
5] a) A. Pastore, P. Temussi, J. Phys. Condens. Matter 2012, 24, 244101;
b) T. P. Knowles, M. Vendruscolo, C. M. Dobson, Nat. Rev. Genet. 2014,
[14] a) M. Bucciantini, D. Nosi, M. Forzan, E. Russo, M. Calamai, L. Pieri, L. For-
migli, F. Quercioli, S. Soria, F. Pavone, J. Savistchenko, R. Melki, M. Stefa-
ni, FASEB J. 2012, 26, 818–831; b) Y. Zhang, R. McLaughlin, C. Goodyer,
A. LeBlanc, J. Cell Biol. 2002, 156, 519–529; c) M. Bucciantini, E. Gianno-
ni, F. Chiti, F. Baroni, L. Formigli, J. Zurdo, N. Taddei, G. Ramponi, C. M.
Dobson, M. Stefani, Nature 2002, 416, 507–511; d) M. Andreasen, N.
Lorenzen, D. Otzen, Biochim. Biophys. Acta 2015, 1848, 1897–1907.
[15] A. Belostozky, M.Sc. Thesis, Bar-Ilan University, Ramat-Gan, 2014.
[16] M. H. Horrocks, L. Tosatto, A. J. Dear, G. A. Garcia, M. Iljina, N. Cremades,
M. Dalla Serra, T. P. Knowles, C. M. Dobson, D. Klenerman, Anal. Chem.
2015, 87, 8818–8826.
[17] a) I. García, A. Sanchez-Iglesias, M. Henriksen-Lacey, M. Grzelczak, S. Pe-
nades, L. M. Liz-Marzan, J. Am. Chem. Soc. 2015, 137, 3686–3692;
b) M. C. Rodriguez, M. Cudic, Methods Mol. Biol. 2013, 1081, 107–136;
c) E. Zacco, J. Hutter, J. L. Heier, J. Mortier, P. H. Seeberger, B. Lepenies,
B. Koksch, ACS Chem. Biol. 2015, 10, 2065–2072; d) D. Yuan, J. Shi, X.
Du, N. Zhou, B. Xu, J. Am. Chem. Soc. 2015, 137, 10092–10095; e) Y.
Kamiya, T. Satoh, K. Kato, Curr. Opin. Struct. Biol. 2014, 26, 44–53; f) Y. F.
Liu, Z. Y. Sun, P. G. Chen, Z. H. Huang, Y. Gao, L. Shi, Y. F. Zhao, Y. X.
Chen, Y. M. Li, Bioconjugate Chem. 2015, 26, 1439–1442.
1
5, 384–396.
6] a) J. F. Díaz-Villanueva, R. Diaz-Molina, V. Garcia-Gonzalez, Int. J. Mol. Sci.
015, 16, 17193–17230; b) J. Ma, F. Wang, Essays Biochem. 2014, 56,
81–191.
2
1
7] a) M. Broncel, J. A. Falenski, S. C. Wagner, C. P. Hackenberger, B. Koksch,
Chem. Eur. J. 2010, 16, 7881–7888; b) F. C. Liang, R. P. Chen, C. C. Lin,
K. T. Huang, S. I. Chan, Biochem. Biophys. Res. Commun. 2006, 342, 482–
4
88; c) S. A. Yuzwa, X. Shan, M. S. Macauley, T. Clark, Y. Skorobogatko, K.
Vosseller, D. J. Vocadlo, Nat. Chem. 2012, 8, 393–399; d) D. F. Wyss, J. S.
Choi, J. Li, M. H. Knoppers, K. J. Willis, A. R. Arulanandam, A. Smolyar,
E. L. Reinherz, G. Wagner, Science 1995, 269, 1273–1278; e) S. R.
Hanson, E. K. Culyba, T. L. Hsu, C. H. Wong, J. W. Kelly, E. T. Powers, Proc.
Natl. Acad. Sci. USA 2009, 106, 3131–3136; f) V. Kayser, N. Chennamset-
ty, V. Voynov, K. Forrer, B. Helk, B. L. Trout, Biotechnol. J. 2011, 6, 38–44;
g) A. Taraboulos, M. Rogers, D. R. Borchelt, M. P. McKinley, M. Scott, D.
Serban, S. B. Prusiner, Proc. Natl. Acad. Sci. USA 1990, 87, 8262–8266;
h) P. M. Rudd, A. H. Merry, M. R. Wormald, R. A. Dwek, Curr. Opin. Struct.
Biol. 2002, 12, 578–586; i) C. J. Bosques, B. Imperiali, Proc. Natl. Acad.
Sci. USA 2003, 100, 7593–7598; j) P. M. Rudd, M. R. Wormald, D. R. Wing,
S. B. Prusiner, R. A. Dwek, Biochemistry 2001, 40, 3759–3766; k) C. Lin,
E. H. Chen, L. Y. Lee, R. L. Hsu, F. Y. Luh, L. L. Yang, C. F. Chou, L. D.
Huang, C. C. Lin, R. P. Chen, Carbohydr. Res. 2014, 387, 46–53; l) C. C.
Ho, L. Y. Lee, K. T. Huang, C. C. Lin, M. Y. Ku, C. C. Yang, S. I. Chan, R. L.
Hsu, R. P. Chen, Proteins 2009, 76, 213–225; m) C. H. Yu, T. Si, W. H. Wu,
J. Hu, J. T. Du, Y. F. Zhao, Y. M. Li, Biochem. Biophys. Res. Commun. 2008,
[18] E. K. Culyba, J. L. Price, S. R. Hanson, A. Dhar, C. H. Wong, M. Gruebele,
E. T. Powers, J. W. Kelly, Science 2011, 331, 571–575.
[19] a) F. A. Quiocho, Curr. Top. Microbiol. Immunol. 1988, 139, 135–148;
b) F. A. Quiocho, Annu. Rev. Biochem. 1986, 55, 287–315; c) V. Voynov,
N. Chennamsetty, V. Kayser, B. Helk, K. Forrer, H. Zhang, C. Fritsch, H.
Heine, B. L. Trout, PLoS ONE 2009, 4, e8425; d) V. Spiwok, P. Lipovova, T.
Skalova, E. Vondrackova, J. Dohnalek, J. Hasek, B. Kralova, J. Comput.
Aided Mol. Des. 2006, 19, 887–901.
3
75, 59–62; n) P. Y. Chen, C. C. Lin, Y. T. Chang, S. C. Lin, S. I. Chan, Proc.
Natl. Acad. Sci. USA 2002, 99, 12633–12638; o) R. Roytman, L. Adler-
Abramovich, K. S. Kumar, T. C. Kuan, C. C. Lin, E. Gazit, A. Brik, Org.
Biomol. Chem. 2011, 9, 5755–5761.
[20] a) H. Inouye, D. Sharma, W. J. Goux, D. A. Kirschner, Biophys. J. 2006, 90,
1774–1789; b) K. Sogawa, R. Okuda, Y. In, T. Ishida, T. Taniguchi, K. Min-
oura, K. Tomoo, J. Biochem. 2012, 152, 221–229.
[
8] a) B. Imperiali, K. W. Rickert, Proc. Natl. Acad. Sci. USA 1995, 92, 97–101;
b) C. J. Bosques, S. M. Tschampel, R. J. Woods, B. Imperiali, J. Am. Chem.
Soc. 2004, 126, 8421–8425; c) S. E. O’Conner, B. Imperiali, Chem. Biol.
[21] a) E. Gazit, FASEB J. 2002, 16, 77–83; b) Y. Porat, Y. Mazor, S. Efrat, E.
Gazit, Biochemistry 2004, 43, 14454–14462; c) P. Marek, A. Abedini, B.
Song, M. Kanungo, M. E. Johnson, R. Gupta, W. Zaman, S. S. Wong, D. P.
Raleigh, Biochemistry 2007, 46, 3255–3261; d) S. M. Tracz, A. Abedini, M.
Driscoll, D. P. Raleigh, Biochemistry 2004, 43, 15901–15908; e) O. S.
Makin, E. Atkins, P. Sikorski, J. Johansson, L. C. Serpell, Proc. Natl. Acad.
Sci. USA 2005, 102, 315–320; f) G. G. Tartaglia, A. Cavalli, R. Pellarin, A.
Caflisch, Protein Sci. 2005, 14, 2723–2734; g) S. Jones, J. Manning, N. M.
Kad, S. E. Radford, J. Mol. Biol. 2003, 325, 249–257; h) G. G. Tartaglia, A.
Cavalli, R. Pellarin, A. Caflisch, Protein Sci. 2004, 13, 1939–1941.
[22] L. Motiei, S. Rahimipour, D. A. Thayer, C. H. Wong, M. R. Ghadiri, Chem.
Commun. 2009, 3693–3695.
1998, 5, 427–437; d) S. Dziadek, C. Griesinger, H. Kunz, U. M. Reinscheid,
Chem. Eur. J. 2006, 12, 4981–4993; e) L. Kirnarsky, O. Prakash, S. M.
Vogen, M. Nomoto, M. A. Hollingsworth, S. Sherman, Biochemistry 2000,
3
9, 12076–12082; f) D. M. Coltart, A. K. Royyuru, L. J. Williams, P. W.
Glunz, D. Sames, S. D. Kuduk, J. B. Schwarz, X. T. Chen, S. J. Danishefsky,
D. H. Live, J. Am. Chem. Soc. 2002, 124, 9833–9844; g) F. Corzana, J. H.
Busto, G. Jimenez-Oses, M. Garcia de Luis, J. L. Asensio, J. Jimenez-Bar-
bero, J. M. Peregrina, A. Avenoza, J. Am. Chem. Soc. 2007, 129, 9458–
9
467; h) C. R. Ellis, W. G. Noid, J. Phys. Chem. B 2014, 118, 11462–11469;
i) D. N. Hebert, L. Lamriben, E. T. Powers, J. W. Kelly, Nat. Chem. Biol.
014, 10, 902–910.
[23] E. Atherton, R. C. Sheppard, Solid Phase Peptide Synthesis: A Practical Ap-
proach, Oxfod University Press, Oxford, 1989.
2
[
9] C. Iannuzzi, G. Irace, I. Sirangelo, Front. Mol. Biosci. 2014, DOI: 10.3389/
fmolb.2014.00009.
[24] a) M. von Bergen, S. Barghorn, L. Li, A. Marx, J. Biernat, E. M. Mandelkow,
E. Mandelkow, J. Biol. Chem. 2001, 276, 48165–48174; b) K. Zhao, G. Ip-
polito, L. Wang, V. Price, M. H. Kim, G. Cornwell, S. Fulenchek, G. A.
Breen, W. J. Goux, S. R. D’Mello, J. Neurosci. Res. 2010, 88, 3399–3413;
c) C. L. Moore, M. H. Huang, S. A. Robbennolt, K. R. Voss, B. Combs, T. C.
Gamblin, W. J. Goux, Biochemistry 2011, 50, 10876–10886.
[
10] a) M. von Bergen, P. Friedhoff, J. Biernat, J. Heberle, E. M. Mandelkow, E.
Mandelkow, Proc. Natl. Acad. Sci. USA 2000, 97, 5129–5134; b) B. Bulic,
M. Pickhardt, B. Schmidt, E. M. Mandelkow, H. Waldmann, E. Mandelkow,
Angew. Chem. Int. Ed. 2009, 48, 1740–1752; Angew. Chem. 2009, 121,
1
772–1785.
[
[
[
11] M. Biancalana, S. Koide, Biochim. Biophys. Acta 2010, 1804, 1405–1412.
12] N. J. Greenfield, Nat. prot. 2007, 1, 2876–2890.
13] L. Bousset, L. Pieri, G. Ruiz-Arlandis, J. Gath, P. H. Jensen, B. Habenstein,
K. Madiona, V. Olieric, A. Bockmann, B. H. Meier, R. Melki, Nat. Commun.
Received: December 9, 2015
2013, 4, 2575.
Published online on February 18, 2016
Chem. Eur. J. 2016, 22, 5945 – 5952
www.chemeurj.org
5952
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim