β-sheet breaker peptides containing α,β- dehydrophenylalanine: Synthesis and in vitro activity studies

Article abstract of DOI:10.1002/cplu.201402072

The synthesis and fibrillogenesis-inhibiting activity of the new peptide derivatives 1-6, containing α,β-unsaturated phenylalanines, are reported. These compounds are related to the pentapeptide Ac-LPFFD-NH ₂ (iAβ5p), which was designed by Soto and co-workers and is commonly accepted as a lead compound for fibrillogenesis inhibition. Their activities are determined by Thioflavin T binding assay, far-UV circular dichroism (CD) spectroscopy, and SEM; in addition, their structures in solution are studied through far-UV CD and FTIR spectroscopy. The presence of two α,β-unsaturated phenylalanines increases the fibrillogenesis inhibiting activity significantly in comparison with the lead compound. The interactions between the Aβ_1-40 and the inhibitors using electrospray ionization mass spectrometry are also studied. The analyses prove the presence of noncovalent complexes of Aβ_1-40 with iAβ5p and its derivatives 1-3 with stoichiometries of 1:1 and 2:1, and the results are independent of time and Aβ_1-40/inhibitor ratio.

Full text of DOI:10.1002/cplu.201402072

CHEMPLUSCHEM
FULL PAPERS
DOI: 10.1002/cplu.201402072
b-Sheet Breaker Peptides Containing a,b-
Dehydrophenylalanine: Synthesis and In Vitro Activity
Studies
Cesare Giordano,*[a] Pasqualina Punzi,[b] Clorinda Lori,[c] Roberta Chiaraluce,[c] and
Valerio Consalvi[c]
The synthesis and fibrillogenesis-inhibiting activity of the new
peptide derivatives 1–6, containing a,b-unsaturated phenylala-
nines, are reported. These compounds are related to the pen-
tapeptide Ac-LPFFD-NH2 (iAb5p), which was designed by Soto
and co-workers and is commonly accepted as a lead com-
pound for fibrillogenesis inhibition . Their activities are deter-
mined by Thioflavin T binding assay, far-UV circular dichroism
(CD) spectroscopy , and SEM; in addition, their structures in so-
lution are studied through far-UV CD and FTIR spectroscopy.
The presence of two a,b-unsaturated phenylalanines increases
the fibrillogenesis inhibiting activity significantly in comparison
with the lead compound. The interactions between the Ab1–40
and the inhibitors using electrospray ionization mass spec-
trometry are also studied. The analyses prove the presence of
noncovalent complexes of Ab1–40 with iAb5p and its derivatives
1–3 with stoichiometries of 1:1 and 2:1, and the results are in-
dependent of time and Ab1–40/inhibitor ratio.
Introduction
The conformational change of proteins from their native form,
a-helix or random coil, to b-sheet is an undesirable process
universally proposed as the cause of b-amyloid fibril formation.
The main components of b-amyloid, normally consisting of
39–42 amino acids, are derived from the processing of a large
ubiquitous type 1 transmembrane protein called Amyloid Pre-
cursor Protein. Many serious pathologies, including the most
common form of senile dementia known as Alzheimer’s dis-
ease, are caused by the self-association of amyloid fibrils in
more consistent deposits, usually named senile plaques.[1–5]
In this context, the incorporation of b-breaker residues into
short synthetic peptides that can interact directly with soluble
oligomers as well as insoluble amyloid aggregates, destabiliz-
ing the amyloidogenic conformer and precluding amyloid
polymerization, has been suggested as a promising pharmaco-
logical approach.[6,7] These synthetic structures, called b-sheet
breaker peptides (BSBps), possess a recognition sequence of
the amyloidogenic protein and are capable of binding b-amy-
loid without becoming part of it.[8–11]
Recently, we have found that amino-terminal modifications
of the synthetic pentapeptide Ac-Leu-Pro-Phe-Phe-Asp-NH2
(iAb5p, Scheme 1), previously synthesized and adopted as
a lead BSBp by Soto and co-workers,[12] can give rise to com-
Scheme 1. Structures of iAb5p and its analogue DM-Tau-iAb5p.
[a] Dr. C. Giordano
Istituto di Chimica Biomolecolare del CNR
“Sapienza” Universitꢀ di Roma, P.le A. Moro 5, 00185 Roma (Italy)
Fax: (+39)06-496-43-268
pounds with enhanced activities and metabolic stabilities.[13] In
particular, by replacing the N-acetyl group of iAb5p with N,N-
dimethyl-taurine (DM-Tau), we obtained the esapeptide DM-
Tau-Leu-Pro-Phe-Phe-Asp-NH2 (DM-Tau-iAb5p, Scheme 1),
which showed considerably increased stability toward proteas-
es and a higher antifibrillogenesis activity than iAb5p itself,
E-mail: cesare.giordano@uniroma1.it
[b] Dr. P. Punzi
Dipartimento di Chimica
“Sapienza” Universitꢀ di Roma, P.le A. Moro 5, 00185 Roma (Italy)
[c] Dr. C. Lori, Dr. R. Chiaraluce, Dr. V. Consalvi
Dipartimento di Scienze Biochimiche
[13]
not only on Ab1–40 but also on its 25–35 fragment.[14] Fur-
“Sapienza” Universitꢀ di Roma, P.le A. Moro 5, 00185 Roma (Italy)
thermore, this Tau-containing derivative hindered the toxic
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cplu.201402072.
effect of Ab25–35 upon incubation in a culture of human SH-
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 1036 – 1043 1036

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
SY5Y neuroblastoma cells and prevented the release of super-
oxide anion radicals from BV2 microglial cells.[14]
far-UV circular dichroism (CD) spectroscopy, and SEM. The
structures in solution of all the BSBps were studied by far-UV
CD and FTIR spectroscopy. The noncovalent interactions be-
tween the Ab1–40 and the inhibitors were investigated by using
electrospray ionization mass spectrometry (ESIMS).
Recently, peptides containing a,b-dehydrophenylalanine
(DPhe), an a,b-unsaturated phenylalanine analogue, were
found to act as b-breakers, inhibiting amyloid polymerization if
inserted into short peptides containing the 16–20 segment of
[15]
Ab1-40
.
a,b-Dehydro amino acids are known to induce specif-
ic secondary structures, and upon insertion into peptides de- Results and Discussion
signed to interact with specific regions of Ab, they may act as
fibrillogenesis inhibitors.[16] In the case of DPhe-containing pep-
Peptide synthesis
tides, this behavior is reasonably attributed to the DPhe ten-
dency, similarly to proline and a-aminoisobutyric acid, to pro-
mote the b-turn conformation in brief peptides or to stabilize
the a-helix in longer sequences, as demonstrated exhaustively
through crystallographic studies.[17,18] On the other hand, the
presence of two consecutive DPhe residues in pentapeptides
produces a folded structure, and the sequence is easily accom-
modated in a helical module.[19] In addition, it has been estab-
lished that the presence of DPhe in peptides confers increased
resistance to enzymatic degradation.[20]
The peptide synthesis was performed manually by convention-
al solid-phase chemistry[21] on Rink-amide resin (65 mM scale,
100 mg). The incorporation of DZPhe residues (Scheme 3) was
effected as the dipeptide 8 or the tripeptide 12 (the latter was
useful for the preparation of the two consecutive DZPhe-con-
taining peptides, 3 and 6) activated as their azlactones. These
blocks were obtained through an acetic-anhydride-mediated
azlactonization/dehydration reaction on their C-terminal b-hy-
droxy-dl-phenylalanine [H-dl-Phe(b-OH)-OH]-containing units
7 and 11, respectively.[15,18] The intermediate tripeptide 11 was
in turn achieved by ring-opening of the C-terminal unsaturated
azlactone 10 with the sodium salt of H-dl-Phe(b-OH)-OH (for
details see Supporting Information).
DPhe is characterized by an achiral planar structure contain-
ing extensive conjugation with the p ring electrons. Conse-
quently, the presence of an a-b double bond not only limits
the conformational flexibility of the specific side chain, but
also affects the backbone conformation, restricting the b-aro-
matic substituent to the trans (E) or cis (Z) orientation. In any
case, as always observed in dehydro peptides synthesized via
azlactone, the double bond of the dehydro residues have the
Z configuration of the phenyl group with respect to the amide
group.
Spectroscopic characterization of BSBps
The BSBps were analyzed by far-UV CD (Figure S1B–H, Sup-
porting Information) and FTIR spectroscopy in H2O and in D2O,
respectively (Figure S2A–C).
The far-UV CD spectra of all the BSBps show a band at
around 200 nm, negative in sign except for 4, and very weak
for 6. The spectra of iAb5p, 2, and 5 have a weak positive
band at around 220 nm (n–p*) and a negative band around
200 nm (p–p*), indeed more intense in iAb5p, which can be
assigned to the polyPro II conformation.[22,23] In these peptides,
the presence of the negative band at 200 nm accompanied by
a weak negative band near 230 nm could be correlated to an
unordered conformation.[24] In
Starting from this evidence and with the aim of obtaining
new potential BSBps, we synthesized six new peptides
(Scheme 2) in which a DZPhe moiety substitutes one or both
of the phenylaniline residues of iAb5p (1–3) and DM-Tau-
iAb5p (4–6).
All the peptides were tested in vitro as inhibitors of fibrillo-
genesis on Ab1À40 using the Thioflavin T (ThT) binding assay,
the spectra of 1, 3, 4, and 6,
a broad negative band is present
at around 235 nm, which is
more intense for 1 and 3, and
less intense for iAb5p and 6;
this may resemble that of b-turn
type II,[23] and could be attribut-
ed to the aromatic contributions
to the far-UV CD spectrum.[25–27]
This type of redshifted negative
(n–p*) band, reported for cyclic
peptides, has been correlated to
a class spectrum composite of b-
turn type II and g-turn.[23] A zero
intercept at around 190 nm is
observed only for 1, 3, 4, and 6.
The far-UV CD spectrum of each
compound is different from that
Scheme 2. Structures of the DZPhe-containing analogues 1–6.
of iAb5p, indicating that the var-
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 1036 – 1043 1037

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
shows the FTIR absorbance spec-
tra of 1, 2, and 3 bearing DZPhe
in position 3, position 4, and
both positions 3 and 4, respec-
tively. The three compounds
show a maximum centered at
1620–1630 cmÀ1 and a band at
around 1450 cmÀ1, correspond-
ing to the amide I’ and amide II’
regions,
respectively.
The
amide I’ region presents two sig-
nificant shoulders at 1672 and at
1650 cmÀ1 for all the com-
pounds. In the amide I’ region,
the main peak at 1630–
1620 cmÀ1 can be assigned to
polyPro II,[30] and the 1672 and
1650 cmÀ1 shoulders to b-turns
and unordered loops,[28,31] re-
spectively (Table 1 and Fig-
ure S2A–C). The region of the
antisymmetric stretching vibra-
tion mode nas(COOÀ) of the un-
protonated carboxyl groups
(1595–1550 cmÀ1)
and
the
region corresponding to the
symmetric stretching vibration
mode ns(COOÀ) of the unproto-
nated carboxyl groups (1420–
1390 cmÀ1) show the typical con-
tributions of the Asp resi-
dues.[29,31] The contributions of
Phe at 1498 cmÀ1, Pro at
1465 cmÀ1, and das(CH3) of Leu at
Scheme 3. Synthesis of azlactone intermediates 8, 10, and 12: 1) (AcO)2O/AcONa, RT, 20 h; 2) H-dl-Phe(b-OH)-OH,
1n NaOH, acetone, 458C, 1 h.
iations in the primary sequence cause significant changes in
the secondary structural elements of the peptides in solution.
The compounds were also analyzed by FTIR spectroscopy in
D2O, in the amide I’, in the amide II’, and in the spectral region
in which the side chains of amino acids are known to make
contributions[28,29] (Table 1 and Figure S2A–C). Figure S2A–C
1440 cmÀ1 are also observed.[29,31] Notably, the presence of
DZPhe in positions 4 and 5 in compounds 4 and 5, respectively
(Table 1), is accompanied by a larger amount of the relative
area of the bands assigned to polyPro II and b-turn with re-
spect to that of unordered loops, suggesting a decreased
structural flexibility in these peptides. Compounds 2, 3 and the
DM-Tau-containing 5 show a higher amount of the
relative area of the bands assigned to b-turn, and,
Table 1. Peak position and secondary structure assignment of the amide I’ of com-
pounds 1–6.
similarly to that reported for other BSBps containing
the peptidosulfonamide moiety,[13]
decreased
a
amount of the relative area for the polyPro II band
(Table 1). In the amide II’ region, the spectra of all the
peptides are comparable and do not show any signif-
icant differences (Table 1).
Peak position
Area [%]
4
Assignment
[cmÀ1
]
1
2
3
5
6
1710
1672
1650
1630–1620
1580
1498
1465
1450
1440–1430
1420–1398
6.64
5.42
14.34 13.76
18.16 26.11
7.03
4.64
5.46
11.50
12.96 31.71
13.85
3.90
9.40
5.00
6.20 19.59
7.00 12.86 14.72
1.00
5.62
11.00 31.72 b-turn
6.95 unordered loops
2.90 Asp[a]
26.90 12.42 13.89 14.40 polyPro II+Pro
ThT binding assay
6.35
1.69
4.75
1.50
7.06
3.10
9.70 10.62
8.34
6.70
5.08
7.08 13.68
4.77
11.33 14.72
9.09
5.52 Asp[b]
5.80 Phe
As shown in Figure 1, the relative fluorescence inten-
sity at 484 nm increases as a function of time in the
presence of 18.4 mm Ab1–40. Upon dissolution of Ab1–
5.08 Pro, Leu[c]
5.50 amide II’
9.70 Pro, Leu[d]
8.14
9.00
in the presence of a 12-molar excess of iAb5p or 1,
0.83 20.00
11.88
3.51 12.43 Asp[e]
40
2, 3, 5, or 6, the increase in ThT fluorescence was
slower than that detected for Ab1–40 alone.
[a] n(C=O); [b] nas(COOH); [c] d(CH2); [d] das(CH3); [e] ns(COOH).
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 1036 – 1043 1038

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
Figure 1. In vitro activity of BSBps 1–6 and iAb5p. Fluorescence changes
were monitored continuously at 208C at 484 nm (450 nm excitation wave-
length) in a final volume of 1 mL containing 10 mm ThT, 18.4 mm Ab1–40 in
20 mm Tris-HCl, pH 7.4, alone (0) or in the presence of a 12-molar excess of
1–6 or iAb5p.
Figure 2. Far-UV CD spectral changes of Ab1–40. Far-UV CD spectra of
18.4 mm Ab1–40 in 20 mm Tris-HCl, pH 7.4, were monitored at 208C, at in-
creasing times from 0 to 50 h.
For Ab1–40 in the presence of 3 and 6, carrying DZPhe in
both positions, the amplitude of the fluorescence changes
measured for 50 h decreased dramatically, and the increase in
ThT fluorescence intensity was significantly slower than that
observed for Ab1–40 alone. In the presence of 2 and 5, the in-
crease in fluorescence intensity of Ab1–40 was slower than that
observed with Ab1–40 alone, and in the presence of 2, the
changes were comparable to those observed in the presence
of iAb5p. In the presence of 4, the fluorescence changes of
Ab1–40 were comparable to those observed for Ab1–40 alone. No-
tably, the amplitudes of the fluorescence changes in the pres-
ence of 1, 2, 4, and 5 were comparable to that observed for
Ab1–40 alone. The kinetics of Ab1–40 aggregation are character-
ized by a sigmoidal profile in which three phases can be distin-
guished: a lag phase, a growth phase, and a final plateau
phase. All the compounds except 4 lead to an increase in the
duration of the lag phase. The length of the lag phase has
been used widely as a diagnostic indicator of the inhibition of
Ab1–40 aggregation, and can be utilized to evaluate the activi-
ties of BSBps as aggregation inhibitors. The lag phase in-
creased from 500 to 760 min upon incubation of Ab1–40 with
1 or iAb5p, and to 920, 1000, 1100, and 1120 min with 2, 5, 6,
and 3, respectively (Figure 1).
containing structural region,[25] and an increase in molar ellip-
ticity at around 215 nm, characteristic of the b-sheet structure.
Two main spectral components were identified after singular
value decomposition (SVD) of the far-UV CD spectra of Ab1–40
alone monitored for 50 h. The changes in the first spectral
component can be attributed mainly to the disappearance of
the turns-containing structure at around 198 nm (Figure S3A),
which is accompanied by the appearance of the b-sheet struc-
ture (Figure S3B), as revealed by the increase in negative
molar ellipticity at 215 nm and in the positive ellipticity of the
second spectral component at around 195 nm.
The in vitro inhibitory activities of 1–6 and iAb5p were
tested by monitoring their efficacy in preventing and/or inter-
fering with the secondary structure conformational transition
of Ab1–40 prior to the formation of fibrils and visible aggregates.
The far-UV CD spectra of Ab1–40 monitored in the presence of
1–6 or iAb5p from 0 to 50 h and analyzed by SVD revealed
two main spectral components, centered at around 198 and
215 nm, respectively, similarly to those observed for Ab1–40
alone. The changes at 198 and 215 nm of the first (Figure S4A)
and second (Figure S4B) spectral components, respectively, are
significantly reduced in amplitude in the presence of all the
BSBps. A plot of the changes in molar ellipticity at 198 nm of
the first spectral component of Ab1–40 alone as a function of
time shows a sigmoidal profile with an amplitude of about
104 degcm2 dmolÀ1 (Figure S4A). In the presence of all the
BSBps, the amplitude of the change at 198 nm is reduced sig-
nificantly with respect to that observed for Ab1À40 alone, and
decreases as the initial molar ellipticity at 198 nm decreases
(Figure S4A). The changes at 215 nm of the second spectral
component in the presence of all the compounds appear to
be negligible in comparison to those observed for Ab1–40 alone
(Figure S4B). These results suggest an interaction between
Ab1–40 and 1–6 or iAb5p, which is further supported by the sig-
nificant difference between the spectra of Ab1–40 in the pres-
ence of the BSBps (Figure S1B–H, upper panels) and the calcu-
Far-UV CD spectroscopy
Conformational studies on the lyophilized synthetic Ab1–40
were performed by monitoring the far-UV CD spectral changes
as a function of time in the absence or presence of a 12-molar
excess of 1–6, in parallel with the ThT binding assay and over
the same time duration. Typically, lyophilized Ab1–40 in Tris-HCl
(20 mm, pH 7.4, final concentration 18.4 mm) exhibits the far-UV
CD spectral changes reported in Figure 2.
Over a period of 50 h, the spectral changes observed can be
summarized as two main events: an overall decrease in molar
ellipticity, mainly at around 198 nm, which is a typical turns-
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 1036 – 1043 1039

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
lated sum of the individual spectra of free Ab1–40 and those of
1–6 or iAb5p alone (Figure S1B–H, lower panels). All the spec-
tra of the BSBps alone show a broad band at around 235 nm,
which is more intense for 1 and 3, and less intense for iAb5p
and 6 (Figure S1B–H). This band, which is absent in the spec-
trum of Ab1–40, disappears upon mixing of Ab1–40 with 1 and 3,
and 50 h after mixing of Ab1–40 with 2, 4, 5, and 6, but remains
unaltered upon mixing of Ab1–40 with iAb5p (Figure S1B–H,
upper panels).
Mass spectrometry
With the aim of detecting the presence of the Ab1–40–inhibitor
noncovalent complex (Ab-I), ESI-MS was utilized (for spectra
see Supporting Information).[34] This method can be used to
evaluate the ability of the BSBp to bind Ab1–40 as well as the
specificity of the Ab-I interaction, using small amounts of
sample to avoid random interactions caused by higher concen-
trations in the gas phase. Furthermore, under suitable condi-
tions, this method allows the determination of the stoichiome-
try of the noncovalently bound complex, because this can be
transferred into the gas phase without breaking.
In these studies, very mild ionization conditions were used,
and the most useful parameters reported for these analyses
were adopted in terms of pH, organic modifier content, and
time.[35]
Scanning electron microscopy
Figure 3 shows SEM images of Ab1–40 after incubation at 208C
in the absence (A, C) and presence of the most active com-
pounds, 3 (B) and 6 (D). In the absence of the inhibitor, the
Ab1–40 sample shows the formation of protofibrils less than
1 mm in length[32,33] (Figure 3A,C).
First, the mass spectrum of Ab1–40 was recorded with a final
concentration of 50 mm at a pH value of around 3.5, dissolved
in a mixture of 1.0 mm acetate buffer (containing 0.5% acetic
acid) and methanol (20% v/v).
In the presence of a BSBp (Figure 3B,D), no protofibrils, or
any kind of fibrillar structure or structure morphologically re-
The analyses in positive-ion
mode confirmed the expected
peak distribution at m/z 866.6 (+
5), 1083.1 (+4), 1443.8 (+3), and
2166.0 (+2) for multiply charged
ions of Ab1–40, corresponding to
a theoretical molecular mass of
4330.0 Da. After the addition of
BSBp to freshly prepared Ab1–40
solutions in equimolar ratios, the
resulting mass spectra showed
the same distribution of the Ab1–
peaks. In addition, for iAb5p
40
and its derivatives 1–3, all the
spectra showed new ion signals
corresponding to the multiply
charged Ab-I (1:1 association
ratio) at charge states of +4 and
+3, and peaks probably corre-
sponding to the multiply
charged (Ab)2-I (2:1 association
ratio) at charge states of +5,
+6, and +7.
In the case of DM-Tau-iAb5p
and its analogues 4–6, no peaks
attributable to the multiply
charged Ab-I interaction were
Figure 3. Effect on Ab1–40 aggregation. SEM images of Ab1–40 (18.4 mm in 20 mm Tris/HCl, pH 7.4) incubated at
208C in the absence (A, C) and presence of a 12-molar excess of compound 3 (B) and 6 (D). 20 mL aliquots were
withdrawn after 50 h, frozen, lyophilized, and gold-coated by sputtering prior to SEM analyses.
detected (data not shown). The
absence of such an interaction
in these cases may be ascribed
to the protonation of the termi-
sembling amyloid aggregates, were observed. The few struc-
tures present are presumably formed by the BSBps, as suggest-
ed by the SEM images of the inhibitors alone (data not
shown).
nal dimethylamino group of the inhibitor owing to the acidic
conditions required by the ESI process, which may prevent
complex formation.
After incubation for 24 h at room temperature under the
same conditions, no substantial differences were observed in
the spectra of any of the samples. Analogously, no evident dif-
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 1036 – 1043 1040

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
ferences were noted if the spectra were recorded in the pres-
ence of a tenfold excess of BSBp (data not shown).
Conclusion
As a parameter directly correlated with the affinity of BSBp
toward Ab1–40, the percentage of bound amyloid was calculat-
ed by applying the following equation: bound Ab1–40 (%)=
100([1:1]+2[2:1])/([Ab1–40]+[1:1]+2[2:1]). Here, [Ab1–40], [1:1],
and [2:1] represent the concentrations of the different species
in solution at equilibrium, and are directly proportional to the
relative intensities of the corresponding peaks found in the
mass spectra (assuming that the response factors of Ab1–40
alone and of Ab-I are the same, as reported previously[34]).
Table 2 shows the Ab-I complexes with 1:1 and 2:1 stoichi-
ometry, as well as the percentage of bound Ab1–40. In general,
for compounds 1–3, the bound Ab1–40 is similar to iAb5p. The
apparent discrepancy with the results obtained by ThT studies
(Figure 1) may be related to the fact that inhibitor binding to
Ab1–40 and inhibition of fibril formation may follow different
pathways, so the two events may not be compared direct-
ly.[36,37]
This report highlights that the presence of DZPhe in iAb5p
and its DM-Tau-containing derivatives significantly increases
the inhibition of Ab1–40 aggregation, and demonstrates the effi-
cacy of these compounds in preventing the secondary struc-
ture conformational transition of Ab1–40 prior to aggregation.
The efficacy seems to be related to the decrease in unordered
loops of the BSBp in solution, probably owing to the constrain-
ed conformation induced by the presence of two consecutive
DZPhe moieties, which generally induce b-bend conformations
in small linear peptides. The analyses with ESI-MS prove the
noncovalent interaction of Ab1–40 with iAb5p or its derivatives
1–3, and the results are dependent neither on time nor Ab1–40/I
ratio. All these findings may be useful for the development of
new molecules containing a,b-dehydroamino acids for Alz-
heimer’s disease therapy, and eventually help prevent the fi-
brillogenesis that follows aggregation in protein misfolding
processes and related pathologies.
Table 2. Percentage of bound Ab1–40
.
Experimental Section
BSBp
Bound Ab1–40 [%]
1:1/2:1 ratio
ThT binding assay
iAb5p
14.57
10.14
11.74
9.57
85/15
83/17
84/16
81/19
The assembly of Ab1–40 was studied by using a ThT binding
assay.[38] The ThT fluorescence intensity was followed continuously
in a PerkinElmer LS50B spectrofluorimeter, thermostated at 208C,
with excitation and emission wavelengths set at 450 and 484 nm,
respectively. Fluorescence changes were monitored in a 1 mL cuv-
ette under continuous stirring at 484 nm every 6 s, with an integra-
tion time of 5 s over at least 13200 s. A ThT stock solution (1 mm
in water) was prepared, aliquoted, and stored at À208C. Before
each experiment, an aliquot of ThT stock solution was thawed at
48C, shielded from light. A solution of lyophilized Ab1–40 (1 mL
final volume, 18.4 mm, 20 mm Tris-HCl, pH 7.4), alone or in the pres-
ence of a 12-molar excess of 1–6 or iAb5p, was transferred to
a 1 mL cuvette, and the ThT stock solution (10 mL) was added.
Dead time for the measurements of fluorescence changes did not
exceed 12 s. In the absence of Ab1–40, the fluorescence intensity at
484 nm of ThT alone was identical to that measured in the pres-
ence of any of the BSBps tested, and did not change over a period
of 50 h.
1
2
3
Upon incubation in the presence of a tenfold excess of
iAb5p, a threefold increase in bound Ab1–40 was found, where-
as after incubation for 24 h at room temperature with a 1:1
ratio of Ab1–40/iAb5p, the Ab2-I association signal disappeared,
and only the Ab-I noncovalent complex was present in addi-
tion to a constant amount of bound Ab1–40. These results con-
firm the immediate Ab-I complex formation, although a lower
Ab1–40/I ratio was used than for the spectroscopic studies, and
demonstrate the good affinity for Ab1–40 of the DZPhe-contain-
ing inhibitors in both 1:1 and 2:1 stoichiometry. These findings
may be related to the higher planarity of the molecule confer-
red by the DZPhe moiety. The increased inhibition of fibrillo-
genesis induced by compound 3, which contains two consecu-
tive DZPhe moieties, may only be in apparent contrast with
the fact that the percentage of Ab1–40 bound to compound 3 is
slightly lower than that of Ab1–40 bound to iAb5p (Table 2).
Indeed, it is tempting to speculate that compound 3, charac-
terized by reduced conformational flexibility as suggested by
the significant decrease in unordered loops (Table 1), upon
binding to Ab1–40 induces this latter to assume a conformation
less prone to fibril formation. Interestingly, the disappearance
of the (Ab)2-I complex and the stability of bound Ab1–40 after
incubation with iAb5p for 24 h suggest the dissociation of one
molecule of amyloid from (Ab)2-I in favor of the formation of
the Ab-I complex. The presence of Ab-I complexes with 1:1
and 2:1 stoichiometry suggests that the inhibitors may interact
Spectroscopy
Far-UV (190–250 nm) CD measurements were performed with
a Jasco J-720 spectropolarimeter in a quartz cuvette (0.1 cm path
length) thermostated at 208C. The results are expressed as the
mean residue ellipticity [V] assuming a mean residue weight of
110 per amino acid residue. Compounds 1–6 were dissolved in
ultra-high-quality water, lyophilized, and stored at À208C. For far-
UV CD spectral measurements, each lyophilized compound was
dissolved in ultra-high-quality water. Aliquots of lyophilized Ab1–40
were dissolved in Tris-HCl (20 mm, pH 7.4, 18.4 mm final concentra-
tion), in the absence or presence of BSBp (12-molar excess). The
far-UV CD spectral changes occurring upon fibril formation were
monitored at 208C, at increasing times from 0 to 50 h. The far-UV
CD spectra of iAb5p and compounds 1–6 monitored from 0 to
50 h did not show any change as a function of time. FTIR spectra
in the amide I’ and amide II’ regions were recorded on a Nicolet
Magna 760 spectrometer (Thermo Nicolet, Madison, WI, U.S.A.)
in more than one way with the hydrophobic region[8] of Ab1–40
;
however, the mode(s) of binding in these complexes cannot
be established.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 1036 – 1043 1041

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
equipped with a liquid-nitrogen-cooled mercury-cadmium telluride
solid-state detector. FTIR transmission spectra of the peptides
(about 20 mgmLÀ1, D2O) were measured at 208C in a CaF2 cell
with a 0.015 mm spacer (Spectra Tech, Madison, WI, U.S.A.). For ex-
change of amide protons with deuterons, each lyophilized com-
pound (1–6) was freeze-dried and dissolved in D2O four times at
208C before measurements. A total of 512 interferograms at a reso-
lution of 2 cmÀ1 were collected for each spectrum, with Mertz
apodization and two levels of zero filling. The sample chamber of
the spectrometer was purged continuously with dry air to avoid
the interference of water vapor on the bands of interest. The back-
ground spectra were collected immediately before the sample
measurements and under the same conditions, with the cell filled
with everything except the peptide sample. Water-vapor spectra
were collected by reduction of the dry-air purge of the clean cell.
The analysis of raw spectra was performed with GRAMS Suite 9.1
(Thermo Fisher Scientific, Philadelphia, PA). The water-vapor contri-
bution was subtracted from each of the sample spectra. The essen-
tially featureless region between 1700 and 1800 cmÀ1 (in which no
peptide bands are present) indicated that water-vapor components
are not responsible for the observed bands in the amide I’ region.
The individual components in the amide region were identified by
second derivative and Fourier self-deconvolution of the raw spec-
trum with GRAMS. Curve-fitting of the raw spectrum with a mixed
Gaussian–Lorentzian function was then performed to determine
the peak positions and parameters of each individual component.
The assignments of the component bands to secondary structure
elements were based on the literature,[28,29,31] and the area under
each peak was used to calculate the percentage of secondary
structure components (Table 1).
Data analysis
Far-UV CD spectra of Ab1–40 in the presence or absence of the syn-
thetic BSBps were analyzed by SVD[39–41] using MATLAB (Math-
Works, South Natick, MA, U.S.A.) software. SVD is useful for finding
the number of independent components in a set of spectra and
for removing high-frequency noise and low-frequency random
error. CD spectra at increasing incubation times in the 190–250 nm
region (0.5 nm sampling interval) were placed in a rectangular
matrix A of n columns, with one column for each spectrum collect-
ed at the selected incubation time, ranging from 0 to 50 h. The A
matrix was decomposed by SVD into the product of three matri-
ces: A=USVT, in which U and V were orthogonal matrices and S
was a diagonal matrix. The columns of matrix U contained the
basis spectra and the columns of matrix V contained the time de-
pendence of each basis spectrum. Both U and V columns were ar-
ranged in terms of decreasing order of the relative weight of infor-
mation, as indicated by the magnitude of the singular values in S.
The diagonal matrix S contained the singular values that quantified
the relative importance of each vector in U and V.
Acknowledgements
This study was supported by funds provided by Sapienza Univer-
sitꢀ di Roma.
Keywords: Alzheimer’s disease · fibrillogenesis inhibitors ·
fluorescence assay · scanning electron microscopy · b-sheet
breakers
[1] K. P. Kepp, Chem. Rev. 2012, 112, 5193–5239.
Mass spectrometry
[2] L. C. Serpell, Biochim. Biophys. Acta 2000, 1502, 16–30.
[3] C. Soto, M. E. Castano, B. Frangione, N. C. Inestrosa, J. Biol. Chem. 1995,
270, 3063–3067.
All the experiments were performed on a Q-TOF MICRO spectrom-
eter (Micromass, now Waters, Manchester, UK) equipped with an
ESI source, in positive-ion mode, adjusting the operating parame-
ters as follows: rate of sample infusion into the mass spectrometer:
20 mLminÀ1; capillary voltage: 2.8 kV; source temperature: 808C;
source pressure: 1.30 mbar; cone voltage: 35 V; and collision
energy: 10 V. Full-scan mass spectra were recorded in the m/z
range between 800 and 2500, with 100 acquisitions per spectrum.
The data were analyzed by using the MassLynx software developed
by Waters.
[4] D. J. Selkoe, Nature 1999, 399, A23–A31.
[5] Y. Takahashi, A. Ueno, H. Mihara, ChemBioChem 2002, 3, 637–642.
[6] J. Liu, W. Wang, Q. Zhang, S. Zhang, Z. Yuan, Biomacromolecules 2014,
15, 931–939.
[7] C. Adessi, C. Soto, Drug. Dev. Res. 2002, 56, 184–193.
[8] V. Minicozzi, R. Chiaraluce, V. Consalvi, C. Giordano, C. Narcisi, P. Punzi,
G. C. Rossi, S. Morante, J. Biol. Chem. 2014, DOI: 10.1074/
jbc.M113.537472.
[9] D. Gordon, S. C. Meredith, Biochemistry 2003, 42, 475–485.
[10] L. D. Estrada, C. Soto, Curr. Pharm. Des. 2006, 12, 2557–2567.
[11] C. Soto, E. M. Sigurdsson, L. Morelli, R. A. Kumar, E. M. CastaÇo, B. Fran-
gione, Nat. Med. 1998, 4, 822–826.
[12] C. Adessi, M. J. Frossard, C. Boissard, S. Fraga, S. Bieler, T. Rucole, F. Vil-
bois, S. M. Robinson, M. Mutters, W. A. Banks, C. Soto, J. Biol. Chem.
2003, 278, 13905–13911.
[13] C. Giordano, A. Masi, A. Pizzini, A. Sansone, V. Consalvi, R. Chiaraluce, G.
Lucente, Eur. J. Med. Chem. 2009, 44, 179–189.
[14] C. Giordano, A. Sansone, A. Masi, A. Masci, L. Mosca, R. Chiaraluce, A.
Pasquo, V. Consalvi, Chem. Biol. Drug Des. 2012, 79, 30–37.
[15] M. Gupta, R. Acharya, A. Mishra, S. Ramakumar, A. Faizan, V. S. Chauhan,
ChemBioChem 2008, 9, 1375–1378.
The samples were dissolved in acetate buffer (1.0 mm, containing
0.5% acetic acid, pH 3.5, 50 mm Ab1–40 final concentration). After
the binding equilibrium was established, methanol (20% v/v) was
added to the mixture just before injection to obtain a stable elec-
trospray signal. The final volume of the sample was 200 mL. As a ref-
erence, a mixture containing only Ab1–40 was prepared. For the
minimization of random errors, each experiment was repeated at
least three times under the same experimental conditions.
[16] V. Rangachari, Z. S. Davey, B. Healy, B. D. Moore, L. K. Sonoda, B. Cusack,
G. M. Maharvi, A. H. Fauq, T. L. Rosenberry, Biopolymers 2009, 91, 456–
465.
[17] K. K. Bhandary, V. S. Chauhan, Biopolymers 1993, 33, 209–217.
[18] S. N. Mitra, S. Dey, S. Karthikeian, T. P. Singh, Biopolymers 1997, 41, 97–
105.
[19] K. R. Rajashankar, S. Ramakumar, R. M. Jain, V. S. Chauhan, Biopolymers
1997, 42, 373–382.
[20] Y. Shimohigashi, M. L. English, C. H. Stammer, T. Costa, Biochem. Biophys.
Res. Commun. 1982, 104, 583–590.
Scanning electron microscopy
The morphologies of Ab1–40 were investigated with a scanning elec-
tron microscope (LEO 1450VP) with a tungsten filament as an elec-
tron emitter, with an acceleration potential of 20 keV and a resolu-
tion of 4 nm. For the SEM study, samples of Ab1–40 incubated in the
absence and presence of inhibitors under the same conditions as
those described for amyloid detection by far-UV CD spectroscopy
were withdrawn after 50 h, frozen, lyophilized, and gold-coated.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 1036 – 1043 1042

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
[21] E. Atherton, R. C. Sheppard, Solid Phase Synthesis–A Practical Approach,
IRL Press, Oxford, 1989.
[32] C. Goldsbury, P. Frey, V. Olivieri, U. Aebi, S. A. Mꢁller, J. Mol. Biol. 2005,
352, 282–298.
[33] R. A. Moore, S. F. Hayes, E. R. Fischer, S. A. Priola, Biochemistry 2007, 46,
7079–7087.
[34] J. M. Daniel, S. D. Friess, S. Rajagopalan, S. Wendt, R. Zenobi, Int. J. Mass
Spectrom. 2002, 216, 1–27.
[22] K. Ma, L. Kann, K. Wang, Biochemistry 2001, 40, 3427–3438.
[23] A. Perczel, M. Hollosi in Circular Dichroism and the Conformational Anal-
ysis of Biomolecules (Eds.: G. D. Fasman), Plenum Press, New York, 1996,
pp. 285–380.
[35] F. N. Bazoti, A. Tsarbopoulos, K. E. Markides, J. Bergquist, J. Mass Spec-
trom. 2005, 40, 182–192.
[36] M. Bartolini, C. Bertucci, M. L. Bolognesi, A. Cavalli, C. Melchiorre, V. An-
drisano, ChemBioChem 2007, 8, 2152–2161.
[37] M. Necula, R. Kayed, S. Milton, C. G. Glabe, J. Biol. Chem. 2007, 282,
10311–10324.
[24] S. Y. Venyaminov, J. T. Yang in Circular Dichroism and the Conformational
Analysis of Biomolecules (Eds.: G. D. Fasman), Plenum Press, New York,
1996, pp. 69–107.
[25] R. W. Woody, A. K. Dunker in Circular Dichroism and the Conformational
Analysis of Biomolecules (Eds.: G. D. Fasman), Plenum Press, New York,
1996, pp. 109–157.
[38] H. Naiki, K. Higuchi, M. Hosokawa, T. Takeda, Anal. Biochem. 1989, 177,
244–249.
[39] E. R. Henry, J. Hofrichter, Methods Enzymol. 1992, 210, 129–192.
[40] W. C. Johnson Jr., Methods Enzymol. 1992, 210, 426–447.
[41] R. M. Ionescu, V. F. Smith, J. C. O’Neill Jr., C. R. Matthews, Biochemistry
2000, 39, 9540–9550.
[26] B. Bochicchio, A. Pepe, A. M. Tamburro, Chirality 2005, 17, 364–372.
[27] S. Vuilleumier, J. Sancho, R. Loewenthal, A. R. Fersht, Biochemistry 1993,
32, 10303–10313.
[28] M. Jackson, H. H. Mantsch, Crit. Rev. Biochem. Mol. Biol. 1995, 30, 95–
120.
[29] A. Barth, Prog. Biophys. Mol. Biol. 2000, 74, 141–173.
[30] Y. Pilpel, O. Bogin, V. Brumfeld, Z. Reich, Biochemistry 2003, 42, 3519–
3526.
Received: March 19, 2014
[31] A. Barth, C. Q. Zscherp, Q. Rev. Biophys. 2002, 35, 369–430.
Published online on May 6, 2014
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 1036 – 1043 1043