The structural basis for optimal performance of oligothiophene-based fluorescent amyloid ligands: Conformational flexibility is essential for spectral assignment of a diversity of protein aggregates

Article abstract of DOI:10.1002/chem.201301463

Protein misfolding diseases are characterized by deposition of protein aggregates, and optical ligands for molecular characterization of these disease-associated structures are important for understanding their potential role in the pathogenesis of the disease. Luminescent conjugated oligothiophenes (LCOs) have proven useful for optical identification of a broader subset of disease-associated protein aggregates than conventional ligands, such as thioflavin T and Congo red. Herein, the molecular requirements for achieving LCOs able to detect nonthioflavinophilic Aβ aggregates or non-congophilic prion aggregates, as well as spectrally discriminate Aβ and tau aggregates, were investigated. An anionic pentameric LCO was subjected to chemical engineering by: 1) replacing thiophene units with selenophene or phenylene moieties, or 2) alternating the anionic substituents along the thiophene backbone. In addition, two asymmetric tetrameric ligands were generated. Overall, the results from this study identified conformational freedom and extended conjugation of the conjugated backbone as crucial determinants for obtaining superior thiophene-based optical ligands for sensitive detection and spectral assignment of disease-associated protein aggregates. Copyright

Full text of DOI:10.1002/chem.201301463

FULL PAPER
DOI: 10.1002/chem.201301463
The Structural Basis for Optimal Performance of Oligothiophene-Based
Fluorescent Amyloid Ligands: Conformational Flexibility is Essential for
Spectral Assignment of a Diversity of Protein Aggregates
Therꢀse Klingstedt,[a] Hamid Shirani,[a] K. O. Andreas ꢁslund,[a] Nigel J. Cairns,[b]
Christina J. Sigurdson,[c] Michel Goedert,[d] and K. Peter R. Nilsson*[a]
Abstract: Protein misfolding diseases
are characterized by deposition of pro-
tein aggregates, and optical ligands for
molecular characterization of these dis-
ease-associated structures are impor-
tant for understanding their potential
role in the pathogenesis of the disease.
Luminescent conjugated oligothio-
phenes (LCOs) have proven useful for
achieving LCOs able to detect non-
AHCTUNGTRENNtGUN hioflavinophilic Ab aggregates or non-
phenylene moieties, or 2) alternating
the anionic substituents along the thio-
phene backbone. In addition, two
asymmetric tetrameric ligands were
generated. Overall, the results from
this study identified conformational
freedom and extended conjugation of
the conjugated backbone as crucial de-
terminants for obtaining superior thio-
phene-based optical ligands for sensi-
tive detection and spectral assignment
of disease-associated protein aggre-
gates.
congophilic prion aggregates, as well as
spectrally discriminate Ab and tau ag-
gregates, were investigated. An anionic
pentameric LCO was subjected to
chemical engineering by: 1) replacing
thiophene units with selenophene or
optical identification of
a broader
Keywords: Alzheimerꢀs disease
·
subset of disease-associated protein ag-
gregates than conventional ligands,
such as thioflavin T and Congo red.
Herein, the molecular requirements for
fluorescent probes luminescent
·
conjugated oligothiophenes · micro-
scopy · protein folding
Introduction
cally defined LCOs are able to pass the blood–brain barrier
and selectively stain protein aggregates of amyloid-b (Ab)[3]
and tau[11] in transgenic mouse models with Alzheimerꢀs dis-
ease (AD)-like pathology.
Accumulation of protein deposits, so-called amyloid, is the
histopathological hallmark of several devastating diseas-
es,[1,2] and the development of sensitive optical ligands to
detect and characterize these disease-associated structures is
of great interest. In this regard, luminescent conjugated
oligo- and polythiophenes (LCOs and LCPs) have been
used to study protein aggregates in a variety of recombinant,
cellular and tissue-based platforms.[3–10] In addition, chemi-
The major advantages of thiophene-based probes over
conventional amyloid ligands, such as Congo red, thiofla-
vin T (ThT) and their derivatives, are their ability to identify
a broader subset of disease-associated protein aggregates
and their utilization for spectroscopic assignment of hetero-
geneous populations of deposits or distinct protein aggre-
gates. The latter property is achieved due to a direct connec-
tion between the conformational geometry of the oligothio-
phene backbone and the emitted light from the dye.[12,13]
This conformation-induced optical read-out has been utiliz-
ed as an indirect measurement of protein aggregate mor-
phology[3–11] as well as for sensing of other biological proc-
esses.[14–17] LCO ligands have proved useful for gaining fun-
damental novel insights regarding the structural heterogene-
ity reported for infectious prions as well as Ab aggre-
gates,[7,8] and optical identification of protein aggregates
with distinct topologies are highly important. For instance, it
was recently proposed that the pathogenesis of both
AD[18,19] and prion[20] diseases might involve two distinct
phases, an initial build-up of amyloid followed by a toxic
pathway, and that these phases entail different aggregate
topologies. Moreover, several lines of evidence indicate that
aggregated oligomeric assemblies preceding the formation
of mature amyloid fibrils are the culprits of many protein
[a] T. Klingstedt,+ Dr. H. Shirani,+ Dr. K. O. A. ꢁslund,
Dr. K. P. R. Nilsson
Department of Chemistry, Linkçping University
581 83 Linkçping (Sweden)
E-mail: petni@ifm.liu.se
[b] Dr. N. J. Cairns
Department of Neurology, Alzheimerꢀs Disease Research Center
Washington University, St. Louis, Missouri 63110 (USA)
[c] Dr. C. J. Sigurdson
Department of Pathology, University of California
San Diego, La Jolla, California 92093-0612 (USA)
[d] Prof. M. Goedert
MRC Laboratory of Molecular Biology
Hills Road, Cambridge CB2 0QH (UK)
[+] These authors contributed equally to this work.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201301463.
Chem. Eur. J. 2013, 00, 0 – 0
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
1
&
ÞÞ
These are not the final page numbers!

aggregation diseases[21–25] and optical ligands with the capaci-
ty to detect these species are therefore highly sought after.
One of the criteria in the definition of amyloid is that the
protein aggregate needs to exhibit both affinity for Congo
red and green birefringence when viewed by polarization
microscopy.[26] However, it was recently shown that Congo
red failed to identify an artificial amyloid structurally indis-
tinguishable from HET-s amyloid[27] and several studies have
demonstrated that Congo red and thioflavins only detect a
limited amount of disease-associated prion aggregates,[8] Ab
assemblies[3,5] as well as protein inclusion bodies.[28] These
reports underline the importance of identifying and under-
standing the variety of amyloid topologies that are identified
with distinct ligands, and highlight the need for developing
versatile amyloid ligands. By a chemical engineering ap-
proach, involving the screening of a library of structurally
related compounds towards defined amyloid targets, insights
regarding the optimal design of such ligands can be
achieved. When comparing the optical performance of a li-
brary of anionic LCOs, it was recently shown that a back-
bone consisting of at least five thiophene units with carboxyl
groups extending the conjugation length was required to
spectrally discriminate Ab plaques and tau neurofibrillary
tangles (NFTs) in AD brain tissue sections.[5] Moreover,
LCOs having at least five thiophene units allowed optical
detection of early in vitro formed aggregated species of Ab,
bovine insulin or lysozyme that went undetected with
ThT.[4,5] Hence, the performance of LCOs as amyloid ligands
is highly dependent on the
have a large impact on its performance, and the findings
might further aid in the understanding of how to design op-
timal amyloid binding ligands.
Results and Discussion
Synthesis and optical characterization of LCOs and thio-
phene/selenophene co-oligomers: The conformation-induced
optical properties of LCO-based amyloid binding ligands
have previously been shown to be highly dependent on the
length of the thiophene backbone and the positioning of car-
boxyl moieties along the conjugated backbone.[3,5] To further
investigate the chemical basis for obtaining the conforma-
tion-induced optical properties that distinguish LCOs from
other conventional amyloid ligands, a novel set of ligands
derived from the previously reported LCOs p-FTAA and p-
HTAA were synthesized. In total, the library contained two
tetramers and six pentamers, including p-FTAA and p-
HTAA (Figure 1).
It has been shown, both theoretically and experimentally,
that oligoselenophenes have a more quinoid character[29]
and a lower band gap,[30–32] and are more difficult to twist
than oligothiophenes.[29,33] Thus, distinct thiophene units
were replaced by a selenophene to render the thiophene/se-
lenophene co-oligomers p-FTAA-Se and p-HTAA-Se
(Figure 1). The synthesis of p-FTAA-Se and p-HTAA-Se is
depicted in Scheme 1 and Scheme 2, respectively. The prepa-
length of the thiophene back-
bone as well as the positioning
of anionic side chains along the
backbone.
Herein we report the synthe-
sis of a set of novel thiophene-
based amyloid ligands designed
with minor variations in the
chemical structure to restrict or
increase the conformational
flexibility of the conjugated
backbone. The effects of the
structural modifications were
investigated by comparing the
optical performances of these
Scheme 1. Synthesis of p-FTAA-Se. Reagents and conditions: a) 1,4-dioxane/MeOH, PEPPSITM-IPr, K2CO3,
708C, 20 min; b) NBS, DMF, 08C to RT, 16 h; c) 1,4-dioxane/MeOH, PEPPSITM-IPr, K2CO3, 708C, 20 min;
d) NaOH (1m), 1,4-dioxane, 508C, 16 h.
ligands in the detection of a va-
riety of protein aggregates, in-
cluding in vitro formed aggre-
gated species of Ab1–42, Ab
deposits and NFTs in AD brain
tissue sections, as well as NFTs and non-congophilic prion
aggregates in transgenic mice. By replacing distinct thio-
phene moieties with selenophene or phenylene units and by
changing the structural symmetry of the thiophene back-
bone, further insights regarding the conformation-induced
optical performance of this class of amyloid ligands were
achieved. The obtained results clearly demonstrate how
subtle changes in the chemical composition of the ligand
ration of pentamers was carried out by following a similar
protocol to that described in our previous work,[3,5] which
takes advantage of Suzuki cross-coupling reactions using
PEPPSI-IPr as the source of palladium and potassium car-
bonate as base (for more details see the Experimental Sec-
tion in the Supporting Information). Accordingly, the initial
efforts were focused on the development of a practical
method towards the trimer (4) through incorporation of a
&
2
&
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Oligothiophene-Based Fluorescent Amyloid Ligands
FULL PAPER
Figure 1. Chemical structures of p-HTAA, p-HTAA-Se, p-FTAA, p-FTAA-Se, HS-68, HS-67, HS-72 and HS-53, and their optical characterization to-
wards in vitro formed recombinant Ab1–42 fibrils. The first panel displays the chemical structure of the oligothiophene and the corresponding thio-
phene/selenophene counterpart. The spectral graphs display excitation (second panel) and emission (third panel) spectra of the different ligands in PBS
with fibrillar Ab1–42. All probes show well-resolved excitation and emission spectra when binding to Ab fibrils. The fourth panel shows the fibrillation
of recombinant Ab1–42 monitored by fluorescence from ThT or the respective ligand. All the pentameric ligands detect early non-thioflavinophilic spe-
cies in the fibrillation pathway, whereas the tetramers show a high emission at the same time as the onset of ThT.
central selenophene within the
desired oligomer. The existing
routes to this type of trimer
generally involve cross-coupling
reactions between 2,5-thio-
phene diboronic acid/ester and
arylbromide compounds. De-
spite the fact that the synthesis
of 2,5-thiophenediylbisboronic
acid is convenient,[34,35] our ini-
tial results confirmed that di-
borylation of selenophene was
unfeasible. Therefore, an alter-
native strategy was used involv-
ing the treatment of 2,5-dibro-
Scheme 2. Synthesis of p-HTAA-Se. Reagents and conditions: a) 1,4-dioxane/MeOH, PEPPSITM-IPr, K2CO3,
708C, 20 min; b) NBS, DMF, 08C to RT, 16 h; c) 1,4-dioxane/MeOH, PEPPSITM-IPr, K2CO3, 708C, 20 min;
d) NaOH (1m), 1,4-dioxane, 508C, 16 h.
Chem. Eur. J. 2013, 00, 0 – 0
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
3
&
ÞÞ
These are not the final page numbers!

K. P. R. Nilsson et al.
moselenophene (2)[36] with methyl 2-(2-(4,4,5,5-tetramethyl-
1,3,2-dioxaborolan-2-yl)thiophen-3-yl)acetate (1). The latter
compound was prepared by taking advantage of protocols
developed by Murata et al. and Christophersen et al.[37,38]
using palladium-catalyzed borylation of pinacolborane and
methyl 2-(2-bromothiophen-3-yl)acetate (12).[5] Subsequent-
ly, the trimer was dibrominated in the a-positions by using a
twofold excess of NBS and then 5-(methoxy carbonyl)thio-
phene-2-boronic acid, pinacol ester (5) was coupled, by
using Suzuki chemistry, to give the pentameric thiophene/se-
lenophene co-oligomer 6. Furthermore, treatment of the
trimer 7[5] with 4,4,5,5-tetramethyl-2-(selenophen-2-yl)-1,3,2-
dioxaborolane (8), which was prepared according to a litera-
ture procedure,[36] gave compound 9, the precursor to p-
HTAA-Se. The two pentamers were hydrolyzed by using
sodium hydroxide in dioxane/water (1.5 equiv NaOH/ester)
to give the target compounds, p-FTAA-Se and p-HTAA-Se,
as the sodium salts.
It is generally known that side chain substitution on
neighboring thiophene rings causes oligothiophenes to
become nonplanar, and so far almost all of the previously
reported LCOs have been synthesized from a trimer build-
ing block, having 3-thiophene acetic acid attached to the 2-
and 5-position of an unsubstituted thiophene.[3,5] To increase
the conformational variability in the LCO library, a set of
asymmetrical tetrameric and pentameric LCOs having only
one 3-thiophene acetic acid unit along the backbone were
generated (Figure 1, Scheme 3). In addition, similar to as de-
scribed above, distinct thiophene units were replaced by se-
lenophene to render the thiophene/selenophene co-oligomer
analogue (Scheme 3). Treatment of commercially available
thiophene-2-boronic acid, pinacol ester 10 and the corre-
sponding selenophene-2-boronic acid, pinacol ester 11[35]
with the monomer 12 and 13 followed by bromination gave
the dimers 16 and 17 or trimers 22 and 23. These intermedi-
ates were subsequently coupled with 5-(methoxy carbon-
yl)thiophene-2-boronic acid, pinacol ester (5) to obtain
target tetramers 18 and 19 or pentamers 24 and 25. The
final products were subsequently transformed to salt form
by hydrolysis with NaOH to provide HS-68, HS-67, HS-72
and HS-53.
When diluted in phosphate buffered saline (PBS, pH 7.4)
all included co-oligomers displayed bathochromic shifts of
excitation and emission spectra as compared with the corre-
sponding oligothiophene (Supporting Information, Fig-
ure S1). The most pronounced red shift in both excitation
and emission mode was accomplished when alternating two
selenophenes and three thiophenes in an asymmetric pen-
tameric backbone (HS-53, Supporting Information, Fig-
ure S1D), whereas peripheral placement of the seleno-
phenes (p-HTAA-Se, Supporting Information, Figure S1A)
exerted the smallest effect on the optical behavior. Earlier
studies have confirmed that an increase in the number of se-
lenophene units in pentameric co-oligomers of thiophene
and selenophene results in a bathochromic shift, as the sele-
nium heteroatom lowers the band gap by stabilizing the
LUMO level.[30,31,36] In addition, the red shift might also be
observed due to an increased quinoid character of the sele-
À
nophene giving the inter-ring C C bond more double-bond
properties and an enhanced efficiency of conjugation
through the backbone.[29] In comparison with their thio-
phene analogues, the selenophene-containing ligands also
showed a decrease in the fluorescence intensity. The phe-
nomenon has been reported previously in a study with thio-
Scheme 3. Synthesis of HS-67, HS-68, HS-53 and HS-72. Reagents and conditions: a) 1,4-dioxane/MeOH, PEPPSITM-IPr, K2CO3, 708C, 20 min; b) NBS,
DMF, 08C to RT, 16 h; c) 1,4-dioxane/MeOH, PEPPSITM-IPr, K2CO3, 708C, 20 min; d) NaOH (1m), 1,4-dioxane, 508C, 16 h.
&
4
&
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Oligothiophene-Based Fluorescent Amyloid Ligands
FULL PAPER
phene and selenophene co-polymers[39] and might be de-
pendent on the lowering of the band gap increasing the
probability of nonradiative emission.
The asymmetric tetramers HS-67 and HS-68 displayed an
Ab1–42 kinetic graph with a short lag time followed by a
temporary increase in the fluorescence at the same time
point as the fluorescence onset of the pentameric ligands,
and then a second continuous growth phase corresponding
to the onset of ThT emission. Hence, HS-67 and HS-68
seem to interact with aggregated non-thioflavinophilic spe-
cies, observed as a slight increase in emission, but as with
ThT, the tetramers need bundles of fibrils in order to con-
vert into a highly emissive state. A similar Ab1–42 amyloid-
formation curve, but with a slightly less pronounced initial
growth phase, was observed for q-FTAA, an asymmetrical
tetramer substituted with a carboxyl group at one of the end
a-positions, whereas with the tetramers t-HTAA and q-
HTAA, which lack this carboxyl group, the early increase in
fluorescence was almost absent.[5] Hence, extending the con-
jugated backbone with carboxyl groups increases the emis-
sion from tetrameric dyes bound to aggregated non-thiofla-
vinophilic Ab1–42 species. However, when the excitation
mode was utilized to monitor the fibrillation kinetics, it was
clearly shown that the conformation-induced highly emissive
state for HS-68 was obtained at a similar time point as for
ThT, whereas p-FTAA and p-HTAA displayed an earlier
conformation-induced onset (Supporting Information, Fig-
ure S2). Overall, in agreement with previous studies,[3–5] we
conclude that a conjugated backbone consisting of at least
five units was necessary to achieve an amyloid ligand dis-
playing strong emission upon binding to aggregated non-
Characterization of in vitro formed recombinant Ab1–42 fi-
brils: In the next series of experiments, LCOs and thio-
phene/selenophene co-oligomers were tested on recombi-
nant Ab1–42 fibrils. Overall, intercalation with the amyloid
fibrils resulted in enhanced fluorescence and well-resolved
excitation and emission spectra for both LCOs and co-
oligomers; hence, all probes showed specific binding to Ab
fibrils and the resulting conformational restriction of the
backbones was mirrored in the defined spectra (Figure 1).
In comparison with probe diluted in PBS, each probe consis-
tently showed a red shift in excitation spectrum when mixed
with Ab fibrils. Additionally, intercalation with the fibrils
caused a blue shift in the emission spectrum for all probes.
Similar to the result in PBS, introducing selenophenes in the
thiophene backbone caused a shift in excitation and emis-
sion spectra towards longer wavelengths when the probes
were mixed with Ab fibrils, and the largest bathochromic
shifts in both modes were seen with HS-53 (Figure 1D). As
mentioned above, HS-53 showed the most pronounced shifts
also in PBS indicating that the design with alternating thio-
phenes and selenophenes in the backbone has the largest
effect on the resulting optical properties when compared to
the thiophene analogue. The tetrameric and pentameric li-
gands were then utilized to monitor the kinetics of recombi-
nant Ab1–42 fibril formation. The fibrillation was per-
formed under physiological conditions (378C, pH 7.4, quies-
cent), a protocol that has previously been reported to result
in the typical sigmoidal curve of amyloid formation when
ThT fluorescence is assessed as read-out for the reaction.[5]
The fluorescence from each probe and the reference ThT
was used to continuously monitor the fibrillation kinetics,
and the emission intensities were plotted against time
(Figure 1). In accordance with earlier observations, all pen-
tameric ligands displayed an Ab1–42 kinetic graph with a
short lag time followed by a steep growth phase and then a
decreasing plateau. The replacement of thiophenes with se-
lenophenes did not have any significant effect, as the result-
ing graphs for pentameric LCOs and corresponding co-
oligomers were more or less identical. All included LCOs
and co-oligomers with a pentameric backbone showed an
onset of fluorescence more than 200 min earlier than ThT.
Previous studies using transmission electron microscopy
have shown the presence of small fibrillar species during the
earlier time-points with observed strong p-FTAA emission,
whereas the onset of ThT fluorescence and decreased p-
FTAA emission was assigned to the presence of larger bun-
dles of amyloid fibrils.[3,5] Hence, in agreement with earlier
findings, all pentameric thiophene- and selenophene-based
ligands displayed a strong emission when bound to non-thio-
flavinophilic species preceding mature Ab fibrils. Introduc-
ing selenophenes or asymmetry along the pentameric back-
bone did not affect the optical performance of LCOs for de-
tection of early in vitro formed aggregated species of Ab.
AHCTUNGTRENGNtUN hioflavinophilic Ab1–42 species.
Characterization of Ab plaques and neurofibrillary tangles
in brain tissue sections: Previous studies have shown that
pentameric and heptameric LCOs identify a broader subset
of Ab deposits in brain tissue sections than conventional
amyloid ligands.[3,5,40] The staining pattern of pentameric
and heptameric LCOs resembled the overall Ab burden ob-
tained with Ab immunohistochemistry.[3,40] In addition, an
LCO consisting of at least five thiophene units with carboxyl
groups extending the conjugation length was required to
spectrally discriminate Ab aggregates and tau NFTs, the two
major histopathological hallmarks of AD, in AD brain
tissue sections.[3,5] To gain further insight regarding the un-
derlying molecular requirements for achieving an LCO-
based ligand for spectral separation of Ab aggregates and
NFTs, the library of ligands was applied to frozen AD brain
sections. All included LCOs and thiophene/selenophene co-
oligomers showed specific binding to Ab aggregates and
NFTs (Figure 2). When excited at 405 nm, all included
probes showed well-resolved emission spectra with charac-
teristic double peaks upon binding to assemblies of Ab and
tau, denoting conformational restrictions of the backbone
(Figure 2). Similar with the spectral results obtained in
buffer and with recombinant Ab fibrils, the thiophene/sele-
nophene co-oligomers displayed a red shift in emission com-
pared to their thiophene analogues owing to the selenium
containing unit(s) (Figure 2).
Chem. Eur. J. 2013, 00, 0 – 0
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
5
&
ÞÞ
These are not the final page numbers!

K. P. R. Nilsson et al.
Figure 2. Fluorescence images and emission spectra of p-HTAA, p-HAA-Se, p-FTAA, p-FTAA-Se, HS-68, HS-67, HS-72 and HS-53 bound to two histo-
pathological hallmarks of AD (Ab plaques and NFTs) in human brain tissue. The emission spectra of the indicated ligand bound to Ab plaques (green
dashed line) or NFTs (red solid line) when excited at 405 nm. The images show the typical pathological protein entities, Ab plaques (white arrows) and
NFTs (white arrow heads), from which the emission spectra were obtained. Green arrowheads in E and F indicate lipofuscin. Scale bars represent
20 mm.
Clearly, introducing selenophenes or asymmetry to the
molecule did not impair the amyloid-binding properties;
however, spectral analysis of the ligands bound to Ab aggre-
gates or NFTs revealed that the structural modifications had
major impact on the ability for spectral separation of Ab
plaques and NFTs. In previous studies,[3,5] p-FTAA has
emerged as the optimal spectral separator of Ab plaques
and NFTs in AD brain tissue. Herein, when p-FTAA was
excited at a single wavelength (405 nm), the emission signa-
ture of both Ab and tau peaked at identical wavelengths
(Figure 2C), however, the higher value of the last tau peak
as well as the more pronounced shoulder at 590 nm con-
firmed the earlier reported red shift of p-FTAA bound to
NFTs in comparison with Ab. The difference in emission
was further assessed by simultaneously exciting the probe at
405 and 538 nm, an operation that resulted in one single
emission spectrum in which both green and red contribu-
tions in emission were highlighted (Figure 3A). By plotting
&
6
&
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Oligothiophene-Based Fluorescent Amyloid Ligands
FULL PAPER
Figure 3. Emission spectra and ratio plot for: A) p-FTAA, B) p-FTAA-Se, C) p-HTAA, D) p-HTAA-Se, E) HS-72, F) HS-53, G) HS-68, and H) HS-67
bound to pathological hallmarks in AD human brain tissue. The emission spectra of the respective ligand bound to Ab plaques (green dashed line) or
NFTs (red solid line), when the ligand has been excited with two distinct wavelengths and the resulting emission spectra have been merged together.
Plot of the ratio of light intensity emitted at the indicated wavelengths for each probe when bound to Ab plaques (green triangles) or NFTs (red squares)
shown with standard deviation. The intensities are obtained from the merged emission spectra.
the ratio of the fluorescence intensity observed for the sepa-
rated peaks, the dissimilarity of the obtained spectral finger-
prints became obvious, and the shift of p-FTAA upon bind-
ing to misfolded tau was clearly illustrated and shown to be
significantly different with a difference between the means
of À0.396Æ0.013 (Figure 3A, and Supporting Information,
Table S1). Interestingly, the spectral analysis of AD brain
tissue stained with p-FTAA-Se revealed almost identical
emission spectra for the two protein entities and the gap be-
tween the mean values (difference between means=
À0.094Æ0.012) in the ratio plot was considerably decreased,
although still significantly different (Figure 3B, and Support-
ing Information, Table S1). In addition, the R2 value from
the unpaired t-test was 0.8901 for p-FTAA meaning that
89% of all the variation among the obtained values could
be ascribed to differences between the two group means,
whereas 11% of the variation resulted from scatter among
the values within the groups. The same number for p-
FTAA-Se was considerably lower (R2 =0.3421). Hence, by
replacing the central thiophene unit with a selenophene, the
distinct spectral separation of Ab and tau achieved with p-
FTAA was substantially reduced. As earlier studies have
shown that oligoselenophenes are more difficult to twist
than oligothiophenes,[29,33] the introduction of a selenophene
unit between the two 3-thiophene acetic moieties might re-
strict the conformational flexibility of the molecule. Ongo-
ing studies with solid state NMR spectroscopy and molecu-
lar modeling (work in progress) have indeed shown that
these groups, in combination with the carboxyl groups at the
a-end positions, are important for achieving an optimal in-
Chem. Eur. J. 2013, 00, 0 – 0
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
7
&
ÞÞ
These are not the final page numbers!

K. P. R. Nilsson et al.
teraction between the LCO and the amyloid. Similar to
what was recently reported for Congo red,[27] the LCOs
seem to preferably intercalate in a groove along the amyloid
fibril axis and the carboxyl moieties interact with cationic
residues flanking this groove. As previously reported, re-
placing the end carboxyl moiety with hydrogen, resulting in
p-HTAA or p-HTAA-Se, also diminished the spectral sepa-
ration of Ab plaques and NFTs (Figure 3C and D). The
emission spectra and ratio plot for p-HTAA demonstrated a
small shift of tau against shorter wavelengths, a phenomen-
on that was observed earlier for the hexameric probe hx-
HTAA.[5] The same tendency, although to a smaller degree,
was also seen in the emission spectra for the selenophene-
containing counterpart, p-HTAA-Se. The statistical values
from the unpaired t-test were also considerably lower (Sup-
porting Information, Table S1) than for the corresponding
analogues having end carboxyl moieties. Previous studies[3,5]
have not reported p-HTAA to be green shifted upon bind-
ing to tau, and this might be explained with the fact that
these earlier studies were not performed with a combination
of excitation wavelengths or a confocal microscope reducing
the light scattering.
The spectral analysis of the ligands with the second modi-
fication of the pentameric scaffold, keeping the end carboxyl
moieties but having only one 3-thiophene acetic acid moiety
in the backbone, verified the necessity of the terminal car-
boxyl groups for efficient spectral separation of Ab aggre-
gates and NFTs. When binding to Ab plaques and NFTs,
the asymmetric LCO HS-72 emitted light with a variation in
color, although the red shift of tau was not as evident as
with p-FTAA (difference between means=À0.1535Æ
0.008382; R2 =0.7529; Figure 3E, and Supporting Informa-
tion, Table S1). Similar to the observation for pentameric li-
gands with a symmetric backbone, introducing selenophenes
at distinct positions reduced the spectral shift between Ab
aggregates and NFTs (difference between means=
À0.07902Æ0.002721; Figure 3F). In addition, the ratio plot
of the resulting co-pentamer HS-53 demonstrated a very
narrow distribution of the data points and 88% (R2 =
0.8846) of all the variation among values was explained by
differences between the two group means (Supporting Infor-
mation, Table S1). By alternating selenophenes with thio-
phenes, the rigidity of the conjugated backbone should in-
crease considerably and this conformational restriction
might be reflected as a decrease in the spectral difference
between Ab aggregates and NFTs as well as a limited spec-
tral variation for individual aggregated entities.
especially within the population of misfolded tau. Analo-
gous to the asymmetric pentamers, the effect on the spectral
properties of replacing one thiophene with a selenophene
was smaller than what was observed with p-FTAA;
however, the co-oligomer HS-67 showed a wide spectral dis-
tribution for Ab aggregates instead of NFTs (Figure 3H,
and Supporting Information, Table S1). These distinct varia-
tions in emission spectra, both between the two protein enti-
ties and within the same protein, are in strong contrast to
the earlier results obtained with the tetramers t-HTAA, q-
HTAA and q-FTAA, as these dyes did not display any sig-
nificant spectral distinction between Ab and tau deposits.[5]
Chemically, HS-68 and q-FTAA contain the identical
number of thiophene units and charged groups. The only
difference is the positioning of the carboxyl groups along
the tetrameric backbone. Firstly, HS-68 has carboxyl groups
at both end a-positions, whereas only one of these positions
is functionalized with a carboxyl group in q-FTAA, and the
importance of these substitutions has been demonstrated
previously. It has been suggested that the red shift of tau is
caused by the carboxyl end-groups interacting with chemical
entities, such as cationic groups, of the protein, and earlier
results have shown that the pentameric LCO p-FTAA sur-
passes both hexameric and heptameric analogues,[5] showing
that the maximum effect of the carboxyl extensions on the
electronic system is obtained with a decreasing length of the
backbone. The result obtained herein for HS-68 verifies
these findings, as this tetrameric dye displayed a pronounced
spectral separation of Ab aggregates and NFTs compared to
p-FTAA and HS-72. Secondly, HS-68 has only one acetic
acid moiety, whereas q-FTAA is synthesized from a trimer
building block having an unsubstituted thiophene ring be-
tween two 3-thiophene acetic acid moieties. In contrast to
the carboxyl end-groups, an interaction between the acetic
acid derived carboxyl moieties with a chemical entity within
the amyloid binding groove will lock the acetic acid substi-
tuted thiophene rings in rather distinct positions. Hence,
upon binding to protein aggregates, the q-FTAA backbone
might be more conformationally restricted compared to the
HS-68 backbone, and this would, together with the diminu-
tion of the shift between Ab aggregates and NFTs when in-
troducing selenophenes, prove the assumption that the spec-
tral phenomenon is highly dependent on the degree of con-
formational freedom of the conjugated backbone.
The chemical design of HS-68 and p-FTAA appears to be
optimal when it comes to separate Ab plaques and NFTs in
AD. However, it should be noted that the specific correla-
tions between chemical design and spectral separation dis-
cussed above should only be applied to the Ab/tau model
system. For example, the most distinct spectral discrimina-
tion of prion aggregates associated with distinct prion strains
has been achieved with polythiophene acetic acid (PTAA),
having acetic acid substitution on all thiophene units.[8]
The most distinct spectral separation of Ab plaques and
NFTs was obtained with the asymmetric tetramers (Fig-
ure 3G and H, and Supporting Information, Table S1).
When the tetrameric LCO HS-68 was excited at 405 nm, the
first peak of the tau emission spectrum was barely noticea-
ble, confirming that the optical signature of the probe upon
binding to the protein was considerably shifted towards
longer wavelengths (Figure 2E). The double excitation pro-
tocol and the resulting ratio plot verified the observation
and revealed that HS-68 displayed a large spectral variation,
Synthesis and characterization of thiophene/phenylene co-
oligomers: Overall, the results described above further
proved the previously shown correlation between LCO
&
8
&
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Oligothiophene-Based Fluorescent Amyloid Ligands
FULL PAPER
Figure 4. A) The chemical structure of p-FTAA-Ph (top) and p-FTAA-MeOPh (bottom). B) Excitation, and C) emission spectra of p-FTAA-Ph (green)
and p-FTAA-MeOPh (red) in PBS (dashed line) or mixed with in vitro formed recombinant Ab1–42 fibrils (solid line). The excitation/emission wave-
lengths were 410/500 nm. D) The fibrillation of recombinant Ab1–42 monitored by fluorescence from ThT, p-FTAA, p-FTAA-Ph or p-FTAA-MeOPh. p-
FTAA detects early non-thioflavinophilic species in the fibrillation pathway, whereas p-FTAA-Ph and p-FTAA-MeOPh showed a high emission at the
same time as the onset of ThT.
structure and amyloid spectral fingerprinting. When moni-
toring the kinetics of Ab1–42, the length of the conjugated
backbone had to exceed four units to strongly fluoresce
upon binding to early formed aggregates not reactive with
ThT. In addition, from the histology experiments, the level
of conformational freedom along the conjugated backbone
was identified as a novel criterion for spectral separation of
Ab aggregates and NFTs. To verify all these findings, p-
FTAA was subjected to further molecular engineering and
as phenyl moieties previously have been utilized to disrupt
the planarity of oligothiophenes,[41–43] two co-oligomers
termed p-FTAA-Ph and p-FTAA-MeOPh with the central
thiophene ring replaced with a phenyl ring, were synthesized
(Figure 4, Scheme 4). The aim with these designs was to
create pentameric analogues to p-FTAA that have a nonpla-
nar backbone with a disrupted conjugation length as well as
a sterically restricted center.
The preparation of p-FTAA-Ph and p-FTAA-MeOPh was
carried out following the similar concept described above,
utilizing Suzuki cross-coupling reactions (Scheme 4). In
Scheme 4. Synthesis of p-FTAA-Ph and p-FTAA-MeOPh. Reagents and conditions: a) 1,4-dioxane/MeOH, PEPPSITM-IPr, K2CO3, 708C, 20 min;
b) NBS, DMF, 08C to RT, 16 h; c) 1,4-dioxane/MeOH, PEPPSITM-IPr, K2CO3, 708C, 20 min; d) NaOH (1m), 1,4-dioxane, 508C, 16 h.
Chem. Eur. J. 2013, 00, 0 – 0
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
9
&
ÞÞ
These are not the final page numbers!

K. P. R. Nilsson et al.
PBS, both the ligands displayed blue-shifted excitation spec-
tra compared to p-FTAA (Figure 4B, and Supporting Infor-
mation, Figure S1B), and when mixed with recombinant
Ab1–42 fibrils, the excitation spectra were only slightly
shifted towards longer wavelengths. Hence, the introduction
of a central phenyl ring diminished the well-resolved excita-
tion spectrum observed for all the other pentameric ligands
interacting with recombinant Ab1–42 fibrils. This observa-
tion was also reflected in the emission spectra, as the inter-
action with the amyloid fibrils resulted only in enhanced flu-
orescence from p-FTAA-Ph and p-FTAA-MeOPh. In the
Ab1–42 fibrillation experiment, p-FTAA-Ph and p-FTAA-
MeOPh displayed an extended lag phase compared to p-
FTAA, and the kinetic graphs almost completely overlapped
with the one obtained with ThT (Figure 4D). Hence, by re-
placing the central thiophene ring with a phenyl ring, the
ability to detect early formed non-thioflavinophilic aggre-
gates of Ab1–42 was abolished and amyloid ligand proper-
ties more similar to ThT than pentameric thiophene-based
molecules were acquired.
To further explore the findings from the in vitro fibrilla-
tion experiments, p-FTAA-Ph and p-FTAA-MeOPh was
next applied to tissue sections from AD brain as well as
prion infected transgenic mice with prion aggregates that
were identified by oligo- and polythiophenes but went unde-
tected by ThT and Congo red.[44] As shown in Figure 5, p-
FTAA emission could easily be utilized to detect the prion
aggregates and the resulting strong fluorescence allowed de-
tection of also very small-sized deposits. These prion aggre-
gates were also detected by all the other oligothiophenes or
thiophene/selenophene co-oligomers (Supporting Informa-
tion, Figure S3). With p-FTAA-Ph, it was very difficult to
distinguish aggregates from the surrounding tissue and the
exposure time used in the fluorescence image was more
than 60-times longer than for p-FTAA (Figure 5). With p-
FTAA-MeOPh, we could not identify any prion aggregates,
indicating that this dye performs similar to ThT and Congo
red. To rule out the possibility that the marked difference in
fluorescence intensity was an inherent property of the
probes, their labeling of congophilic Ab plaques in AD
brain tissue was compared (Figure 5C). p-FTAA-Ph showed
weaker fluorescence also from Ab deposits, but the expo-
sure time was only three-times longer than the one used for
p-FTAA. In addition, congophilic Ab plaques could be read-
ily identified with p-FTAA-MeOPh. Hence, the low intensi-
ty from p-FTAA-Ph upon binding to the prion deposits and
the lack of binding of p-FTAA-MeOPh to the same aggre-
gated species were only observed for deposits that are not
recognized by conventional amyloid ligands. Taken together
with the previous findings for the thiophene/selenophene
co-oligomers, these observations indicate that an extended
conjugation of the backbone is necessary to obtain strong
fluorescence signals from non-congophilic and non-thioflavi-
nophilic protein aggregates. The presumed disordered struc-
ture found in early in vitro formed Ab aggregates or in the
prion deposits may allow the thiophene/phenylene co-pen-
tamers to bind, but prevents them from adopting a fully
emissive conformation. Alternatively, pentameric thiophene/
selenophene co-oligomers might identify a wider range of
aggregate topologies than p-FTAA-Ph and p-FTAA-
MeOPh. In either case, we conclude that replacing the
center thiophene ring with a phenyl moiety will diminish
sensitive fluorescence detection of a variety of early formed
aggregated species or disease-associated protein aggregates.
Next we investigated whether the introduced phenyl
moiety in the p-FTAA scaffold had an effect on spectral fin-
gerprinting of Ab and tau deposits. Both Ab and tau aggre-
gates were easily identified with p-FTAA-Ph (Figure 5D).
However, the emission spectra from p-FTAA-Ph demon-
strated a maximum peak at 469 nm and a pronounced
shoulder at 495 nm for both these aggregated species (Fig-
ure 5E). Hence, the distinct spectral separation of Ab and
tau obtained with p-FTAA was completely abolished,
although p-FTAA-Ph has carboxyl groups extending the
conjugated backbone. As a similar effect, although not as
pronounced, was seen when increasing the rigidity of the
main chain by replacing the central thiophene with seleno-
phene (Figure 3), it could be concluded that conformational
freedom along the conjugated backbone in combination
with end carboxyl moieties is necessary for achieving superi-
or spectral separation of the two pathological hallmarks ob-
served in AD. Frozen AD brain sections were also stained
by p-FTAA-MeOPh, and as mentioned earlier, Ab deposits
were easily identified due to strong emission from the probe
(Figure 5C and D). However, aggregated tau species, NFTs
and dystrophic neurites, were very difficult to detect. In ad-
dition, increased background fluorescence, most likely from
lipofuscin, was also observed. To verify the almost complete
lack of aggregated tau staining from p-FTAA-MeOPh, brain
sections from transgenic mice expressing human P301S tau
protein[45] were stained with p-FTAA and its phenyl ana-
logues (Figure 5F). Analogous to the results obtained for
the non-thioflavinophilic and non-congophilic prion aggre-
gates, intracellular tau aggregates in the brain stem could
easily be identified by p-FTAA fluorescence, whereas p-
FTAA-Ph stained aggregates displayed only weak fluores-
cence and p-FTAA-MeOPh showed an almost complete
lack of aggregated tau staining (Figure 5F). These observa-
tions strengthen our conclusions that substitution of the cen-
tral thiophene ring with a phenyl moiety will diminish sensi-
tive fluorescence detection of distinct protein aggregates.
Furthermore, introducing methoxy substituents on the
phenyl ring will completely abolish fluorescence detection
of certain protein deposits. Interestingly, a similar methoxy
modification on the central phenyl ring of the Congo red an-
alogue, X-34,[46] has rendered an amyloid ligand, MeO-X-
04,[47] that allows fluorescence detection of Ab and tau de-
posits in postmortem sections of AD brain with good specif-
icity. This is in contrast to our finding for conformationally
flexible amphiphilic thiophene-based ligands. Recently, it
was also reported that amphiphilic Nꢀ-benzylidene-benzo-
AHCTUNGTREGhNNNU ydrazide (NBB)-based fluorescent amyloid ligands with
bulky hydrophilic groups showed selectivity towards tau ag-
gregates, whereas non-amphiphilic NBB ligands were rather
&
10
&
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Oligothiophene-Based Fluorescent Amyloid Ligands
FULL PAPER
Figure 5. Fluorescence images of: A), B) prion infected transgenic mice, C), D) AD, or F) tauP301S transgenic mice brain sections stained with p-FTAA, p-
FTAA-Ph or p-FTAA-MeOPh. White arrows and arrowheads indicate Ab plaques and NFTs, respectively, whereas the green arrowhead identifies lipo-
fuscin. Scale bars represent 50 mm. E) Emission spectra collected from Ab plaques (green dashed line) and tau tangles (red solid line) labeled with p-
FTAA-Ph.
Chem. Eur. J. 2013, 00, 0 – 0
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
11
&
ÞÞ
These are not the final page numbers!

K. P. R. Nilsson et al.
Table 1. Amyloid topologies detected and spectrally assigned by the respective ligand.
Probe
Chemical structure
Detection of ThT negative Ab Spectral separation of
Detection of ThT negative
prion aggregates
aggregates
Ab and tau
p-FTAA
+ + +
+ + +
+ + +
p-FTAA-Se
p-FTAA-Ph
+ + +
+
+ +
À
À
À
À
p-FTAA-
MeOPh
À
À
p-HTAA
p-HTAA-Se
HS-72
+ + +
+ + +
+ + +
+
+ + +
+
+ +
+ +
+ +
HS-53
HS-68
HS-67
+ + +
+
+ +
À
À
+ + +
+ + +
+
+
selective for Ab deposits.[48] Similar to our thiophene li-
gands, the NBBs display a high degree of rotational free-
dom, suggesting that introduction of hydrophobic moieties,
such as methoxy groups, in amyloid ligands with a flexible
molecular scaffold will abolish fluorescence detection of tau
aggregates. Whether this phenomenon is associated with
lack of binding, inability of the molecule to adopt a highly
emissive conformation or a combination of these factors
needs to be investigated further. Overall, we conclude that
the central thiophene ring in pentameric LCOs is an essen-
tial chemical element for sensitive fluorescent assignment of
a variety of disease-associated protein aggregates.
&
12
&
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Oligothiophene-Based Fluorescent Amyloid Ligands
FULL PAPER
[6] V. Mahajan, T. Klingstedt, R. Simon, K. P. R. Nilsson, A. Thuering-
er, K. Kashofer, J. Haybaeck, H. Denk, P. M. Abuja, K. Zatloukal,
Gastroenterology 2011, 141, 1080–1090.
[7] K. P. R. Nilsson, A. ꢁslund, I. Berg, S. Nystrçm, P. Konradsson, O.
Inganꢃs, F. Stabo-Eeg, M. Lindgren, G. T. westermark, L. Lannfelt,
L. N. Nilsson, P. Hammarstrçm, ACS Chem. Biol. 2007, 2, 553–560.
[8] C. J. Sigurdson, K. P. R. Nilsson, S. Hornemann, G. Manco, M. Poly-
menidou, P. Schwarz, M. Leclerc, P. Hammarstrçm, K. Wꢄthrich, A.
Aguzzi, Nat. Methods 2007, 4, 1023–1030.
[9] K. P. R. Nilsson, K. Ikenberg, A. ꢁslund, S. Fransson, P. Konrads-
son, C. Rçcken, H. Moch, A. Aguzzi, Am. J. Pathol. 2010, 176, 563–
574.
[10] I. Berg, K. P. R. Nilsson, S. Thor, P. Hammarstrçm, Nat. Protoc.
2010, 5, 935–944.
[11] B. M. Wegenast-Braun, A. Skodras, G. Bayraktar, J. Mahler, S. K.
Fritschi, T. Klingstedt, J. J. Mason, P. Hammarstrçm, K. P. R. Nils-
son, C. Liebig, M. Jucker, Am. J. Pathol. 2012, 181, 1953–1960.
[12] T. Klingstedt, K. P. R. Nilsson, Biochim. Biophys. Acta. 2011, 1810,
286–296.
[13] E. E. Nesterov, J. Skoch, B. T. Hyman, W. E. Klunk, B. J. Bacskai,
T. M. Swager, Angew. Chem. 2005, 117, 5588–5592; Angew. Chem.
Int. Ed. 2005, 44, 5452–5456.
[14] A. Fin, A. Vargas Jentzsch, N. Sakai, S. Matile, Angew. Chem. 2012,
124, 12908–12911; Angew. Chem. Int. Ed. 2012, 51, 12736–12739.
[15] P. Yan, A. Xie, M. Wei, L. M. Loew, J. Org. Chem. 2008, 73, 6587–
6594.
[16] M. Zambianchi, F. Di Maria, A. Cazzato, G. Gigli, M. Piacenza, F.
Della Sala, G. Barbarella, J. Am. Chem. Soc. 2009, 131, 10892–
10900.
[17] M. Dal Molin, S. Matile, Org. Biomol. Chem. 2013, 11, 1952–1957.
[18] H. Zetterberg, P. Hammarstrçm, Neurochem. 2012, 122, 231–232.
[19] H. Zetterberg, K. Blennow, Alzheimers Dement. 2012; DOI:
10.1016/j.jalz.2012.07.002.
[20] M. K. Sandberg, H. Al-Doujaily, B. Sharps, A. R. Clarke, J. Collinge,
Nature 2011, 470, 540–542.
[21] J. Hardy, D. J. Selkoe, Science 2002, 297, 353–356.
[22] D. M. Walsh, I. Klyubin, J. V. Fadeeva, W. K. Cullen, R. Anwyl,
M. S. Wolfe, M. J. Rowan, D. J. Selkoe, Nature 2002, 416, 535–539.
[23] R. M. Koffie, M. Meyer-Luehmann, T. Hashimoto, K. W. Adams,
M. L. Mielke, M. Garcia-Alloza, K. D. Micheva, S. J. Smith, M. L.
Kim, V. M. Lee, B. T. Hyman, T. L. Spires-Jones, Proc. Natl. Acad.
Sci. USA 2009, 106, 4012–4017.
[24] B. Winner, R. Jappelli, S. K. Maji, P. A. Desplats, L. Boyer, S.
Aigner, C. Hetzer, T. Loher, M. Vilar, S. Campioni, C. Tzitzilonis,
A. Soragni, S. Jessberger, H. Mira, A. Consiglio, E. Pham, E. Mas-
liah, F. H. Gage, R. Riek, Proc. Natl. Acad. Sci. USA 2011, 108,
4194–4199.
[25] F. Bemporad, F. Chiti, Chem. Biol. 2012, 19, 315–327.
[26] J. D. Sipe, M. D. Benson, J. N. Buxbaum, S. Ikeda, G. Merlini, M. J.
Saraiva, P. Westermark, Amyloid 2012, 19, 167–170.
[27] A. K. Schꢄtz, A. Soragni, S. Hornemann, A. Aguzzi, M. Ernst, A.
Bçckmann, B. H. Meier, Angew. Chem. 2011, 123, 6078–6082;
Angew. Chem. Int. Ed. 2011, 50, 5956–5960.
[28] T. Klingstedt, C. Blechschmidt, A. Nogalska, S. Prokop, B. Hꢃggqv-
ist, O. Danielsson, W.-K. Engel, V. Askanas, F. L. Heppner, K. P. R.
Nilsson, ChemBioChem 2013, 14, 607–616.
Conclusion
Understanding the optical performance of fluorescent amy-
loid ligands is crucial for obtaining molecular probes that
identify a variety of disease-associated protein aggregates.
Herein, we have verified and identified distinct molecular
requirements that are highly important for achieving thio-
phene-based ligands that are superior, compared to conven-
tional ligands, for optical assignment of disease-associated
protein aggregates. In addition, we have shown that small
chemical modifications of the molecular scaffold significant-
ly influence the diversity of amyloid topologies that are de-
tected or spectrally assigned by the ligand (Table 1). Firstly,
replacing the central thiophene ring with a phenyl moiety
abolished fluorescent detection of non-thioflavinophilic Ab
species and non-congophilic prion aggregates. Secondly, the
level of conformational freedom along the conjugated back-
bone was identified as a novel criterion for spectral separa-
tion of Ab aggregates and NFTs, the two major pathological
hallmarks of AD. Overall, we have demonstrated that a
combination of fluorescent ligands will facilitate the detec-
tion of a broader subset of amyloid topologies, and the re-
sults presented also underline the importance of understand-
ing the variety of disease-associated protein aggregates that
are identified with distinct ligands.
Experimental Section
Full experimental details including additional characterization data and
NMR spectra of new compounds are given in the Supporting Informa-
tion.
Acknowledgements
Our work is supported by the Swedish Foundation for Strategic Research
(K.P.R.N, T.K.). K.P.R.N is financed by an ERC Starting Independent
Researcher Grant (Project: MUMID) from the European Research
Council. We would also like to thank Prof. David M. Holtzman and the
Knight Alzheimerꢀs Disease Research Center (NIH grants P50AG05681
and 5PO1-AG03991) Washington University (St. Louis, MO, USA) for
supplying us with human brain tissue. A generous gift from Astrid and
Georg Olsson is also gratefully acknowledged. T.K. is enrolled in the doc-
toral program Forum Scientum. We also acknowledge the Swedish Na-
tional Laboratory of Forensic Science for the assistance with high-resolu-
tion mass spectrometry.
[29] A. Patra, M. Bendikov, J. Mater. Chem. 2010, 20, 422–433.
[30] M. Heeney, W. Zhang, D. J. Crouch, M. L. Chabinyc, S. Gordeyev,
R. Hamilton, S. J. Higgins, I. McCulloch, P. J. Skabara, D. Sparrowe,
S. Tierney, Chem. Commun. 2007, 5061–5063.
[31] S. Das, S. S. Zade, Chem. Commun. 2010, 46, 1168–1170.
[32] S. S. Zade, N. Zamoshchik, M. Bendikov, Chem. Eur. J. 2009, 15,
8613–8624.
[1] C. A. Ross, M. A. Poirier, Nat. Med. 2004, 10, S10–S17.
[2] F. Chiti, C. M. Dobson, Annu. Rev. Biochem. 2006, 75, 333–366.
[3] A. ꢁslund, C. J. Sigurdson, T. Klingstedt, S. Grathwohl, T. Bolmont,
D. L. Dickstein, E. Glimsdal, S. Prokop, M. Lindgren, P. Konrads-
son, D. M. Holtzman, P. R. Hof, F. L. Heppner, S. Gandy, M. Jucker,
A. Aguzzi, P. Hammarstrçm, K. P. R. Nilsson, ACS Chem. Biol.
2009, 4, 673–684.
[33] Y. H. Wijsboom, A. Patra, S. S. Zade, Y. Sheynin, M. Li, L. L. W.
Shimon, M. Bendikov, Angew. Chem. 2009, 121, 5551–5555; Angew.
Chem. Int. Ed. 2009, 48, 5443–5447.
[34] J. Hellberg, T. Remonen, F. Allared, J. Slatt, M. Svensson, Synthesis
2003, 14, 2199–2205.
[4] P. Hammarstrçm, R. Simon, S. Nystrçm, P. Konradsson, A. ꢁslund,
K. P. R. Nilsson, Biochemistry 2010, 49, 6838–6845.
[5] T. Klingstedt, A. ꢁslund, R. A. Simon, L. B. G. Johansson, J. J.
Mason, S. Nystrçm, P. Hammarstrçm, K. P. R. Nilsson, Org. Biomol.
Chem. 2011, 9, 8356–8370.
Chem. Eur. J. 2013, 00, 0 – 0
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
13
&
ÞÞ
These are not the final page numbers!

K. P. R. Nilsson et al.
[35] H. Usta, G. Lu, A. Facchetti, T. J. Marks, J. Am. Chem. Soc. 2006,
128, 9034–9035.
[43] H. Zgou, S. M. Bouzzine, S. Bouzakraoui, M. Hamidi, M. Bouachr-
ine, Chin. Chem. Lett. 2008, 19, 123–126.
[44] C. J. Sigurdson, S. Joshi-Barr, C. Bett, O. Winson, G. Manco, P.
Schwarz, T. Rꢄlicke, K. P. R. Nilsson, I. Margalith, A. Raeber, D.
Peretz, S. Hornemann, K. Wꢄthrich, A. Aguzzi, J. Neurosci. 2011,
31, 13840–13847.
[36] S. Haid, A. Mishra, C. Uhrich, M. Pfeiffer, P. Bꢃuerle, Chem. Mater.
2011, 23, 4435–4444.
[37] C. Christophersen, M. Begtrup, S. Ebdrup, H. Petersen, P. Vedsø, J.
Org. Chem. 2003, 68, 9513–9516.
[45] B. Allen, E. Ingram, M. Takao, M. J. Smith, R. Jakes, K. Virdee, H.
Yoshida, M. Holzer, M. Craxton, P. C. Emson, C. Atzori, A. Migheli,
R. A. Crowther, B. Ghetti, M. G. Spillantini, M. Goedert, J. Neuro-
sci. 2002, 22, 9340–9351.
[46] S. D. Styren, R. L. Hamilton, G. C. Styren, W. E. Klunk, J. Histo-
chem. Cytochem. 2000, 48, 1223–1232.
[47] W. E. Klunk, B. J. Bacskai, C. A. Mathis, S. T. Kajdasz, M. E. McLel-
lan, M. P. Frosch, M. L. Debnath, D. P. Holt, Y. Wang, B. T. Hyman,
J. Neuropathol. Exp. Neurol. 2002, 61, 797–805.
[48] A. Taghavi, S. Nasir, M. Pickhardt, R. Heyny-von Haussen, G. Mall,
E. Mandelkow, E. M. Mandelkow, B. Schmidt, J. Alzheimer’s Dis.
2011, 27, 835–843.
[38] M. Murata, T. Oyama, S. Watanabe, Y. Masuda, J. Org. Chem. 2000,
65, 164–168.
[39] J. Hollinger, A. A. Jahnke, N. Coombs, D. S. Seferos, J. Am. Chem.
Soc. 2010, 132, 8546–8547.
[40] S. Nystrçm, K. M. Psonka-Antonczyk, P. G. Ellingsen, L. B. G. Jo-
hansson, N. Reitan, S. Handrick, S. Prokop, F. L. Heppner, B. M.
Wegenast-Braun, M. Jucker, M. Lindgren, B. T. Stokke, P. Hammar-
strçm, K. P. R. Nilsson, ACS Chem. Biol. 2013; DOI: 10.1021/
cb4000376.
[41] S. A. Lee, S. Hotta, F. Nakanishi, J. Phys. Chem. A 2000, 104, 1827–
1833.
[42] C. B. Nielsen, A. Angerhofer, K. A. Abboud, J. R. Reynolds, J. Am.
Chem. Soc. 2008, 130, 9734–9746.
Received: April 17, 2013
Published online: && &&, 0000
&
14
&
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Oligothiophene-Based Fluorescent Amyloid Ligands
FULL PAPER
Freedom! Conformational freedom
and extended conjugation of the conju-
gated backbone are important determi-
nants for thiophene-based optical
ligands for spectral assignment of dis-
ease-associated protein aggregates.
Replacing the central thiophene ring
with a phenyl moiety (see figure) abol-
ished the spectral assignment of Ab
and tau aggregates as well as the
Fluorescent Probes
T. Klingstedt, H. Shirani,
K. O. A. ꢁslund, N. J. Cairns,
C. J. Sigurdson, M. Goedert,
K. P. R. Nilsson* ............. &&&& — &&&&
The Structural Basis for Optimal
Performance of Oligothiophene-Based
Fluorescent Amyloid Ligands: Confor-
mational Flexibility is Essential for
Spectral Assignment of a Diversity of
Protein Aggregates
detection of non-congophilic and non-
thioflavinophilic protein aggregates.
Chem. Eur. J. 2013, 00, 0 – 0
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
15
&
ÞÞ
These are not the final page numbers!