Dicyanovinylnaphthalenes for neuroimaging ofamyloids and relationships of electronic structuresand geometries to binding affinitiesAndrej Petri�ca,b, Scott A. Johnsonc, Hung V. Phamc, Ying Lic, Simon �Ceha, Amalija Golobi�ca, Eric D. Agdeppad,Gerald Timbold, Jie Liud, Gyochang Keumd,e, Nagichettiar Satyamurthyd, Vladimir Keped, Kendall N. Houkc,1,and Jorge R. Barriod

aDepartment of Chemistry and Chemical Technology, University of Ljubljana, 1000 Ljubljana, Slovenia; bEN-FIST Centre of Excellence, Dunajska 156, SI-1000Ljubljana, Slovenia; cDepartment of Chemistry and Biochemistry, University of California, Los Angeles, CA 90095; dDepartment of Molecular and MedicalPharmacology, The David Geffen School of Medicine, University of California, Los Angeles, CA 90095; and eCenter for Neuro-Medicine, Brain Science Institute,Korea Institute of Science and Technology Hwarangro 14-gil 5 Seongbuk-gu, Seoul 136-791, Republic of Korea

Contributed by Kendall N. Houk, August 20, 2012 (sent for review July 22, 2012)

The positron-emission tomography (PET) probe 2-(1-[6-[(2-fluo-roethyl)(methyl)amino]-2-naphthyl]ethylidene) (FDDNP) is usedfor the noninvasive brain imaging of amyloid-β (Aβ) and otheramyloid aggregates present in Alzheimer’s disease and other neu-rodegenerative diseases. A series of FDDNP analogs has been syn-thesized and characterized using spectroscopic and computationalmethods. The binding affinities of these molecules have beenmeasured experimentally and explained through the use of a com-putational model. The analogs were created by systematicallymodifying the donor and the acceptor sides of FDDNP to learnthe structural requirements for optimal binding to Aβ aggregates.FDDNP and its analogs are neutral, environmentally sensitive, fluo-rescent molecules with high dipole moments, as evidenced by theirspectroscopic properties and dipole moment calculations. The pre-ferred solution-state conformation of these compounds is directlyrelated to the binding affinities. The extreme cases were a nonplanaranalog t-butyl-FDDNP, which shows low binding affinity for Aβaggregates (520 nM Ki) in vitro and a nearly planar tricyclic analogcDDNP, which displayed the highest binding affinity (10 pM Ki). Usinga previously published X-ray crystallographic model of 1,1-dicyano-2-[6-(dimethylamino)naphthalen-2-yl]propene (DDNP) bound to anamyloidogenic Aβ peptide model, we show that the binding affin-ity is inversely related to the distortion energy necessary to avoidsteric clashes along the internal surface of the binding channel.

density functional theory | M06-2X | docking

Extracellular amyloid-β (Aβ) senile plaques (SPs) and neuro-fibrillary tangles (NFTs) of intraneuronal hyperphosphorylated

tau peptide aggregates are a characteristic pathology found inthe brains of patients with Alzheimer’s disease (AD) and otherneurodegenerative diseases (1). Accurate detection of theseaggregates in the brains of living AD patients allows early di-agnosis and potential treatments to reverse or retard diseaseprogression. The fluorescent thioflavin T and Congo red dyes aretraditionally used to label protein aggregates in brain slices forpostmortem diagnosis of AD. Unlike these charged molecules,the uncharged fluorescent naphthalene derivative 1,1-dicyano-2-[6-(dimethylamino)naphthalen-2-yl]propene (DDNP) (Fig. 1),also an excellent staining dye in vitro, is capable of crossing theblood–brain barrier. The fluorinated DDNP analog, 2-(1-[6-[(2-fluoroethyl)(methyl)amino]-2-naphthyl]ethylidene) (FDDNP), canbe synthesized with the 18F positron emitter and was used as thefirst molecular imaging probe for the regional assessment of Aβand tau deposition in the brains of living AD patients (2–5).Previous studies have shown that it is the neutral and hydro-

phobic properties of DDNP and its analogs that enable it torecognize the β-sheet domains in amyloid-like aggregates,making it an effective in vitro and in vivo imaging probe (6, 7).Recently, DDNP and other amyloid-imaging probes have been

cocrystallized with short amyloid-forming segments of the Aβand tau polypeptides that form fibrils similar to those formedwith the full protein (8). These truncated systems contain stacksof self-complementary β-sheets tightly bound to each other ina motif called a steric zipper. The fibrils formed by the stericzippers are similar to those formed by the full-length parentsequence in morphology, diameter, helical pitch, cross-β dif-fraction pattern, fibril seeding capacity, stability, and dye binding(8–10). Numerous investigations have shown that although thesteric zippers do not contain all of the elements of complexitycontained in the parent systems, they serve as excellent modelsfor full-length fibrils (8, 11). Recently, the VQIVYK steric zippermodel from tau has been used to develop a small polypeptideinhibitor of not only truncated peptide aggregation but also ag-gregation of two tau constructs as well (12). The X-ray structuressolved with DDNP bound to the VQIVYK steric zipper sug-gested transient binding of DDNP along a hydrophobic channelthat runs the length of the longitudinal fibril axis (or fibril spine).The presence of a “smear” of electron density along the bindingchannel indicated that there may be multiple sites that DDNPbinds, unlike the more discrete binding sites associated withcharged molecular probes that preferred highly localized sitesnear complementary charged residues. Molecular docking wasused in the crystallographic study to identify the preferred bindingmodes and sites of DDNP along the fibril channel, while alsomanaging to recapitulate the experimentally observed transientdocking phenomenon.We have synthesized a number of FDDNP analogs (Fig. 1)

that differ in size both at the electron-donating amine and electron-withdrawing dicyanovinyl moieties. Herein, we report the effectsof chemical modifications on the molecular geometries, spec-troscopic properties, and binding to Aβ fibrils. We then use mo-lecular docking to study the binding modes of these molecules tothe established tau steric zipper model. The binding informationis used to rationalize the effects of chemical substitutions andchanges in molecular geometry on the binding affinity.

Author contributions: A.P., K.N.H., and J.R.B. designed research; A.P., S.A.J., H.V.P., Y.L.,S.�C., A.G., E.D.A., G.T., J.L., G.K., N.S., and V.K. performed research; S.A.J., H.V.P., andK.N.H. analyzed data; and S.A.J., H.V.P., Y.L., K.N.H., and J.R.B. wrote the paper.

The authors declare no conflict of interest.

Data deposition: The DDNP structures have been deposited in the Cambridge StructuralDatabase (CSD), www.ccdc.cam.ac.uk/products/csd (CSD reference code NAFZUP).Supplementary crystallographic data for compounds 5, 6b, 7b, 8b, 10b(r), 10b(o) and11b have also been deposited with the Cambridge Crystallographic Data Centre assupplementary publication numbers CCDC 669016 - CCDC 669021 and CCDC 673983,respectively. These data can be obtained free of charge at http://www.ccdc.cam.ac.uk/conts/retrieving.html.1To whom correspondence should be addressed. E-mail: houk@chem.ucla.edu.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1214134109/-/DCSupplemental.

16492–16497 | PNAS | October 9, 2012 | vol. 109 | no. 41 www.pnas.org/cgi/doi/10.1073/pnas.1214134109

Dow

nloa

ded

by g

uest

on

Janu

ary

22, 2

020

Dow

nloa

ded

by g

uest

on

Janu

ary

22, 2

020

Dow

nloa

ded

by g

uest

on

Janu

ary

22, 2

020

Experimental Results and DiscussionNaphthalene Derivatives. Fluorescent compounds 4–11 (Fig. 1)were prepared based on an established synthetic methodology(for details see Materials and Methods and SI Appendix) (3, 13,14). FDDNP and its analogs are environmentally sensitive fluo-rophores that undergo a significant change in absorption/emis-sion wavelength upon binding to SPs or NFTs (3, 7, 15).Fluorescence microscopy revealed an intense fluorescence of SPsand NFTs with minimal background fluorescence in postmortemAD brain specimens stained with FDDNP analogs. The bindingaffinity of each FDDNP analog was measured using radioactivecompetitive-binding assays with [18F]FDDNP. Addition of non-radioactive 4b, 6b, 7b, 8b, 9c, 10b, or 11b displaces [18F]FDDNPfrom the Aβ fibril-binding site. This indicates that the analogsshare the same [18F]FDDNP-binding site on the Aβ aggregates(Fig. 2, Upper). All compounds, except 9c, exhibit rather similarnanomolar binding affinity to Aβ fibrils (Ki; Table 1). Compound9c has only weak binding to Aβ fibrils, also reflected in its poorability to label Aβ aggregates in human brain specimens (Fig.2, Lower).The ketone intermediates did have good affinities for amyloid

fibrils in vitro (16). We tested the [18F]fluoroethyl (1-{6-[(2-flu-oroethyl)(methyl)amino]naphthalen-2-yl}ethanone) (ADMAN)derivative in vivo in humans, which showed significant nonspecific(nonamyloid) binding, consistent with the membrane-intercalat-ing properties of the 2-(dialkylamino)-6-acylnaphthalenes (17).This reduces ketone effectiveness for in vivo use in humans,and this report concentrates on the binding affinities and prop-erties of the amyloid-specific dicyanovinyl compounds.

NMR Spectra. 1H NMR spectra reflect the structural changesupon formal substitution of the acetyl group for the dicyanovinylacceptor group, most notably through a decrease in chemical

shift of proton H-1 from ∼8.3 to 8.0 ppm attributable to modifiedmagnetic anisotropy of the side chain (see SI Appendix). The 1HNMR spectra also provide insights into the ground-state mo-lecular geometry in solution, particularly around the nitrogenatom. Planar geometries support efficient delocalization of thenitrogen electron pair into the aromatic ring and result in increasedshielding and smaller chemical shifts for protons H-5 and H-7(δ; Table 1). In compounds with nonplanar arrangement of thesubstituents around the donor nitrogen (5, 8a, 8b), chemicalshifts of the corresponding protons are substantially larger. Thedistinct coupling pattern in solution 1H NMR spectra of com-pound 8b and related piperidine and piperazine derivativesshows that at room temperature, the six-membered ring is fixedin a chair conformation (18). This is in agreement with theground-state calculations for 8b, in which the heavy atom anglesin the six-membered rings are shown to be between 108° and 112°(see Structure and Modeling). Similar NMR spectral results wereobtained for compounds with analogous electron-donatinggroups, which were concluded to contain nonplanar amines (13,14). For compound 6b, which exists both in planar and nonplanarforms in the solid state, 1H chemical shifts for protons H-5 andH-7 were found to be relatively small in solution. This indicatesa greater propensity toward the planar arrangement about thenitrogen, allowing for efficient conjugation with the naphthalenering system.

Absorption and Emission Spectra. For all compounds except 9c,replacement of the acetyl acceptor group by the dicyanovinyl sidechain resulted in considerable bathochromic shifts of the ab-sorption maxima from 350–380 to 400–470 nm (λmax; Table 1).This observation is consistent with previous studies that dem-onstrated a propensity for the dicyanovinyl to be planar andconjugated with the donor group (19). The lack of red shift in 9c is

HO

O

R

HO

O

X

Y

N N N N N N

1 R = Me2 R = t-Bu

N

Y

XN

Y

N

Y

H

3 4 ~ 8

N

CNNC

F

FDDNP

ADMAN Y = O DDNP Y = C(CN)2

4a Y = O4b Y = C(CN)2

5 Y = O 6a Y = O6b Y = C(CN)2

7a Y = O7b Y = C(CN)2

8a Y = O8b Y = C(CN)2

9a Y = O, X = CH2CH2OH9b Y = C(CN)2, X = CH2CH2OH9c Y = C(CN)2, X = CH2CH2F

10a Y = O10b Y = C(CN)2

11a Y = O11b Y = C(CN)2

Fig. 1. Structures of synthesized FDDNP analogs.

Petri�c et al. PNAS | October 9, 2012 | vol. 109 | no. 41 | 16493

BIOCH

EMISTR

YCH

EMISTR

Y

Dow

nloa

ded

by g

uest

on

Janu

ary

22, 2

020

the result of the acceptor being forced out of the plane of conju-gation because of the bulky t-butyl substituent. The absorptionmaxima are also dependent on the extent of conjugation of thedonor group, although the effect is less pronounced.The emission maxima of these compounds are independent of

the geometry about the amine but occur at a higher wavelengthfor analog 9c, in which the acceptor group is highly nonplanar.Emission maxima are also red-shifted with increasing solventpolarity (SI Appendix, Table S4). Large red shifts indicate anincreased dipole moment in the excited state compared with thatof the ground state, consistent with the formation of an intra-molecular charge-transfer excited state (ICT) (see SI Appendix)(20, 21). It should be noted that for strongly solvatochromaticcompounds, an anomalous blue shift in water is observed be-cause of interactions with the solvent cage. This blue-shiftingphenomenon has been long established in the literature (3).Fluorescence-emission intensities depend on the rate of ICT

excited-state depopulation. In viscous microenvironments orupon binding to amyloid aggregates, double-bond isomerizationor rotational relaxation is restricted and becomes much slower,leading to ∼10-times enhanced fluorescence yields over in bulksolution (ϕ; Table 1) (16).

Structure and Modeling. Single crystal X-ray diffraction providedsolid-state structures of DDNP, 5, 6b, 7b, 8b, 10b, and 11b.[The structures for DDNP have been deposited in the CSD,

www.ccdc.cam.ac.uk/products/csd (reference code NAFZUP).All available X-ray data, including coordinates, are available forall structures in SI Appendix.] In the case of 6b, two conformerswere present in the crystal. The distinct conformations are at-tributable to the orientation of the dicyanovinyl, which can berotated syn or anti with respect to the C6-C7 aromatic bond(Fig. 3). We have defined the torsion angle responsible for theorientation of the dicyanovinyl as ωtorsion. Analog 10b crystallizedinto two distinct crystal polymorphs, colored red and orange,which arises because of cyclohexenyl-ring conformations thatorient the dicyanovinyl either above or below the plane of thenaphthalene ring.The X-ray structures exhibit an sp2-like planar arrangement of

the substituents around the naphthalene ring, providing maximalconjugation of the nitrogen lone pair with the aromatic system.The X-ray structures revealed the syn conformation of 6b to beplanar and the anti to be slightly puckered. Azetidine rings, suchas the one in 6b, are nonplanar, with a slight 1.3 kcal/mol ni-trogen inversion barrier (22). However, planar azetidine ringshave been reported previously in other crystal structures in whichthe azetidine nitrogen is conjugated with a π system (23). Thepyrrolidine ring in 7b adopts an envelope conformation. Theother cyclic analogs do not allow planarization about the aminenitrogen without a higher energetic penalty. For these analogs,distortion from planarity at the amine results in a slight loss ofconjugation, which causes a lengthening of the bond distancebetween the amine nitrogen and aromatic carbon of the naph-thalene ring (N-Car; Table 1). The mean N-Car X-ray distancesfor the planar amines listed in Table 1 are 1.368 Å, and 1.393 Åfor the nonplanar amines. This difference in bond distances forthe planar and nonplanar compounds has been observed ina related set of DDNP analogs with structures determined byneutron diffraction (13). These molecules demonstrated N-Cardistances of 1.371 Å for the planar amines and 1.426 Å for thenonplanar amines.Quantum mechanical (QM) geometry optimizations were

performed at the M06–2X/6–311+G(d,p) level in the gas phase(details in Materials and Methods). The compounds were allmodeled in both the syn and anti orientations of the dicyanovinylgroup. The optimized structures agreed well with the geometriesof the available X-ray structures (Table 1 and SI Appendix, Fig.S1), yielding an average heavy atom rmsd of 0.15 Å. The QMgeometry optimizations showed that the syn and anti orientationsare nearly isoenergetic, with an average difference of 0.4 kcal/mol(SI Appendix, Table S1). It has been proposed previously that thedipole moment of DDNP is related to binding affinity (2, 24). Weanalyzed the charge distribution of these molecules by mappingthe electrostatic potential (ESP) to the molecular surface of eachanalog (SI Appendix, Fig. S2). The ESP surfaces of all of theanalogs are very similar, with the primary difference betweenthem being their shapes. This is most prominently highlighted bycomparing the two most planar analogs with the highest bindingaffinities, 10b and 11b, to the least planar analog with the lowestbinding affinity, 9c.

Amyloid Binding ModelA steric zipper pseudofibril of the VQIVYK tau sequence wasbuilt as a model in which to rationalize the relative binding af-finities of the analogs in Fig. 1 (see Materials and Methods). Twounique channels run along the fibril spine, which we have denotedas the tyrosine and lysine channels, named for the prominentamino acids characterizing the respective channels (SI Appendix,Fig. S3 and S4). Eisenberg and coworkers (8) observed that thenegatively charged fluorescence probe Orange-G preferentiallybound the electrostatically positive lysine channel, whereas theneutral probes, DDNP and curcumin, preferred the tyrosinechannel. The binding preferences were revealed by a registrationshift in β-sheet mating in the steric zipper, enlarging either the

Fig. 2. (Upper) Confocal fluorescence micrograph of a NFT and SP labeledwith 10 μM FDDNP vs. constrained DDNP (10b) and t-butyl DDNP (9c). Digi-tally captured image produced by laser-scanning Leica TCS SP MP invertedconfocal microscope with an argon laser (excitation wavelength, 488 nm).(Scale bar: 50 μm.) Note the poor labeling ability of the t-butyl analog 9c, asopposed to the more planar FDDNP and 10b. (Lower) Radioactive competi-tive binding curves of [18F]FDDNP vs. nonradioactive FDDNP (red circles), 10b(blue triangles), and 9c (green squares) in the presence of Aβ(1–40) fibrils.Each symbol represents the mean value of three determinations per eachconcentration of competitor. Error bars indicate ±SD.

16494 | www.pnas.org/cgi/doi/10.1073/pnas.1214134109 Petri�c et al.

Dow

nloa

ded

by g

uest

on

Janu

ary

22, 2

020

tyrosine channel, as in the case of DDNP and curcumin, or thelysine channel, as in the case of Orange-G, to facilitate binding.We have only considered binding to the tyrosine channel. Theprogram AutoDock Vina (25) was used to flexibly dock all of themolecules in Fig. 1 to the rigid VQIVYK pseudofibril model. Sidechains were kept rigid during the docking in the same fashion asfor Eisenberg and coworkers, because several computational testsshowed that the tightly packed environment of the fibril channeldoes not provide room for side-chain reorganization upon ligandbinding. This greatly diminishes the effect that an induced fitwould have on ligand binding.The docking results for each analog recapitulated the multiple

binding modes along the fibril spine observed in the X-raycrystallographic density (Fig. 4A). It is likely that DDNP and itsanalogs bind to localized sites along the fibril spine, which issuggested by the ability of the nonsteroidal anti-inflammatories(NSAIDs) naproxen and ibuprofen to displace FDDNP fromfibrils at low concentrations (5.7 nM and 11 μM, respectively)(26). However, because of the absence of well-resolved crystal-lographic density in the fibril channel and the lack of easilyidentifiable anchors on FDDNP and its analogs [such as the

negative charges of Orange-G that clearly place it proximal tothe lysine residues (8)], it is difficult to exactly delineate thepreferred binding mode. Analysis of the binding modes of eachanalog immediately justifies the little effect that drastic changesat the amine position have on binding affinity. The amine pointsdirectly along the spine of the binding channel, so changes in sizealong this axis can easily be accommodated without introducingsteric clashes with the sides of the binding channel (Fig. 4B). Thisis most prominently demonstrated by the similarity in bindingaffinity between 11b, containing a dimethyl-substituted amine,and 8b, containing a methylpiperidine ring. It was noted abovethat the large dipole moments of DDNP and its analogs (com-puted range of 8–12 Debye) could be important in binding. Thepolar contacts present between the bound probe, tyrosine oxygenatoms (Fig. 4 D and F), and valine N termini and stabilization ofthe negative (dicyanovinyl) and positive (amine) ends of themolecule could be significant factors in binding.Analysis of the top-ranking poses of each molecule showed

that the smaller and more planar molecules (e.g., 11c; Fig. 4 Cand D) better fit the binding channel than the larger molecules(e.g., 9c, Fig. 4 E and F). Molecules with larger substitutions atthe dicyanovinyl position prefer greater degrees of nonplanarityin the free global minimum conformation and, therefore, re-quired more distortion of the ωtorsion dihedral to prevent stericclashes of the C6 side chain with the binding channel uponbinding. We further examined the magnitude of the ωtorsion dis-tortion by performing QM/molecular mechanical (QM/MM)optimizations of the docked poses at the M06–2X/6–311+G(d,p):Universal Force Field (UFF) level of theory. The QM regionwas defined as the ligand, whereas the MM system was defined asthe rest of the system and held rigid during the calculation.During the QM/MM optimizations, noticeable differences in thepotential energy surfaces between the docking force field andQM/MM were observed. For example, the lowest-energy docked

Table 1. Selected structural, spectroscopic, and binding data for synthesized DDNP analogs

NMR Spectroscopic Binding X-ray Calculated

Compound δ (H-5) δ (H-7) λmax (abs*) λmax (em) ϕ† Δν (cm−1) Ki (nM){ NCar‡ (Å) Σωi

§ (°) ωtorsion (°) NCar (Å) ωtorsion (°) rms (Å)jj

DDNP 6.85 7.18 438 560 0.030 4,970 10 1.362 359.9 31.0 1.375 37.1 0.101.375 360.0 143.7 1.374 143.8 0.11

FDDNP 6.95 7.20 431 553 0.016 5,065 0.20 n/d n/d n/d 1.388 37.7 —

1.388 143.2 —

4b 6.82 7.12 455 564 0.021 4,250 0.46 n/d n/d n/d 1.372 36.2 —

1.372 145.1 —

6b 6.58 6.81 434 567 0.018 5,400 0.43 1.369 359.8 33.3 1.375 37.3 0.251.374 348.4 152.6 1.374 143.8 0.10

7b 6.71 7.03 455 566 0.020 4,310 0.19 n/o n/o n/o 1.364 36.3 —

1.369 359.9 148.8 1.364 145.1 0.098b 7.07 7.33 428 569 0.014 5,790 0.30 n/o n/o n/o 1.401 37.9 —

1.411 347.8 146.0 1.402 143.1 0.259c 6.91 7.04 374 595 0.013 9,930 520 n/d n/d n/d 1.390 75.4 —

1.390 110.6 —

10b** 6.84 7.18 443 541 0.006 4,090 0.01 1.377 360.0 35.6 1.377 31.9 0.15†† 1.366 359.2 −10.3 1.373 −29.0 0.2811b 6.83 7.16 470 566 0.137 3,610 0.12 1.360 360.0 2.9 1.369 0.0 0.06

n/o n/o n/o 1.369 180.0 —

n/d, not determined; n/o, not observed; —, uncalculated value.*Absorption (abs) and emission (em) measured in CH2Cl2 (nm), a solvent that mimics the microenvironment of the Aβ fibril-binding site.†Quantum yields determined relative to DDNP.‡Distance between amine nitrogen and aromatic naphthalene carbon.§Sum of bond angles about amino nitrogen.{High binding affinity to Aβ fibrils. Binding affinities were determined in PBS (pH 7.4) containing up to 1% ethanol.jjHeavy atom rmsd with respect to the X-ray structure.**Orange crystal.††Red crystal.

N

R

anti

NC CN

N

R

syn

CN

CN

torsion

56

7

88a

1

2

34

4a

Fig. 3. Syn and anti orientations of the dicyanovinyl group with respect tothe C7-C6 aromatic bond. The syn conformation is defined as 0° ≤ jωtorsionj <90° and the anti conformation as 90° ≤ jωtorsionj ≤ 180°.

Petri�c et al. PNAS | October 9, 2012 | vol. 109 | no. 41 | 16495

BIOCH

EMISTR

YCH

EMISTR

Y

Dow

nloa

ded

by g

uest

on

Janu

ary

22, 2

020

pose of 11b in the tyrosine channel had a ωtorsion value of −37.3°and an out-of-plane twisting of the amine group on the naph-thalene, whereas after QM/MM optimization, ωtorsion decreasedto 1.5° and was also accompanied by a planarization of the aminegroup. The planarization of ωtorsion after the QM/MM optimi-zation corresponds to a lower distortion energy on the M06–2X/6–311+G(d,p) potential energy surface (SI Appendix, Fig. S5).Upon examination of all of the QM/MM optimized com-

plexes, a qualitative trend was found between the energy re-quired to distort ωtorsion and the binding affinity of the molecule(Fig. 5). The constrained DDNP analog 10b is locked into a near-planar conformation and has the highest binding affinity. Analog11b, which has a hydrogen substitution at the dicyanovinyl posi-tion, prefers a planar conformation and, correspondingly, alsohas a high binding affinity. The molecules with a methyl sub-stitution at the dicyanovinyl position, 4b, 6b, 8b, FDDNP, andDDNP, fall within a similar region of the graph. DDNP has 50-fold lower affinity than FDDNP and the other methyl-substitutedanalogs (10 vs. ∼0.2 nM). The program QikProp (version 3.0.001w;Schrodinger) was used to compare the molecular properties ofthese compounds (SI Appendix, Table S2). DDNP has notablyless hydrophobic surface area and thus a lower predicted water/

octanol partition coefficient. Similar analysis can also be appliedto distinguish the difference in binding affinity between 10b and11b. Both of these molecules require little distortion to fit thebinding pocket, yet 11b has both a lower predicted log(P) valueand less hydrophobic contact surface area (SI Appendix, TableS3). Analog 9c had the lowest binding affinity but also requiredsignificantly more distortion energy to fit the fibril channel thanany of the other analogs.

ConclusionsA series of FDDNP analogs have been synthesized (Fig. 1) andcharacterized using NMR, spectroscopic, and computationalmethods. Two of these molecules, 10b and 11b, showed improvedaffinity to amyloid fibrils over the parent molecule FDDNP. Im-proved binding affinities are essential for imaging probes to vi-sualize and appropriately quantify amyloid aggregation withinthe tissue target. Through the use of the steric zipper–bindingmodel for DDNP bound to the VQIVYK segment of tau (8), thedifferences in relative binding affinities of these imaging probeshas been attributed to the distortion required for the moleculesto fit within the binding channels that run along the fibril spine.Molecules with larger substitutions at the dicyanovinyl position,such as 9c, preferred highly nonplanar conformations in solutionand required the largest magnitude of distortion to fit withinthe binding channel. Efforts to (i) modify (lower or increase) thebinding affinities to some amyloids (e.g., Aβ) and (ii) providea differential binding modification between amyloids (e.g., Aβand tau aggregates) could produce imaging probes with selectivesensitivity and specificity for different imaging agents. This approachoffers a framework for fine-tuning the binding properties ofneurofibrillary tau-specific (or Aβ-specific) imaging probes inparallel with X-ray microcrystallography at atomic resolution incocrystallization experiments.

Materials and MethodsSyntheses of Naphthalene Derivatives. In preparing fluorescent compounds4–11 (Fig. 1), the 6-acyl-2-naphthols 1–3 were subjected to the Buchererreaction with open-chain and cyclic amines that are not sensitive to slightlyacidic aqueous environment. Aziridine and azetidine rings are too reactiveunder the Bucherer reaction conditions, and, therefore, the ketones 5 and6a were prepared by direct nucleophilic substitution of the methoxy groupin (6-methoxy-2-naphthyl)methyl ketone with the respective aziridine andazetidine lithium salts. In the last step of the synthesis, the ketones weresubjected to the Knoevenagel reaction with malononitrile to yield thedicyanovinyl naphthalene analogs. Because of the high reactivity of theaziridine ring, ketone 5 could not be transformed into the expected product,and decomposition occurred. Naphthaldehyde 11a, prepared by a modifica-tion of a known procedure (27), gave its Knoevenagel product in good yield.Full experimental details for the synthesis and characterization of eachmolecule are described in SI Appendix.

Fig. 4. (A) Docking of each dicyanovinyl analog from Fig. 1 yieldeda “smear” of poses down the binding channel. (B) The channel accom-modates planar molecules of different sizes at the amine position, as shownby the substitution of a methylpiperidine in 8b for the dimethylamino groupin 11b, which have similar affinities. Differences in binding affinity can berationalized by the ability of the smaller and more planar analogs, such as11b (C and D), to fit within the channel with minimal distortion. Larger andless planar molecules, such as 9c (E and F), require greater distortion to fitwithin the pocket. In C and F, solid spheres show the atomic radii of the li-gand and the transparent surface represents the molecular surface of theprotein. Dashed lines represent polar contacts within 3.0 Å between the li-gand and pseudofibril.

4b, 6b, 8b,FDDNP11b

10b

DDNP

9c

7b

Ki (

nM

)

0.01

1.00

100.00

Edist (kcal/mol)0 0.5 1.0 1.5 2.0 2.5 3.0

Fig. 5. Plot of the Ki vs. the energy to distort ωtorsion from the global minimumgas-phase conformation to the value in the QM/MM optimized structure.

16496 | www.pnas.org/cgi/doi/10.1073/pnas.1214134109 Petri�c et al.

Dow

nloa

ded

by g

uest

on

Janu

ary

22, 2

020

QM Structure Calculations. FDDNP analogs from Fig. 1 with X-ray structures ineither the syn or anti conformation were directly optimized to the nearestground-state minimum using M06–2X/6–311+G(d,p) in Gaussian09 (28) withtight convergence. For analogs with a crystal structure of only the syn or anticonformation, the relationship ωsyn ∼ 180 − ωanti was used to estimate thestarting structure. This relationship held quite well for the optimized struc-tures (R2 = 0.99; SI Appendix, Fig. S2). For analogs without X-ray structuraldata, the global minimum conformation was located using a 5,000-stepMonte Carlo Multiple Minimum (MCMM) gas-phase conformational searchwith OPLS2005 performed in MacroModel (version 9.9.111; Schrodinger) wasused as the starting point. Geometries of the QM-calculated and X-raystructures were superimposed using Maestro (version 9.2; Schrodinger) giv-ing an average heavy atom RMSD of 0.15 Å (SI Appendix, Fig. S1).

Amyloid Binding Model Construction. The biological unit of the VQIVYK stericzipper was obtained from ref. 8 without DDNP explicitly modeled. The bi-ological unit was duplicated into a 4 × 3 × 8 steric zipper with PyMOL (version1.3; Schrodinger) using the crystallographic symmetry information includedwith themodel. This was done in an identical manner as that of Eisenberg andcoworkers (8), and a superposition of ourmodel to theirs shows they are nearlyidentical, with corresponding atom–atom deviations of ≤ 0.5 Å (SI Appendix,Fig. S2).We constructed a largermodel tomore comprehensively include long-range electrostatics in both the docking and QM/MM simulations.

Molecular Docking and QM/MM Calculations. AutoDock Vina was used toperform the docking simulations (25) with the QM-optimized analogs

serving as the input ligands. The receptor was prepared for docking usingthe PyMOL AutoDock plugin (29). Tyrosine and lysine channels located to-ward the center of the model were chosen for docking, each with a 12 × 39 ×21 Å rectangular docking region, more than large enough to encompasseach entire channel separately. The pseudofibril was held rigid during bothdocking and QM/MM optimization, because the close packing of the aminoacid residues within the channel prevents the side chains from reorganizingupon ligand binding. Individual docking simulations were performed foreach channel. Default docking parameters were used for AutoDock Vina,and the top-ranked/lowest-energy docked conformation was used as thebest pose. We confirmed that the best pose for each analog was locatedtoward the center of each channel. This pose served as the input to the QM/MM calculations performed in Gaussian09 (28) at the M06–2X/6–311+G(d,p):UFF level of theory.

ACKNOWLEDGMENTS. We thank Drs. B. Kralj and D. Zigon (Jozef StefanInstitute) for MS measurements. This work was supported by the Ministry ofHigher Education, Science, and Technology of the Republic of Slovenia andSlovenian Research Agency Grant P1-0230-0103; US Department of EnergyGrant DE-FC0837-ER60615; the Korean Research Foundation [Ministry ofEducation and Human Resources Development (MOEHRD) Grant KRF-2006-611-C00004 (to G.K.)]; National Institutes of Health (NIH) Grant P01AG025831;and National Institute of General Medical Sciences/NIH Grant GM36700.S.A.J. and H.V.P. are supported by US Public Health Service (USPHS) NationalResearch Service Award GM08496. J.R.B. is supported by The Elizabeth andThomas Plott Chair Endowment in Gerontology.

1. Goedert M, Spillantini MG (2006) A century of Alzheimer’s disease. Science 314:777–781.

2. Bresjanac M, et al. (2003) Molecular-imaging probe 2-(1-[6-[(2-fluoroethyl)(methyl)amino]-2-naphthyl]ethylidene) malononitrile labels prion plaques in vitro. J Neurosci23:8029–8033.

3. Jacobson AF, Petri�c A, Hogenkamp D, Sinur A, Barrio JR (1996) 1,1-Dicyano-2-[6-(di-methylamino)naphthalen-2-yl]propene (DDNP): A solvent polarity and viscosity sen-sitive fluorophore for fluorescence microscopy. J Am Chem Soc 118:5572–5579.

4. Liu J, et al. (2007) High-yield, automated radiosynthesis of 2-(1-6-[(2-[18F]fluoroethyl)(methyl)amino]-2-naphthylethylidene)malononitrile ([18F]FDDNP) ready for animal orhuman administration. Mol Imaging Biol 9:6–16.

5. Small GW, et al. (2006) PET of brain amyloid and tau in mild cognitive impairment.N Engl J Med 355:2652–2663.

6. Shoghi-Jadid K, et al. (2002) Localization of neurofibrillary tangles and beta-amyloidplaques in the brains of living patients with Alzheimer disease. Am J Geriatr Psychiatry10:24–35.

7. Teplow DB (1998) Structural and kinetic features of amyloid beta-protein fibrillo-genesis. Amyloid 5:121–142.

8. Landau M, et al. (2011) Towards a pharmacophore for amyloid. PLoS Biol 9:e1001080.9. Ivanova MI, Thompson MJ, Eisenberg D (2006) A systematic screen of beta(2)-micro-

globulin and insulin for amyloid-like segments. Proc Natl Acad Sci USA 103:4079–4082.10. Gazit E (2005) Mechanisms of amyloid fibril self-assembly and inhibition. Model short

peptides as a key research tool. FEBS J 272:5971–5978.11. Gazit E (2002) Mechanistic studies of the process of amyloid fibrils formation by the

use of peptide fragments and analogues: Implications for the design of fibrillizationinhibitors. Curr Med Chem 9:1725–1735.

12. Sievers SA, et al. (2011) Structure-based design of non-natural amino-acid inhibitorsof amyloid fibril formation. Nature 475:96–100.

13. Petri�c A, Spes T, Barrio JR (1998) Novel fluorescent reactive dyes as intermediates forthe preparation of UV and Vis wavelength fluorescent probes. Monatsh Chem 129:777–786.

14. Petri�c A, Jacobson AF, Barrio JR (1998) Functionalization of a viscosity-sensitive flu-orophore for probing of biological systems. Bioorg Med Chem Lett 8:1455–1460.

15. Selkoe DJ (1994) Cell biology of the amyloid beta-protein precursor and the mecha-nism of Alzheimer’s disease. Annu Rev Cell Biol 10:373–403.

16. Agdeppa ED, et al. (2001) Binding characteristics of radiofluorinated 6-dialkylamino-

2-naphthylethylidene derivatives as positron emission tomography imaging probes

for beta-amyloid plaques in Alzheimer’s disease. J Neurosci 21:RC189.17. Krasnowska EK, Gratton E, Parasassi T (1998) Prodan as a membrane surface fluo-

rescence probe: Partitioning between water and phospholipid phases. Biophys J 74:

1984–1993.18. Petri�c A, Barrio JR (1994) Synthesis and NMR Characterization of 4-[(2-tetrahy-

drophyranyloxy)-methyl]piperidine and Intermediates. J Heterocycl Chem 31:545–548.19. Bure�s F, et al. (2007) Property tuning in charge-transfer chromophores by systematic

modulation of the spacer between donor and acceptor. Chemistry 13:5378–5387.20. Reichardt C, Welton T (2003) Solvents and Solvent Effects in Organic Chemistry (Wi-

ley-VCH, New York), 3rd Ed.21. Volchkov VV, Uzhinov BM (2008) Structural relaxation of excited molecules of het-

eroaromatic compounds. High Energy Chem 42:153–169.22. Bansal RK (1999) Heterocyclic Chemistry (New Age International, New Delhi, India).23. Bartnik R, Faure R, Gebicki K (1998) Synthesis and crystal structure of 1-acetyl-3-bromo-

3-phenylazetidine and 1-phenyl-2-(N-acetyl-N-formyl)-aminoethanone. J Chem Crys-

tallogr 28:119–123.24. �Skofic P, et al. (2005) Syntheses of 4-(2-naphthyl)pyridine derivatives from DDNP. Acta

Chim Slov 52:391–397.25. Trott O, Olson AJ (2010) AutoDock Vina: Improving the speed and accuracy of

docking with a new scoring function, efficient optimization, and multithreading. J

Comput Chem 31:455–461.26. Agdeppa ED, et al. (2003) In vitro detection of (S)-naproxen and ibuprofen binding to

plaques in the Alzheimer’s brain using the positron emission tomography molecular

imaging probe 2-(1-6-[(2-[18F]fluoroethyl)(methyl)amino]-2-naphthylethylidene)ma-

lononitrile. Neurosci 117:723–730.27. Umezawa H, et al. (2000) Synthesis of stilbazolium derivatives having 2-(6-dimethyl

amino) naphthyl group for nonlinear optics. Mol Cryst Liq Cryst Sci Technol B 22:

251–254.28. Frisch MJ, et al. (2009) Gaussian 09 Revision B.01 (Gaussian, Wallingford CT).29. Seeliger D, de Groot BL (2010) Ligand docking and binding site analysis with PyMOL

and Autodock/Vina. J Comput Aided Mol Des 24:417–422.

Petri�c et al. PNAS | October 9, 2012 | vol. 109 | no. 41 | 16497

BIOCH

EMISTR

YCH

EMISTR

Y

Dow

nloa

ded

by g

uest

on

Janu

ary

22, 2

020

Corrections

BIOCHEMISTRY, CHEMISTRYCorrection for “Dicyanovinylnaphthalenes for neuroimaging ofamyloids and relationships of electronic structures and geome-tries to binding affinities,” by Andrej Petri�c, Scott A. Johnson,Hung V. Pham, Ying Li, Simon �Ceh, Amalija Golobi�c, Eric D.Agdeppa, Gerald Timbol, Jie Liu, Gyochang Keum, NagichettiarSatyamurthy, Vladimir Kepe, Kendall N. Houk, and Jorge R.Barrio, which appeared in issue 41, October 9, 2012, of Proc NatlAcad Sci USA (109:16492–16497; first published September 25,2012; 10.1073/pnas.1214134109).The authors note that an additional affiliation should be listed

for Gyochang Keum. The new affiliation should appear as Centerfor Neuro-Medicine, Brain Science Institute, Korea Institute ofScience and Technology, Hwarangro 14-gil 5 Seoungbuk-gu, Seoul136-791, Republic of Korea. The corrected author and affiliationlines appear below. The online version has been corrected.

Andrej Petri�ca,b, Scott A. Johnsonc, Hung V. Phamc,Ying Lic, Simon �Ceha, Amalija Golobi�ca, Eric D. Agdeppad,Gerald Timbold, Jie Liud, Gyochang Keumd,e,Nagichettiar Satyamurthyd, Vladimir Keped,Kendall N. Houkc, and Jorge R. Barriod

aDepartment of Chemistry and Chemical Technology, University of Ljubljana,1000 Ljubljana, Slovenia; bEN-FIST Centre of Excellence, Dunajska 156, SI-1000Ljubljana, Slovenia; cDepartment of Chemistry and Biochemistry, University ofCalifornia, Los Angeles, CA 90095; dDepartment of Molecular and MedicalPharmacology, The David Geffen School of Medicine, University of California,Los Angeles, CA 90095; and eCenter for Neuro-Medicine, Brain ScienceInstitute, Korea Institute of Science and Technology Hwarangro 14-gil 5Seoungbuk-gu, Seoul 136-791, Republic of Korea

www.pnas.org/cgi/doi/10.1073/pnas.1302644110

CELL BIOLOGYCorrection for “Increased lipolysis and altered lipid homeostasisprotect γ-synuclein–null mutant mice from diet-induced obesity,”by Steven Millership, Natalia Ninkina, Irina A. Guschina, JessicaNorton, Ricardo Brambilla, Pieter J. Oort, Sean H. Adams,Rowena J. Dennis, Peter J. Voshol, Justin J. Rochford, andVladimir L. Buchman, which appeared in issue 51, December 18,2012, of Proc Natl Acad Sci USA (109:20943–20948; first pub-lished December 3, 2012; 10.1073/pnas.1210022110).The authors note that the author name Ricardo Brambilla

should instead appear as Riccardo Brambilla. The corrected au-thor line appears below. The online version has been corrected.

Steven Millership, Natalia Ninkina, Irina A. Guschina,Jessica Norton, Riccardo Brambilla, Pieter J. Oort,Sean H. Adams, Rowena J. Dennis, Peter J. Voshol,Justin J. Rochford, and Vladimir L. Buchman

www.pnas.org/cgi/doi/10.1073/pnas.1302920110

IMMUNOLOGYCorrection for “Previously undescribed grass pollen antigens arethe major inducers of T helper 2 cytokine-producing T cells inallergic individuals,” by Véronique Schulten, Jason A. Greenbaum,Michael Hauser, Denise M. McKinney, John Sidney, Ravi Kolla,Cecilia S. Lindestam Arlehamn, Carla Oseroff, Rapheul Alam,David H. Broide, Fatima Ferreira-Briza, Howard M. Grey,Alessandro Sette, and Bjoern Peters, which appeared in issue 9,February 26, 2013, of Proc Natl Acad Sci USA (110:3459–3464;first published February 11, 2013; 10.1073/pnas.1300512110).The authors note that the author name Fatima Ferreira-Briza

should instead appear as Fatima Ferreira. The corrected authorline appears below. The online version has been corrected.

Véronique Schulten, Jason A. Greenbaum,Michael Hauser, Denise M. McKinney, John Sidney,Ravi Kolla, Cecilia S. Lindestam Arlehamn, Carla Oseroff,Rapheul Alam, David H. Broide, Fatima Ferreira,Howard M. Grey, Alessandro Sette, and Bjoern Peters

www.pnas.org/cgi/doi/10.1073/pnas.1303818110

IMMUNOLOGYCorrection for “Lymphocyte-derived ACh regulates local innatebut not adaptive immunity,” by Colin Reardon, Gordon S.Duncan, Anne Brüstle, Dirk Brenner, Michael W. Tusche, PederOlofsson, Mauricio Rosas-Ballina, Kevin J. Tracey, and Tak W.Mak, which appeared in issue 4, January 22, 2013, of Proc NatlAcad Sci USA (110:1410-1415; first published January 7, 2013;10.1073/pnas.1221655110).The authors note that the author name Peder Olofsson should

instead appear as Peder S. Olofsson. The corrected author lineappears below. The online version has been corrected.

Colin Reardon, Gordon S. Duncan, Anne Brüstle,Dirk Brenner, Michael W. Tusche, Peder S. Olofsson,Mauricio Rosas-Ballina, Kevin J. Tracey, and Tak W. Mak

www.pnas.org/cgi/doi/10.1073/pnas.1302921110

NEUROSCIENCECorrection for “Representational gain in cortical area underliesincrease of memory strength,” by Kasia M. Bieszczad andNorman M. Weinberger, which appeared in issue 8, February 23,2010, of Proc Natl Acad Sci USA (107:3793–3798; first publishedFebruary 4, 2010; 10.1073/pnas.1000159107).The authors note that the National Institutes of Health Grant

DC-02938 should instead appear as DC-010013.

www.pnas.org/cgi/doi/10.1073/pnas.1303439110

www.pnas.org PNAS | March 26, 2013 | vol. 110 | no. 13 | 5269

CORR

ECTIONS