Alzheimer's disease: Identification and development of β-secretase (BACE-1) binding fragments and inhibitors by dynamic ligation screening (DLS)

CHEMMEDCHEM

FULL PAPERS

DOI: 10.1002/cmdc.201300078

Alzheimer’s Disease: Identification and Development of b-

Secretase (BACE-1) Binding Fragments and Inhibitors by

Dynamic Ligation Screening (DLS)

Marꢀa Isabel Fernꢁndez-Bachiller,^[a]Andrꢂ Horatscheck,^[a]Michael Lisurek,^[a]and

Jçrg Rademann*^{[a, b]}

The application of dynamic ligation screening (DLS), a method-

ology for fragment-based drug discovery (FBDD), to the aspart-

ic protease b-secretase (BACE-1) is reported. For this purpose,

three new fluorescence resonance energy transfer (FRET) sub-

strates were designed and synthesized. Their kinetic parame-

ters (V_max, K_M, and k_cat) were determined and compared with

a commercial substrate. Secondly, a peptide aldehyde was de-

signed as a chemically reactive inhibitor (CRI) based on the

Swedish mutation substrate sequence. Incubation of this CRI

with the protease, a FRET substrate, and one amine per well

taken from an amine library, which was assembled by a maxi-

mum common substructure (MCS) approach, revealed the frag-

ment 3-(3-aminophenyl)-2H-chromen-2-one (1) to be a compet-

itive BACE-1 inhibitor that enhanced the activity of the CRI. Ir-

reversibly formed fragment combination products of 1 with

the initial peptide sequence were active and confirmed the tar-

geting of the active site through the ethane-1,2-diamine iso-

stere. Finally, structure-assisted combination of fragment

1 with secondary fragments that target the S1 site in hit opti-

mization yielded novel, entirely fragment-based BACE-1 inhibi-

tors with up to 30-fold improved binding affinity. Interactions

with the protein were explained by molecular modeling stud-

ies, which indicate that the new fragment combinations inter-

act with the catalytic aspartic acid dyad, as well as with the ad-

jacent binding sites required for potency.

Introduction

Alzheimer’s disease (AD) is a chronic neurodegenerative disor-

der of the central nervous system, which results in severe cog-

nitive deficits along with psychiatric complications and repre-

sents the most common form of dementia. Today, more than

37 million individuals in the world suffer from this condition;

the number is expected to continue growing dramatically and

it is predicted to reach 80 million by 2040, with AD being rec-

ognized as a major global social and financial burden.^[1,2]Cur-

rently, there is no treatment available for AD and medical

therapies are usually limited to acetylcholinesterase (AChE) in-

hibitors and N-methyl-d-aspartate (NMDA) antagonists. In this

context, five drugs are currently on the market, the cholinester-

ase inhibitors donepezil, rivastigmine, galantamine, and ta-

crine, supplemented by the NMDA receptor modulator mem-

antine. All of these clinically admitted inhibitors, however,

were only approved for the symptomatic treatment of AD and

are unable to halt or reverse the disease progression.

The major pathological hallmark of AD is the deposition and

aggregation of amyloid-b peptide (Ab40,42) in the brain tissue,

which produces the extracellular formation of senile plaques.^[3]

These Ab peptides consist in fragments of 39–43 amino acid

residues and are produced under pathological conditions by

the sequential action of two proteases, named b- and g-secre-

tase, on amyloid precursor protein (APP).^[4]The b-secretase

cleavage of APP is the first step in the generation of Ab pep-

tides. In this context, b-secretase, also called b-amyloid-con-

verting enzyme 1 (BACE-1), has been identified as the rate-lim-

iting step in the production of Ab peptides and the main ap-

proach for avoiding Ab formation involves targeting b-secre-

tase.^[5,6]In addition, g-secretase inhibitors have been found to

interfere with the Notch signaling pathway and were thus dis-

qualified from drug use.^[7]For these reasons, the inhibition of

BACE-1 is considered to be one of the most prominent targets

for intervention in AD.^[8,9]Over the past decade, many BACE-

1 inhibitors have been reported and these inhibitors are divid-

ed into two classes, peptidomimetic and non-peptidomimetic

inhibitors.^[10]Although the first class of inhibitors showed the

highest potency against BACE-1, their relatively large molecular

size, low metabolic stability, and poor bioavailability render

their development into therapeutic drug candidates diffi-

cult.^[11,12]On the other hand, a great deal of effort has been

put into the discovery of nonpeptidic organic compounds with

better pharmacological properties as drugs leads.^[13–15]Some of

these compounds have been identified as potent BACE-1 inhib-

itors with balanced adsorption, distribution, metabolism, and

[a] Dr. M. I. Fernꢀndez-Bachiller, Dr. A. Horatscheck, Dr. M. Lisurek,

Prof. Dr. J. Rademann

Medicinal Chemistry Department

Leibniz Institut fꢁr Molekulare Pharmakologie (FMP) Campus Berlin-Buch

Robert-Rçssle Str. 10, 13125 Berlin (Germany)

[b] Prof. Dr. J. Rademann

Free University Berlin, Medicinal Chemistry, Institute of Pharmacy

Kçnigin-Luise Str. 2+4, 14195 Berlin (Germany)

E-mail: joerg.rademann@fu-berlin.de

Supporting information for this article is available on the WWW under

http://dx.doi.org/10.1002/cmdc.201300078.

ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemMedChem 2013, 8, 1041 – 1056 1041

CHEMMEDCHEM

FULL PAPERS

www.chemmedchem.org

excretion toxicity (ADMET) properties, which will make the

chemicals more favorable for further development into leads

and drugs.^[16]

a nucleophilic fragment (for example, an amine, VI) at the CRI

aldehyde functionality in presence of the protein to yield, re-

versibly, a dynamic ligation product VIII, which competed with

a fluorogenic substrate for binding to the protein surface. For

an aspartic protease like BACE-1, the assay has to be modified

at several points. Peptide aldehyde inhibitors cannot react

with an aspartic protease under the formation of a hemithioa-

cetal. Instead, the aldehyde hydrate has to be formed and

binds to both aspartic acid residues of the active site (VII)

through hydrogen bonding.^[29]Dynamic ligation of the peptide

aldehyde with an amine nucleophile can nevertheless form an

imine product reversibly, which can possibly be hydrated into

the respective hemiaminal in equilibrium. Both products

should be able to interact with the active site and with other

binding pockets close to the active site like the imine bound

by the cysteine proteases. In addition, the substrate must be

modified for the BACE-1 assay. Whereas 7-aminomethylcou-

marin peptides served as fluorogenic substrates for cysteine

proteases, BACE-1 requires fluorescence resonance energy

transfer (FRET) substrates, in which the fluorescence of a donor

fluorophore is suppressed by resonance energy transfer to

a quencher dye. After the substrate has been cleaved by the

protein, the donor fluorophore and the quencher are separat-

ed and the fluorescence of the fluorogenic part is restored (IX).

Thus, to extend the DLS methodology from cysteine proteases

to an aspartic protease such as b-secretase (BACE-1) and apply

it for the identification of new chemical entities (NCEs) or frag-

ments as BACE-1 inhibitors, the following steps have to be

taken: a) the design, synthesis, and kinetic parameter determi-

nation for new FRET BACE-1 substrates I–III, b) the design and

synthesis of peptide aldehyde V as the chemically reactive in-

hibitor, c) assay optimization, d) assay validation, e) hit identifi-

cation, f) hit validation with a secondary assay, and finally,

g) hit-to-lead optimization.

Generally, these nonpeptidic inhibitors have emerged by the

implementation of high-throughput screening (HTS) programs

and, more recently, due to the application of fragment-based

drug discovery (FBDD) methodologies.^[17–20]In the latter case,

different detection methods have been used, such as X-ray

analysis, nuclear magnetic resonance (NMR) spectroscopy, and

surface plasmon resonance (SPR) spectroscopy. These biophysi-

cal methods directly provide thermodynamic and, at least par-

tially, structure information on the fragment–protein interac-

tion. However, the low throughput and the high time and pro-

tein consumption are limiting factors for the application of

each technique for high-throughput screening (HTS). Moreover,

these biophysical methods have been demonstrated to pro-

vide low-affinity binders of proteins; however, they do not pro-

vide information about the linking of primary fragment hits to

transform low-affinity fragments into high-affinity fragment

combinations. Therefore, exhaustive synthetic efforts are re-

quired for the chemical optimization of primary fragment

hits.^[21–25]

More recently, dynamic ligation screening (DLS) has been in-

troduced as a method in FBDD that combines dynamic, target-

assisted formation of fragment combinations as inhibitory spe-

cies with detection through a biochemical assay.^[26,27]For exam-

ple, the enzymatic turnover of a fluorogenic substrate has

been used to amplify the active fragment combination, which

allowed sensitive detection of low-affinity fragments with

a drastic decrease in the amount of protein required.^[26]In ad-

dition, DLS has enabled the implementation of this methodol-

ogy in an HTS format (Figure 1). In the initial setup, the

method was established for cysteine proteases. Peptide alde-

hydes V were used as chemically reactive inhibitors (CRIs). The

CRIs served as directing probes; they were able to bind to the

protease’s active cleft through backbone and side-chain inter-

actions while the aldehyde electrophile was attacked by the

thiolate nucleophile of the active site cysteine residue to yield

a hemithioacetal intermediate in equilibrium.^[28]The reversibili-

ty of the hemithioacetal formation finally permitted attack of

Results and Discussion

BACE-1 FRET substrates I–III: design, synthesis, and determi-

nation of kinetic parameters

To facilitate the research on spe-

cific BACE-1 inhibitors and to fa-

cilitate studies of the BACE-1 en-

zymology, it is necessary to have

a sensitive and continuous assay

for enzyme activity. For these

reasons, fluorogenic peptide

substrates that contain suitable

fluorophore

and

quencher

groups in the same molecule

should be developed. In this

context, the FRET substrates I–III

were synthesized. Their kinetic

parameters were determined

and compared with the commer-

cial substrate IV with a similar

structure. The new FRET sub-

Figure 1. The concept of dynamic ligation screening (DLS) applied to the identification of BACE-1 inhibitors.

ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemMedChem 2013, 8, 1041 – 1056 1042

CHEMMEDCHEM

FULL PAPERS

www.chemmedchem.org

strates were designed based on a peptide containing ten

amino acid residues of the putative substrate belonging to the

Swedish mutant (SEVNLDAEFR; Michaelis–Menten constant

were investigated. Thus, various amounts of the protein (25–

100 ng; 8.6–74.4 nm) were incubated for 60 min at 258C with

25 mm substrate I and the development of the product was

monitored through continuous measurements every minute.

After these experiments, 25 ng (18.6 nm, final concentration) of

the protein were used for the end-point estimation for the

next experiments. The experimental conditions such as pH

value and detergent were adapted from published data.^[31,32]

With these results in mind, substrates I–III were digested in

vitro with BACE-1 and their kinetic parameters were deter-

mined and compared with those of the commercial substrate

IV with a similar amino acid sequence but with a different fluo-

rophore and quencher pair (Table 1). The initial velocity (V) for

each substrate was calculated from the linear part of the fluo-

rescence curve recorded over the time between 1–20 min. The

K_Mand maximal velocity (V_max) values were determined by

using the Michaelis–Menten representation. The k_catvalue was

calculated from the equation k_cat=V_max/[BACE-1]_total, in which

[BACE-1]_totalis the total enzyme concentration in mm. The kinet-

ic parameters for the cleavage of these fluorogenic substrates

are represented in Table 1. Relative to the commercial sub-

strate IV, our best substrate I displayed a twofold increase in

the substrate affinity (K_M) and an almost fivefold increase in

the rate of the enzymatic reaction (k_cat/K_M), which determines

the sensitivity of the substrate in the enzyme assay.

(K_M)=9 mm, catalytic rate (k_cat)=0.002 s )

^{À1 [30]}and were modified

by the introduction of a fluorophore (7-methoxy-coumarin-4-

yl-acetyl (MOCAc) or 7-amino-4-methylcoumarin-2-yl-acetyl

(AMCA)) and a quencher like 2,4-dinitrophenyl (DNP), 2,4-dini-

trobenzoyl (DNB), or 2-(2,4-dinitrophenylamino)acetyl (DNPA)

at the N- and C-terminal positions, respectively (Table 1). The

Table 1. Kinetic parameters for the hydrolysis of the synthesized BACE-

1 fluorescence substrates I–III mediated by the BACE-1 full protein and in

comparison with commercial substrate IV.^[a,b]

Substrate

V_max

[mms^À1]ꢄ10^À2

K_M

[mm]

k_cat

]

k_cat/K_M

[s^À1m^À1

]

[s^À1

I^[c]

0.142Æ0.016

0.092Æ0.003

0.398Æ0.006

0.053Æ0.005

0.038Æ0.002

4.97Æ1.01

6.71Æ1.51

40.2Æ3.05

128Æ12.0

9.27Æ1.56

12.8Æ2.53

0.08

0.05

0.26

0.215

0.03

0.02

15687

7705

6388

28

3405

2078

I^[d]

II^[c]

III^[c,e]

IV^[c]

IV^[d]

[a] The data points are an average of three independent experiments.

Substrate I: (MOCAc)SEVNLDAEFK(DNPA)RR; II: (MOCAc)SEVNLDAEFK-

(DNB)RR; III: (DNPA)SEVNLDAEFRRG(AMCA); IV: (MOCAc)SEVNLDAEFRK-

(DNP)RR. The P1 and P1’ residues are shown in bold. [b] Unless otherwise

stated, the enzyme reaction was carried out in 100 mm sodium acetate

(pH 4.0) containing 0.01% Triton X-100, in the presence of 5–100 mm sub-

strate and 18.6 nm enzyme (25 ng) with 5 or 10% DMSO at room temper-

ature. [c] 5% DMSO. [d] 10% DMSO. [e] The enzyme reaction was carried

out in 100 mm sodium acetate (pH 4.0) containing 0.01% Triton X-100, in

the presence of 5–500 mm substrate and 148.8 nm enzyme (200 ng) with

5% DMSO at room temperature.

This effect was achieved by replacement of 2,4-dinitrophenyl

(DNP) with 2-(2,4-dinitrophenylamino)acetyl (DNPA) as

a quencher in the C-terminal substrate position. For compari-

son, the kinetic parameters for our first FRET substrate III,

which employed the synthetically favorably DNPA/AMCA pair

in the reversed positioning, were decreased by a factor of 26

in terms of substrate affinity and by 560-fold with respect to

the rate of the enzymatic reaction. Likewise, the substitution of

a 2-(2,4-dinitrophenylamino)acetyl moiety by a dinitrobenzoyl

group increased the K_Mvalue eightfold and reduced the k_cat/K_M

value 2.5-fold (compare substrates I and II). The kinetic param-

eters were also modified by the DMSO content in the assay.

The K_Mvalues of substrates I and IV against BACE-1 were in-

creased 1.3-fold when the DMSO concentration was increased

(Table 1). The highest affinity (lowest K_Mvalue) and enzymatic

activity (highest k_cat/K_Mvalue) of substrate I proved to be the

most suitable for a sensitive, rapid, and specific substrate-

based BACE-1 assay. All of the FRET substrates I–IV were di-

gested in vitro with BACE-1 and the generated peptide frag-

ments were analyzed and identified by LC–MS (electrospray

ionization), with the conclusion that they have a single cleav-

age site between the leucine and aspartate residues (LﬂD; fig-

ure S11 in the Supporting Information).

fluorescence signals in the uncleaved substrates were

quenched by resonance energy transfer between the fluoro-

phore and the quencher groups. After cleavage of the peptide

chain, the quenching efficiency decreased, which resulted in

an increase of the fluorescence signal at 398 nm. To increase

the solubility of the FRET substrates in the buffer assay, the

peptides were modified by the introduction of additional argi-

nine residues into the sequence. The synthesis of FRET sub-

strates was conducted by using 2-chloro-2-trityl resin (sub-

strates I and II) or Rink amide (substrate III) as solid supports

(see schemes S1 and S2 in the Supporting Information and the

Experimental Section) for 9-fluorenylmethoxycarbonyl (Fmoc)-

based solid-phase peptide synthesis (SPPS). The products were

obtained in good yields (60%). Knowledge of the reaction ki-

netics and the K_Mvalue of the substrate is a mandatory re-

quirement for defining the assay conditions of the enzymatic

reaction. Next, the FRET substrates I–III were characterized

with respect to their kinetic properties by measuring the initial

velocities at various substrate concentrations (figures S6–S10 in

the Supporting Information). For this purpose, the BACE-1_1–501

(BACE-1, His*Tag, human, recombinant, NSO cells, Calbiochem,

PF125) full protein was selected. The effects of enzyme con-

centration and incubation time, as well as the influence of di-

methyl sulfoxide (DMSO), on the rate of substrate cleavage

Assay validation

The sensitivity of the biochemical assay was validated by re-

producing the IC₅₀value of the peptide inhibitor H-Glu-Val-

Asn-[(2R,4S,5S)-5-amino-4-hydroxy-2,7-dimethyloctanoyl]-Ala-

Glu-Phe-OH, an OM99-2 derivative. This statine inhibitor has

a peptide sequence similar to a known enzyme substrate, but

the scissile bond is replaced by the statine isostere to mimic

ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemMedChem 2013, 8, 1041 – 1056 1043

CHEMMEDCHEM

FULL PAPERS

www.chemmedchem.org

the tetrahedral reaction intermediate of aspartic proteases. The

compound has been reported to be a potent inhibitor of

human brain BACE-1, with inhibition constant (K_i) values of

1.2–9.8 nm.^[33,34]Thus, the BACE-1 full protein was treated with

different inhibitor concentrations in the range of 1–250 nm

and the product progress curves in the presence of this prod-

uct afforded a K_ivalue of 5.47 nm, which is in good agreement

with the published data.

Table 2. IC₅₀determination for some modified peptide–coumarin deriva-

tives against BACE-1.

Compound

IC₅₀[mm]^[a]

K_i[mm]^[b]

1

2

5

157Æ6.5

72Æ3

146Æ5^[c]

15.2Æ0.6

>250

7

8

96Æ3

20.3Æ0.6

>250

20.7Æ0.8

>250

V (CRI)

98Æ4

[a] For IC₅₀determination, full-length protein BACE-1 (His*Tag, human, re-

combinant NSO cells, PF 125, [E]=18.6 nm, 25 ng), (MOCAc)SEVNLDAEFK-

(DNPA)RR, FRET substrate I ([S]=25 mm), and 100 mm sodium acetate

(pH 4) were used. The IC₅₀values are the mean (n=3) ÆSD and were de-

termined by the GraphPad Prism program by using log[Inhibitor] versus

normalized responseÀvariable slope; l_abs=328 nm, l_emi=398 nm. The

assay was carried out with 10% DMSO (final concentration). [b] Unless

noted, the K_ivalues for these inhibitors were calculated by using the

equation of Cheng and Prusoff (K_i=IC₅₀/(1+[S]/K_M)), taking into account

a competitive binding mode.^[59][S]=25 mm and K_M=(6.71Æ1.51) mm for

substrate I (see Table 1). [c] This K_ivalue was calculated from the Linewea-

ver–Burk representation by using five different inhibitor concentrations

(25–250 mm) and increasing substrate concentrations (1–25 mm).

Peptide aldehyde V as a chemically reactive inhibitor of

BACE-1: design, synthesis, and biological evaluation

Peptide aldehydes have been described frequently as inhibi-

tors of serine and cysteine proteases and have been studied

intensively as biochemical tools and potential peptide

drugs.^[35,36]In general, peptide aldehydes inhibit cysteine pro-

teases through the reversible formation of covalent, tetrahe-

dral intermediates such as hemiacetals (serine proteases) or

hemithioacetals (cysteine proteases), both of which mimic the

transition states of the cleaved substrates.^[37]Peptide aldehydes

have also been suggested as inhibitors of aspartic proteases

because they are potential transition state analogues by them-

selves.^[29a]For these cases, it was assumed that the hydrated

form of the aldehyde accounts for its activity.^[38]To our knowl-

edge, for the aspartic protease BACE-1, peptide aldehydes

have not yet been described as inhibitors. Thus, for our pur-

poses, we have investigated a peptide aldehyde as a chemically

reactive inhibitor (CRI) of BACE-1 and as a directing probe for

dynamic ligation screening. Based on the reported cleavage

site of BACE-1 in the Swedish mutant substrate SEVNLDAEFR,

the N-acetylated pentapeptide aldehyde N-acetyl-Ser-Glu-Val-

Asn-Leu-H (V) was synthesized as a potential inhibitor. By

using the preloaded aldehyde resin H-Leu-HNovaSyn TGResin

as a solid support,^[39]product V was obtained in 38% overall

yield and with good purity by Fmoc solid-phase peptide syn-

thesis. The biological activity of this peptide aldehyde against

BACE-1 protease was determined and an IC₅₀value of (98Æ

4) mm was measured (Table 2). This result was in agreement

with our initial hypothesis that peptide aldehydes can be in-

hibitors of aspartic proteases such as b-secretase. The inhibito-

ry activity can be attributed to the formation of a tetrahedral,

hydrated form of the peptide aldehyde, which may act as

a transition state analogue recognized by the two active site

Asp residues (Figure 1, VII).

Drug Index (WDI) revealed 561 cyclic substructures that were

classified as privileged bioactive scaffolds. The small fragment

library was assembled in order to represent the substructure

composition of the WDI, together with an application of filters

for reactivity and for physicochemical properties, such as the

number of hydrogen bonding donors ꢀ3, the clogP value,

and molecular weights ꢀ300 Da. For the assay, 18.6 nm BACE-

1 protease was incubated with peptide aldehyde V (75 mm,

final concentration) and a 2.6-fold excess of one nucleophilic

fragment (200 mm, final concentration) per well in a 384-well

microtiter plate for one hour at room temperature. FRET sub-

strate I (25 mm) was then added and the rate of the enzymatic

reaction was recorded for each fragment–enzyme combina-

tion. Rate differences in the turnover of the substrate were ob-

served and quantified to identify the active-fragment hits.

Under these experimental conditions, the fragment 3-(3-amino-

phenyl)coumarin (APC; 1) was identified as an inhibitory frag-

ment due to its ability to decrease the enzymatic activity in

the presence of the peptide aldehyde. To our surprise, several

coumarin-containing structures have been discovered as novel

b-secretase inhibitors over the last few years; however, there

has been no indication of the binding site of these molecules

yet.^[41–45]Thus, fragment 1 was validated as a potential starting

point for further ligand development. For this purpose, the

binding mode and K_ivalue were first determined by conduct-

ing concentration-dependent inhibition assays. The initial ve-

locities of the enzymatic reaction at five different concentra-

tions of fragment 1 (25–250 mm) and increasing substrate con-

centrations (1–25 mm) were represented in a Lineweaver–Burk

plot (Figure 2a). Values of the reciprocal velocity (1/V) for each

inhibitor concentration were linear with respect to the recipro-

cal substrate concentration (1/[S]) and all of the lines intercept-

ed the Y axis in roughly the same point and the X axis at differ-

ent points (À1/K_M(1+I/K_i)), which indicated that this compound

displayed a competitive inhibition mode. In a second graph,

Fragment screening: hit identification and subsequent SAR

development by using the protein structure input

Next, a dynamic ligation assay was implemented by using sub-

strate I and peptide aldehyde V. The directing probe V was

employed at a concentration of 75 mm, which resulted in par-

tial inhibition of the protease (63% of initial activity) and left

room for the detection of additional inhibitory effects exerted

by interacting nucleophilic fragments. For screening, a collec-

tion of 450 amine nucleophiles was assembled by employing

a maximum common substructure (MCS) concept described re-

cently.^[40]In this concept, a structural analysis of the World

ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemMedChem 2013, 8, 1041 – 1056 1044

CHEMMEDCHEM

FULL PAPERS

www.chemmedchem.org

of 1 was confirmed by liquid chromatography/electrospray-

ionization time-of-flight mass spectrometry (LC-ESI-TOF-MS).

FRET substrate I was cleaved by BACE-1 to generate the two

cleavage products CP1 and CP2 (figure S11 in the Supporting

Information). The extracted ion chromatogram (EIC) peak areas

for each cleavage products were normalized to 100% enzyme

activity (see figure S13 and table S1 in the Supporting Informa-

tion). In the presence of increasing concentrations of inhibitor

1 (0–250 mm), the peak areas for each cleavage product were

decreased in correspondence with the enzyme activity. At

50 mm inhibitor 1, the enzyme activity was decreased by 20%

(for CP1) or by 17% (for CP2), whereas at 250 mm 1, the

enzyme activity was completely deleted. Analysis of peak areas

for different concentrations yielded a mean IC₅₀value of

135 mm for fragment 1.

In the next step, the concentration-dependent effects of

fragment 1 were validated in the dynamic ligation assay

(Figure 3). For this purpose, different concentrations of frag-

ment 1 (0–150 mm) were added to buffer solutions containing

the enzyme, substrate, and peptide aldehyde (PA; 0 or 75 mm)

and the rates of the enzymatic reactions were determined. En-

zymatic activity was decreased to 63% in the presence of PA

alone (75 mm) and to a similar degree by the inhibitory frag-

ment APC alone at the same concentration (60%, 75 mm).

However, if equal concentrations of PA and APC (75 mm each)

were investigated in the enzyme reaction, an additive effect

was recorded, with 48% enzyme activity, not the cooperative

(over-additive) effect reported earlier for other dynamic liga-

tions assays. In the presence of a double concentration of

1 with PA (75 mm), stronger inhibition was observed, with only

35% of the enzyme activity maintained.

Figure 2. a) Kinetics of b-site amyloid precursor protein cleaving enzyme 1

inhibition with the 3-(3-aminophenyl)-2H-chromen-2-one (1) inhibitor: Line-

weaver–Burk plot. The experiment was carried out with increasing substrate

!

~

^

concentrations ([S]; 1–25 mm) ( ) and 1 at concentrations of 25 ( ), 50 ( ),

&

*

100 ( ), and 250 mm ( ) with a final enzyme concentration of 37.2 nm

(50 ng; V_max=0.019ꢄ10^À2mms^À1and K_M=14.3 mm). The data obtained with

this FRET assay confirm the competitive binding mode. V: velocity. b) Deter-

mination of the K_ivalue (146 mm) for the BACE-1 inhibitor (1): Dixon-plot.

The experiment was carried out with increasing inhibitor concentrations ([I];

Thus, it remained to be investigated whether fragment

1 and peptide aldehyde V indeed formed a ligation product,

such as an imine or hemiaminal, as the active species responsi-

ble for decreasing the enzymatic activity in the assay. Isolation

~

!

&

*

25–250 mm) and substrate at concentrations of 2.5 ( ), 5 ( ), 10 ( ) 15 ( ),

^

and 25 mm ( ). Data are the mean Æ of two experiments.

the Dixon plot, the reciprocal initial velocity was represented

as a linear function of the inhibitor concentration (Figure 2b).

The intersection point of all of these lines with the X axis yield-

ed a K_ivalue for this compound of 146 mm.

In addition, analysis of the inhibition mechanism was also

conducted by using nonlinear least-squares fitting (NLSF) with

the Solver macro in the Excel program.^[46]Thus, experimental

data in x/y form and data calculated from a regression equa-

tion are inputted and plotted in a Microsoft Excel worksheet

and the addition of squares residual (SSR) is computed and

minimized by using the Solver add-in to obtain the parameter-

values set that best describes the experimental data. Thus,

with consideration of the five different inhibitor concentrations

(25–250 mm) and substrate concentrations (1–25 mm) and the

Michaelis–Menten constant for substrate I (kinetic parameters

Figure 3. The turnover of substrate I ([S]=25 mm) was recorded with BACE-

&

*

1 ([E]=18.6 nm) alone ( , positive control), in the presence of PA ( , 75 mm),

of

V a final

_max=0.019ꢄ10^À2mms^À1and K_M=14.3 mm for

~

!

or in the presence of increasing concentrations of inhibitor 1 ( , 25 mm;

,

^

75 mm; , 150 mm). The substrate turnover was decreased in the mixtures

enzyme concentration of 37.2 nm (50 ng)), the best fit was ob-

tained for a competitive binding mode, with SSR=7.967ꢄ

10^À6mm. As a third control experiment, the inhibitory capacity

~

&

*

containing PA (75 mm) and inhibitor 1 ( , 25 mm; , 75 mm; ], 150 mm).

These changes in activity suggest that the formation of an intermediate be-

tween both inhibitors (PA and 1) is responsible for the observed changes.

ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemMedChem 2013, 8, 1041 – 1056 1045

CHEMMEDCHEM

FULL PAPERS

www.chemmedchem.org

of diamino derivative 5 was obtained by treatment

with acetic anhydride in the absence of base at room

temperature.

Inhibition of the enzyme BACE-1 by these com-

pounds was determined by using the FRET assay de-

scribed above.^[30,32]The IC₅₀values obtained for each

compound are represented in Table 2. Reductive ami-

nation product 2 displayed a K_ivalue of 15 mm,

which indicated a significantly stronger inhibition

than both the initial directing probe V and the amine

hit fragment 1. For comparison, the N-acetylated

pentapeptide amine AcSEVNL-CH₂NH₂(8), the formal

reductive amination product of aldehyde V with am-

monia, was synthesized. The compound was ob-

tained in 71% yield from SPPS with tritylamine resin

as the solid support (Scheme 1). The inhibitory ca-

pacity of 8 against BACE-1 was tested and the com-

pound was inactive, with an IC₅₀value higher than

250 mm. Evidently, the amino group of the reduced

amide alone was not sufficient to establish binding

to the active site but required the contribution of

either the APC fragment or the aldehyde functionali-

ty for activity. These results confirmed our initial hy-

pothesis and indicated that the hit fragment is capa-

ble of interacting with an additional binding site of

the enzyme next to the binding site of the peptide

aldehyde.

Scheme 1. Reagents and conditions: a) N₂, molecular sieves, DMF, NaCNBH₃(2 equiv),

10% AcOH, 18 h, RT; b) 20% piperidine/DMF, 30 min, RT; c) N₂, HOBt (1.2 equiv), DIC

(1.2 equiv), CH₂Cl₂, 12 h, RT; d) N₂, CH₂Cl₂/TFA/EDT (80:17.5:2.5), 2 h, RT; e) N₂, Ac₂O

(1 equiv), CH₂Cl₂, 2 h, RT; f) DMF, NaCNBH₃(2.5 equiv), 24 h, RT; g) Fmoc SPPS, (AA)_n

(5 equiv), HOBt (5 equiv), DIC (5 equiv), DMF, 2 h, RT; h) 20% piperidine/DMF (1ꢄ1 min,

2ꢄ20 min); i) Ac₂O (5 equiv; 2ꢄ20 min), DMF, RT; j) TFA/TIS/H₂O/EDT (94:1:2.5:2.5), 2 h,

RT. DMF: N,N-dimethylformamide; HOBt: N-hydroxybenzotriazole; DIC: N,N’-diisopropyl-

carbodiimide; TFA: trifluoroacetic acid; EDT: ethanedithiol; AA: amino acid; TIS: triisopro-

pylsilane; trt: triphenylmethyl (trityl).

The potential binding mode of compound 2 with

the catalytic active site of the protein target was in-

vestigated by molecular modeling (Figure 4). For the

purpose, the X-ray co-crystal structure of BACE-1 pro-

tein and a peptidic statine inhibitor (PDB ID: 1FKN)

available in the Protein Data Bank was analyzed.^[33]

Binding simulations were carried out with the Surflex

docking program within SYBYL (see the Experimental

of these reversibly formed intermediates was not possible, so

secondary amine 2 was synthesized as a stable ligation prod-

uct instead (Scheme 1). The hit fragment, 1, was synthesized

by using a reported procedure.^[47]Fmoc-leucinal, 3, was pre-

pared in two steps from Fmoc-leucine.^[29a]First, treatment with

isobutylchloroformate (IBCF), N-methylmorpholine (NMM), and

sodium borohydride in situ furnished Fmoc-leucinol. Subse-

quently, oxidation of the intermediary alcohol with Dess–

Martin periodinane afforded aldehyde 3 in almost quantitative

yield (95%).^[20]By using this protocol, racemization of aldehyde

3 could be avoided, as reported.^[37]Treatment of 3 with 1 in

the presence of sodium cyanoborohydride afforded the secon-

dary amine 4 in moderate yield (65%). Deprotection of 4 with

20% piperidine in N,N’-dimethylformamide furnished 3-(3-((S)-

2-amino-4-methylpentylamino)phenyl)-2H-chromen-2-one (5),

which was finally coupled with the fully protected tetrapeptide

AcS(tBu)E(OtBu)VN(trt)-OH 6 in the presence of HOBt and DIC.

Deprotection of the side chains with TFA and EDT afforded the

final product 2 in moderate yield. The protected tetrapeptide

Ac-S(tBu)E(OtBu)VN(trt)-OH 6 was synthesized by solid-phase

peptide synthesis in good yield (75%) by using a 2-chlorotrityl

chloride resin as the solid support. The N-acetylation product 7

Section for details).^[48–51]All resulting docking conformations

were inspected visually and the most probable docking confor-

mations were chosen.

Figure 4. Proposed binding mode of the pentapeptide amine 2 (carbon

atoms in bold blue) docked at the active site of BACE-1 (cyan). The hydro-

gen bonds are depicted as dashed line (magenta). Amino acids are indicated

by three-letter codes.

ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemMedChem 2013, 8, 1041 – 1056 1046

CHEMMEDCHEM

FULL PAPERS

www.chemmedchem.org

In Figure 4, the potential binding mode of compound 2 in

the active site of BACE-1 is visualized. The pentapeptide-de-

rived secondary amine 2 fitted well into the active site of the

enzyme. Firstly, the acylated ethane-1,2-diamine isostere was

able to interact with aspartic acid residues Asp32 and Asp228,

present in the catalytic site. The nitrogen atom of the aniline is

protonated, so hydrogen bonds interaction with both amino

acids were supplemented by a Coulomb attraction. This result

constituted an extension of earlier findings by us and others,

which had described ethane-1,2-diamine isosteres incorporated

in peptides as potent inhibitors of BACE-1.^[52]Secondly, the leu-

cine residue at the P1 position of 2 was located in the large

and hydrophobic pocket S1, establishing CH–p interactions

with aromatic residues such as Phe108, Trp115, and Tyr71,

whereas the rest of the amino acid residues establish an hydro-

gen bond network with the S2–S5 pockets of the enzyme,

such as Arg235, Gly230, Thr232, and Gln73. Thirdly, the cou-

marin moiety was always exposed to the protein surface, ac-

quiring different orientations, and its carbonyl group interacted

with Tyr198, located in the S2’ pocket, through a hydrogen

bond.

Hit optimization

In the next phase, the fragment combination product 7 was

optimized by fragment variations by employing input derived

from the docking experiments with the target protein. The

active site of BACE-1 is composed by the catalytic dyad formed

by Asp32 and Asp228. Directly adjacent to this catalytic site,

the S1 pocket is located. This S1 pocket is a large hydrophobic

site lined by the side chains of Phe108, Trp115, Ile118, and

Leu30.^[12]In close proximity to the S1 pocket, the flap region

can be found, which is composed of (among others) the

amino acid residue Tyr71. The presence of these aromatic resi-

dues suggested the introduction of an aromatic side chain,

which might be able to establish p-stacking interactions, into

the fragment-based inhibitor 7.^[53]This strategy was supported

by the size of the S1 pocket, which was not completely filled

by the leucine residue (Figure 4) and should be able to accom-

modate an aromatic moiety. Additional interactions might be

contributed by fluorine substituents.^[54–56]Thus, by starting

from inhibitor 7, we planned the synthesis of derivatives 12–

28 (Scheme 2). In these new derivatives, the leucine side chain

in the P1 position was replaced by benzyl, 3,5-difluorobenzyl,

and biphenyl groups. In addition, substituents at the terminal

amino group were varied and some modifications in the cou-

marin scaffold were introduced.^[54,55]

Next, compounds 5 and 7, the entirely nonpeptidic combi-

nations of hit fragment 1 and the ethane-1,2-diamine isostere,

were tested for BACE-1 inhibition. Compound 5, the isobutyl-

substituted ethane-1,2-diamine derivative, which was obtained

by reductive amination of l-leucinal with fragment 1 and con-

tained a free primary amine, did not show any inhibitory activi-

ty. This observation was in

The syntheses of the designed compounds started from the

commercially available N-Fmoc- or N-Boc-protected l-a-amino

agreement with our earlier find-

ings indicating that the ethane-

1,2-diamine isosteres needed

a peptide extension for activi-

ty.^[52]To our great surprise, how-

ever, compound 7, formally the

reductive amination product of

N-acetyl-leucinal and fragment 1,

almost maintained the affinity of

the tetrapeptide-derived frag-

ment ligation product, with an

IC₅₀value of 96 mm (K_ivalue of

20 mm). The better inhibitory af-

finity observed for the acetyl de-

rivative 7 could be explained by

a postulated hydrogen bond in-

teraction of the acetyl carbonyl

group with the side chain of

Gln73, found in the catalytic site.

This interaction cannot be

formed by the nonacylated dia-

mine 5. Furthermore, compound

7 establishes one extra hydrogen

bond with Thr231. This finding

indicated that compound 7 was

a favorable starting point for fur-

ther optimization of a nonpepti-

Scheme 2. Reagents and conditions: a) N₂, DMF, AcOH, NaCNBH₃(2 equiv), 30 min, RT; b) 20%piperidine/DMF,

30 min, RT; c) 4m HCl/dioxane, 2 h, RT; d) Ac₂O (1 equiv), CH₂Cl₂, 3 h, RT; e) (Boc)₂O, CH₂Cl₂, 4 h, RT; f) CbzCl, CH₂Cl₂,

2 h, RT. Boc: tert-butoxycarbonyl; Cbz: benzyloxycarbonyl.

dic structure.

ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemMedChem 2013, 8, 1041 – 1056 1047

CHEMMEDCHEM

FULL PAPERS

www.chemmedchem.org

acids 29a–c, which contained benzyl, biphenyl, or 3,5-difluoro-

benzyl side chains, respectively. Activation of the protected

amino acids 29a–c with IBCF and NMM, followed by reduction

with NaBH₄in situ afforded the corresponding alcohols in

good yields. Subsequent oxidation of these intermediates with

Dess–Martin periodinane furnished the aldehydes 30a–c in

quantitative yield.^[57,52]Treatment of 30a–c with the 3-(3-ami-

nophenyl)coumarin derivatives 1, 9 (6-MeO), or 10 (7-MeO), or

with 3-(4-aminophenyl)coumarin (11) under reductive amina-

tion conditions provided the derivatives 12–17 in moderate

yields (45–70%). The aminophenylcoumarin derivatives 1 and

9–11 were prepared by following the experimental conditions

described in the literature.^[47,58]Treatment of the Fmoc deriva-

tives 12 and 15–17 with 20% piperidine in DMF yielded the

deprotected products 18 and 21–23 in good yields, whereas

deprotection of Boc derivatives 13 and 14 with a solution 4m

HCl in dioxane afforded diamines 19 and 20. N-Acetylation of

the amines 18–20 with acetic anhydride in the absence of

base at room temperature provided the acetylated derivatives

24–26 in good yields. The Boc derivative 27 was obtained by

treatment of 18 with Boc anhydride. Acylation of 18 with ben-

zyloxycarbonyl chloride in the absence of base afforded the

Cbz derivative 28 in moderate yield.

potent inhibitors of human BACE-1 (K_ivalues of 5–50 mm) than

the starting fragment alone or in combination with the com-

plete peptide aldehyde V (compound 5).

In agreement with our hypothesis, the introduction of aro-

matic rings at P1, the extended and lipophilic biphenyl group,

and the fluorine atoms as electron-acceptor groups into the

phenyl ring increased the activity of these new compounds

against BACE-1 and decreased their K_ivalues.^[53]In a compari-

son of the different fragments in the P1 position of com-

pounds 18–20 containing the nonacylated, primary amine, the

biphenyl derivative 20 was the most active compound, with

a K_ivalue of 11 mm, whereas the benzyl- and the 3,5-difluoro-

benzyl derivatives showed little difference in their K_ivalues

(18: 40.1 mm; 19: 33.7 mm). A similar tendency for the three

side chains was observed if the N terminus of the fragment

combinations was modified, for example, in the Boc derivatives

27 (benzyl, 19.2 mm), 13 (3,5-difluorobenzyl, 8.8 mm), and 14

(biphenyl, 5.3 mm). The most active N-terminal substituent was

the N-acetyl group (compounds 24–26). The latter three deriv-

atives were the most potent inhibitors with K_ivalues of 5.1

(24: benzyl), 16.5 (25: 3,5-difluorobenzyl), and 3.7 mm (26: bi-

phenyl). The potency of these three compounds against BACE-

1 was two- to sevenfold higher than that of the corresponding

products without N-acylation. Introduction of a Cbz group

(compound 28; IC₅₀>200 mm), however, was not tolerated by

the protein and decreased the inhibitory capacity of this com-

pound. Variations or additional substitutions on the aminophe-

nylcoumarin moiety also led to a distinct improvement of the

inhibitory activities. First, the new initial fragments 9–11 were

investigated as inhibitors of BACE-1 and showed activities in

the same mm range as the initial hit 1 (data not shown). Incor-

poration of the fragment derivatives in the ethane-1,2-diamine

structure provided fragment combinations with a significant

increase in affinity (decrease in K_ivalues). Relative to the activi-

ty of compound 18 (K_i=40.1 mm), the introduction of the 6-

methoxy group in 21 increased the activity twofold (K_i=

19.3 mm). This enhancement in the activity could be explained

by formation of one additional hydrogen bond with Arg128

present on the protein surface (Figure 5). The 7-methoxy deriv-

ative 22 displayed an almost fourfold increase in activity (K_i=

11.5 mm) relative to that of reference compound 18, which can

be rationalized by the formation of two additional hydrogen

bonds between the carbonyl group of the chromone ring and

the hydroxy group of Tyr198 and between the methoxy

oxygen atom and Arg128, respectively. The shift of the amino

substituent from the meta to the para position in compound

23 also led to an almost fourfold rise in activity (K_i=11 mm)

and the docking of the inhibitor (Figure 6) indicated a good fit

into the S1 pocket, including hydrogen bonding with the cata-

lytic dyad (Asp32 and Asp228), as well as with Gly230 and

Thr72. Additional studies are required to investigate whether

these modifications lead to further improved inhibitors if they

are combined with each other and with N-acylation of the ter-

minal amino group.

The activity of these synthesized compounds against BACE-

1 was evaluated in the cell-free inhibition assay in a concentra-

tion-dependent manner. Inhibition constants (K_ivalues) were

calculated from the IC₅₀values by using the equation of Cheng

and Prusoff (K_i=IC₅₀/1+[S]/K_M),^[59]which is valid for competi-

tive inhibitors. Results for the IC₅₀and K_ivalues are summar-

ized in Table 3. In general, fragment combinations of the hit 3-

(3-aminophenyl)-coumarin fragment

1

(K_i=146 mm) with

amino acid aldehydes were found to be significantly more

Table 3. IC₅₀determination for 3-(3-aminophenyl)coumarin derivatives

with aromatic rings in the P1 position (see Scheme 2 for structures).^[a,b]

Compd

R

X

R’

IC₅₀[mm]

K_i[mm]^[c]

13

14

18

19

20

21

22

23

24

25

26

27

28

Boc

H

Ac

Boc

Cbz

b

c

a

b

c

a

b

c

H

25.1Æ1

15.1Æ0.5

114.9Æ10

96.4Æ3.5

31.5Æ1.8

55.2Æ3.3

33Æ1

8.8

5.3

40.1

33.7

11.0

19.3

11.5

11.0

5.1

H

6-OCH₃

7-OCH₃

H

31.6Æ1.7

14.7Æ0.5

47.2Æ1.3

10.7Æ0.4

55.0Æ3

16.5

3.7

a

19.2

>200

[a] For the IC₅₀determination, full-length protein BACE-1 (His*Tag, human,

recombinant NSO cells, PF 125, [E]=18.6 nm, 25 ng), RhoEVNLDAEFR-

Quencher ([S]=250 nm), and 100 mm sodium acetate (pH 4) were used.

[b] The IC₅₀values are the mean (n=3) ÆSD; l_abs=535 nm, l_emi

=

590 nm. The assay was carried out with 5% DMSO (final concentration).

[c] The K_ivalues for these inhibitors were calculated by using the equa-

tion of Cheng and Prusoff (K_i=IC₅₀/(1+[S]/K_M)), taking into account com-

petitive binding mode.^[59][S]=250 nm and K_M=0.134 mm (see the Sup-

porting Information).

ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemMedChem 2013, 8, 1041 – 1056 1048

CHEMMEDCHEM

FULL PAPERS

www.chemmedchem.org

I

(V_max=0.142ꢄ10^À2mms^À1

,

K_M=4.97 mm, and k_cat/K_M=

15687 s^À1m^À1) was selected due to a fivefold increase in sensi-

tivity compared with similar, published substrates. The use of

the pentapeptide aldehyde as a directing probe in the dynam-

ic ligation assay in presence of the protein target (BACE-1) was

established. The assay was used to screen a library of nucleo-

philic fragments, which had been chosen by using a maximum

common substructure concept. The best fragment hit, 3-(3-

aminophenyl)coumarin (APC, 1; K_i=146 mm), was selected for

validation. Competitive binding of 1 to the active site of the

enzyme BACE-1 was confirmed by kinetic analysis and by the

synthesis and investigation of the irreversible fragment combi-

nation product with the directing probe.

For hit-to-lead optimization, combinations of fragment

1 with various nonpeptidic fragments targeting the S1 site of

the protein were prepared. Entirely nonpeptidic fragment com-

binations containing the N-acyl-ethane-1,2-diamine motif were

found to up to 30-fold more active than the initial fragment

hit. Considering earlier findings, which had indicated that the

ethane-1,2-diamine motif required an extended peptide se-

quence for activity, these observations were completely unex-

pected.^[56]Clearly, fragment 1 is capable of favorably replacing

the amino acids in the S’ direction. As secondary fragments tar-

geting the S1 site of BACE-1, 2-methylpropyl, benzyl, 3,5-di-

fluorobenzyl, and biphenyl were found to enhance binding,

with K_ivalues in the low micromolar range. Preference for the

S1 position depended on the N-acyl substituent. With N-Boc

substitution, the 3,5-difluorobenzyl (13, 8.8 mm) and the bi-

phenyl fragments (14, 5.3 mm) were preferred, whereas with N-

acetyl substitution, the unsubstituted benzyl derivative (24,

5.1 mm) was found to be superior to the substituted 3,5-difluor-

obenzyl (25, 16.5 mm) and isobutyl fragments (7, 20.3 mm). The

strongest activity for a compound with an acylated, primary

aliphatic amine was detected for the biphenyl derivative 26

(K_i=3.7 mm). Further improvement is indicated for a modifica-

tion of the hit fragment itself (compare, for example, com-

pounds 18 and 23) and by variation of the N-acyl substituent.

The chemical optimization process was supported by the cal-

culated binding modes of the fragment combinations to the

protein. According to molecular modeling and the recorded

structure–activity relationships, the compounds can interact fa-

vorably with Asp32 and Asp228, which define the active cata-

lytic site of the enzyme, through a protonated aromatic amine.

Binding is further sustained by additional hydrogen bonds and

hydrophobic interactions of the molecules. Exchange of an iso-

butyl residue in the P1 position by several aromatic residues

was found to be advantageous due to better space-filling of

the hydrophobic S1 pocket.

Figure 5. Cartoon representation of the BACE-1 active site. Key residues, be-

longing to the different subsites, are highlighted in black for clarity. a) Dock-

ing pose of compound 21. b) Docking pose of compound 22. Hydrogen

bonds are depicted as dashed lines (magenta). The binding mode was de-

rived from the Surflex program within the SYBYL software.

Figure 6. Most reasonable docked pose of compound 23 in the BACE-1 bind-

ing site. The amino acid residues, belonging to the different subsides, are la-

beled in black for clarity. The hydrogen bonds are depicted as dashed lines

(magenta).

Conclusions

In this work, we have investigated the application of dynamic

ligation screening, a method for fragment-based drug discov-

ery, to the identification of active-site binding fragments of the

aspartic protease BACE-1 and their development into nonpep-

tidic, small-molecule inhibitors of this protein. The implementa-

tion of DLS for BACE-1 involved the design and synthesis of

new FRET substrates I–III, with strongly enhanced sensitivity, as

well as the development of a peptide aldehyde inhibitor of

BACE-1 as a chemically reactive directing probe. FRET substrate

In summary, we have demonstrated in this work that dy-

namic ligation screening can be applied successfully to aspartic

proteases such as BACE-1 as targets and represents a valuable

approach for the site-directed discovery of inhibitory frag-

ments. Peptide aldehydes can be used as directing probes for

this enzyme class and hit fragments can be extended to frag-

ment-based compounds containing the N-acylated ethane-1,2-

diamine motif, which can act as potent inhibitors. Herein, the

identified 3-(aminophenyl)coumarin fragment has been estab-

ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemMedChem 2013, 8, 1041 – 1056 1049

CHEMMEDCHEM

FULL PAPERS

www.chemmedchem.org

Fmoc-Arg(Pbf)-OH and Fmoc-Lys(Mtt)-OH (Mtt: 4-methyltrityl) by

using HOBt and DIC as coupling reagents (fivefold molar excess) in

DMF for 2 h at room temperature. The completion of all coupling

was confirmed with a negative ninhydrin test result. The resin was

then split into two parts (100 mg) and the N-Mtt group on the

lysine side chain was removed by using a cocktail of TFA/TIS/

CH₂Cl₂(2:5:93) for 2 h at room temperature. After the washing, the

quencher groups 2-(2,4-dinitrophenyl)acetic acid (substrate I) or

2,4-dinitrobenzoic acid (substrate II) were coupled with a fivefold

molar excess of HOBt and DIC in a mixture of DMF/THF (2:1) for

10 h at room temperature. After that, the peptide was elongated

by SPPS coupling of Fmoc-Phe-OH, Fmoc-Glu(OtBu)-OH, Fmoc-Ala-

OH, Fmoc-Asp(OtBu)-OH, Fmoc-Leu-OH, Fmoc-Asn(trt)-OH, Fmoc-

Val-OH, Fmoc-Glu(OtBu)-OH, and Fmoc-Ser(tBu)-OH by using the

same coupling reagents (fivefold molar excess) in DMF for 2 h at

room temperature. After SPPS, the substrate was then N-terminally

labeled with the fluorophore MOCAc. Thus, the resin was treated

with 7-methoxycoumarin-4-acetic acid (fivefold molar excess) in

the presence of HOBt (5 equiv) and DIC (5 equiv) in DMF for 10 h

at room temperature. Finally, the peptide was cleaved from the

resin and the protective amino acid side chains were removed

with a cocktail TFA/TIS/H₂O/EDT (94:1:2.5:2.5) for 3 h at room tem-

perature. The filtrate was collected and the solvents were removed

under reduced pressure. The residue was purified by reverse HPLC

(solvent A and solvent B, from 5% B to 99% B over 45 min, at

a flow of 20 mLmin^À1with a linear gradient). Thus, substrate I

(60 mg, 52%) was isolated as a yellow powder after lyophilization:

HPLC purity: 98% (UV, t_R=2.62 min); HRMS (electrospray ioniza-

tion, positive mode): m/z calcd for C₈₂H₁₁₇N₂₄O₂₉: 1901.8418 [M+

H]⁺; found: 1902.8376. Substrate II was isolated as a yellow

powder after lyophilization (53 mg, 47%): HPLC purity: 99% (UV,

t_R=2.54 min); HRMS (electrospray ionization, positive mode): m/z

calcd for C₈₁H₁₁₄N₂₃O₂₉: 1872.8153 [M+H]⁺; found: 1873.8212.

lished as a successful starting point for the development of en-

tirely nonpeptidic, drug-like inhibitors of BACE-1 with low-mi-

cromolar activity and further potential for optimization.

Experimental Section

Synthesis

General: Chemicals and dry solvents were purchased from Sigma–

Aldrich, Fluka, or Novabiochem and used without further purifica-

tion or distillation unless otherwise stated. Fmoc solid-phase pep-

tide syntheses were performed in polypropylene 2–20 mL syringes

fitted with polyethylene porous filter disks. Solvents and soluble re-

agents were removed by suction. The Fmoc group was removed

with piperidine/DMF (2:8; 1ꢄ1 min, 2ꢄ10 min). Washings between

coupling and deprotection steps were performed with DMF (5ꢄ

0.5 min), tetrahydrofuran (THF; 5ꢄ0.5 min), and CH₂Cl₂(5ꢄ0.5 min)

by using at least 10 mL of solvent per gram of resin each time. For

elemental analyses or Fmoc determination, the resins were washed

as described above with additional washings with MeOH (5ꢄ

0.5 min) and Et₂O (5ꢄ0.5 min) and with brief drying under suction

of air and finally under high vacuum. Loadings were quantitatively

determined by using a Jasco V-550 UV/Vis spectrophotometer with

Hellma Suprasil cuvettes with a path length of 1 mm. Solid-phase

reactions were monitored by FTR-IR spectrometry of the resin by

using a Bruker Tensor 27 FTR-IR instrument equipped with a Pike

MIRacle ATR apparatus fitted with a ZnSe crystal plate. Frequencies

(n˜) are expressed in cm^À1. Solution-phase reactions were monitored

either by LC–MS techniques or by thin layer chromatography (TLC)

with Analtech silica gel plates (60 F₂₅₄) and the spots were exam-

ined under UV light at 254 nm or stained with developing re-

agents. LC–MS data were recorded on an Agilent 1100 series chro-

matography workstation (Agilent Technologies) equipped with

a single quadrupole mass spectrometer and electrospray ionization

(ESI). Solvents A (0.1% TFA in Millipore water), B (0.1% TFA in ace-

tonitrile), and C (0.1% TFA in MeOH) were used in a linear gradient

(5–99% B or C in 5 min or 30 min for preparative runs). Peptides

were purified either on a semipreparative HPLC column (VP 250/10

Nucleodur 100-5 C₁₈ec Macherey–Nagel) by employing individual

gradients derived from analytical runs (eluents A–C). HRMS meas-

urements were conducted on an Agilent 6220 ESI-TOF mass spec-

trometer. Nuclear magnetic resonance (NMR) spectra (¹H and

¹³C NMR) were recorded on a Bruker AVANCE 300 MHz instrument

and chemical shifts (d) were measured in parts per million (ppm)

relative to tetramethylsilane (TMS) used as an internal reference.

Coupling constants (J) are expressed in Hertz (Hz). The following

abbreviations are used to describe peak patterns when appropri-

ate: s: singlet; d: doublet; t: triplet; qt: quintet; m: multiplet; br:

broad. The commercial FRET substrates IV ((MOCAc)SEVNLDAEFRK-

(DNP)RR) and RhoEVNLDAEFR-Quencher were purchased from R&D

Systems, Inc. (ES004) and Invitrogen (P2985), respectively. Triton X-

100 was purchased from Aldrich Chemical Co. Inc.

(DNPA)SEVNLDAEFRRG(AMCA) (III): A Rink amide resin (500 mg,

f=0.70 mmolg^À1) placed in a 25 mL polypropylene syringe fitted

with a polyethylene filter disk was treated with 20% piperidine in

DMF (5 mL; 1ꢄ1 min, 2ꢄ20 min). After that, the resin was filtered

off, washed several times (five times) with DMF, MeOH, THF, and

CH₂Cl₂, and dried under high vacuum. Subsequently, it was treated

with

2-{7-[(9H-fluoren-9-ylmethoxy)carbonyl]amino}-4-methyl-2-

oxo-2H-chromen-3-yl)acetic acid (5 equiv) in presence of HOBt

(5 equiv) and DIC (5 equiv) in DMF for 15 h at room temperature,

until the ninhydrin test result was negative. When the reaction was

finished, the loading was calculated by UV spectroscopy (f=

0.62 mmolg^À1). The chain was elongated by sequentially coupling

Fmoc-Gly-OH, Fmoc-Arg(Pbf)-OH, Fmoc-Arg(Pbf)-OH, Fmoc-Phe-

OH, Fmoc-Glu(OtBu)-OH, Fmoc-Ala-OH, Fmoc-Asp(OtBu)-OH, Fmoc-

Leu-OH, Fmoc-Asn(trt)-OH, Fmoc-Val-OH, Fmoc-Glu(OtBu)-OH, and

Fmoc-Ser(tBu)-OH with HOBt and DIC as a coupling reagents (five-

fold molar excess) in DMF for 2 h at room temperature. The com-

pletion of all couplings was confirmed with negative ninhydrin test

results. After the last amino acid coupling, the Fmoc protective

group was removed with 20% piperidine in DMF. The amino termi-

nus was modified by the introduction of the fluorescence quench-

er by treatment with 2-(2,4-dinitrophenylamino)acetic acid

(5 equiv), HOBt (5 equiv), and DIC (5 equiv), in THF/DMF (3:2) for

2 h at room temperature. The resin was filtered off, the peptide

was cleaved from the resin, and the protective amino acid side

chains were removed with a cocktail TFA/TIS/H₂O/EDT (94:1:2.5:2.5)

for 3 h at room temperature. The filtrate was collected and the sol-

vents were removed under reduced pressure. The residue was pu-

rified by reverse HPLC (solvent A and solvent B, from 5% B to 99%

B over 45 min, at a flow of 20 mLmin^À1with a linear gradient).

(MOCAc)SEVNLDAEFK(DNPA)RR (I) and (MOCAc)SEVNLDAEFK-

(DNB)RR (II): A Rink amide resin (300 mg, f=0.7 mmolg^À1) placed

in a 25 mL polypropylene syringe fitted with a polyethylene filter

disk was treated with Fmoc-Arg(Pbf)-OH (Pbf: 2,2,4,6,7-pentame-

thyldihydrobenzofuran-5-sulfonyl; 5 equiv) in the presence of HOBt

(5 equiv) and DIC (5 equiv) for 4 h at room temperature, until the

ninhydrin test for primary amines was negative. When the reaction

was finished, the resin was filtered off, washed several times (five

times) with DMF, MeOH, THF, and CH₂Cl₂, and dried under high

vacuum. The loading was calculated by UV spectroscopy (f=

0.60 mmolg^À1). The chain was elongated by sequentially coupling

ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemMedChem 2013, 8, 1041 – 1056 1050

CHEMMEDCHEM

FULL PAPERS

www.chemmedchem.org

Thus, substrate III (336 mg, 53%) was isolated as a yellow powder

after lyophilization: HPLC purity: 98% (UV, t_R=3.01 min); HRMS

(electrospray ionization, positive mode): m/z calcd for C₇₉H₁₁₀N₂₄O₂₇:

915.3955 [M+2H]²⁺/2; found: 915.3972.

tion was finished, the loading was calculated by UV spectroscopy

(f=0.8 mmolg^À1; IR: n˜ =1727.91, 1663.08, 1118.07, 737.31 cm^À1).

The chain was elongated by sequentially coupling Fmoc-Asn(trt)-

OH, Fmoc-Val-OH, Fmoc-Glu(OtBu), and Fmoc-Ser(tBu)-OH by using

HOBt and DIC as coupling reagents (fourfold molar excess) in DMF

for 2 h at room temperature. The completion of all couplings was

confirmed with a negative ninhydrin test result. After the last

amino acid coupling, the Fmoc protective group was removed

with 20% piperidine in DMF. The amino terminus was acetylated

with acetic anhydride (400 mL, 4.25 mmol, 5 equiv; 2 times, until

the ninhydrin test was negative). Finally, the peptide was cleaved

from the resin and the protective groups at the amino acid side

AcSEVNL–CHO (V): An H-Leu-H-NovaSyn TG resin (500 mg, f=

0.23 mmolg^À1) placed in a 10 mL polypropylene syringe fitted with

a polyethylene filter disk was treated with 20% piperidine in DMF

(5 mL; 1ꢄ1 min, 2ꢄ20 min). After that, the resin was filtered off,

washed with several times (five times) with DMF, MeOH, THF, and

CH₂Cl₂, and dried under high vacuum. The chain was elongated by

sequentially coupling Fmoc-Asn(trt)-OH, Fmoc-Val-OH, Fmoc-Glu-

(OtBu)-OH, and Fmoc-Ser(tBu)-OH (fourfold molar excess) by using

HOBt and DIC as coupling reagents (fourfold molar excess) in DMF

for 2 h at room temperature. The completion of all coupling was

confirmed with a negative ninhydrin test result. The Fmoc protec-

tive group for the last amino acid was removed by 20% piperidine

in DMF. The N terminus was acetylated with acetic anhydride

(4 equiv) in DMF (2 times, until the ninhydrin test was negative). Fi-

nally, the peptide was cleaved from the resin and the protective

groups at the amino acid side chains were removed with a cocktail

of TFA/CH₂Cl₂(80:20) for 3 h at room temperature. The filtrate was

then treated with cool Et₂O and the precipitate was filtered off and

washed three times with Et₂O to yield compound V (24.3 mg,

36%) as a white powder after lyophilization: HPLC purity: 98% (UV,

t_R=1.61 min); HRMS (electrospray ionization, positive mode): m/z

calcd for C₂₅H₄₃N₆O₁₁: 587.3035 [M+H]⁺; found: 587.3053.

chains were removed with

a cocktail of TFA/TIS/H₂O/EDT

(94:1:2.5:2.5) for 3 h at room temperature, followed by successive

washings with CH₂Cl₂(five times). The filtrates were collected, the

solvents were evaporated, and the resulting residue was then

treated with cool Et₂O. The precipitate was filtered off and washed

three times with cold Et₂O to yield compound 8 (235 mg, 71%) as

a white powder after lyophilization: HPLC purity: 99% (UV, t_R=

1.49 min); HRMS (electrospray ionization, positive mode): m/z calcd

for C₂₅H₄₂N₆O₁₀: 588.3352 [M+H]⁺; found: 588.3367.

General procedure for the synthesis of derivatives 4 and 12–17:

In a round-bottom flask, the molecular sieves were activated under

high vacuum and heating. After that, under an N₂atmosphere, a so-

lution of 3-(3-aminophenyl)-2H-chromen-2-one derivatives 1, 9, or

10 or 3-(4-aminophenyl)-2H-chromen-2-one (11; 1 equiv) and the

appropriate aldehyde (1.4 equiv) in

a mixture of DMF/AcOH

AcS(tBu)E(OtBu)VN(trt)-OH (6): 2-Chlorotrityl chloride resin (1.5 g,

f=1.2–1.5 mmolg^À1) placed in a 10 mL polypropylene syringe

fitted with a polyethylene filter disk was treated with Fmoc-Asn-

(trt)-OH (1.3 g, 1.8 mmol, 1.0 equiv) and with diisopropylethylamine

(DIPEA; 1.2 mL, 7.2 mmol, 4.0 equiv) in DMF for 15 min at room

temperature. Subsequently, an additional portion of DIPEA (2.5 mL,

14.4 mmol, 8.0 equiv) was added and the reaction mixture was

stirred for 3 h at room temperature. After that, the resin was treat-

ed with MeOH (1.18 mL) and the reaction mixture was stirred for

another 15 min. The resin was filtered off, washed five times with

DMF, MeOH, THF, and CH₂Cl₂, and dried under high vacuum. When

the reaction was finished, the loading was calculated by UV spec-

troscopy (f=0.9 mmolg^À1). The chain was elongated by sequential-

ly coupling Fmoc-Val-OH, Fmoc-Glu(OtBu)-OH, and Fmoc-Ser(tBu)-

OH by using HOBt and DIC as coupling reagents (fourfold molar

excess) in DMF for 2 h at room temperature. The completion of all

couplings was confirmed with a negative ninhydrin test result.

After the last amino acid coupling, the Fmoc protective group was

removed with 20% piperidine in DMF and the amino terminus was

acetylated with Ac₂O (679 mL, 7.2 mmol, 4 equiv; 2 times, until the

ninhydrin test was negative). Finally, the peptide was cleaved from

the resin, and the protected amino acid side chains were main-

tained with a cocktail of TFE/AcOH (1:1; 30%) and CH₂Cl₂(70%) for

1 h at room temperature and successive washings with CH₂Cl₂

(5 times). The filtrates were collected and the solvents were evapo-

rated. The resulting residue was treated with cool Et₂O and the

precipitate was filtered off and washed three times with cold Et₂O

to afford compound 6 (685 mg, 90%) as a white powder after lyo-

philization: HPLC purity: 99% (UV, t_R=4.59 min); HRMS (electro-

spray ionization, positive mode): m/z calcd for C₄₆H₆₂N₅O₁₀:

844.4491 [M+H]⁺; found: 844.4478.

(1 mL:100 mL) was added. The reaction mixture was stirred for 1 h

at room temperature and then NaCNBH₃(2 equiv) was added in

one portion. The resulting mixture was stirred for an additional

12 h at room temperature. The reaction mixture was then diluted

with CH₂Cl₂(75 mL) and washed sequentially with brine (3ꢄ50 mL)

and H₂O (3ꢄ50 mL). The organic phases were dried over Na₂SO₄

and filtered. The solvents were removed under reduced pressure.

The mixture was purified by flash column chromatography with

the solvent mixture indicated for each case.

(9H-Fluoren-9-yl)-methyl-(S)-1-(3-(2H-chromen-2-one-3-yl)pheny-

lamino)-4-methylpentan-2-ylcarbamate (4): Reagents were 3-(3-

aminophenyl)-2H-chromen-2-one (1; 28.5 mg, 0.12 mmol, 1 equiv),

Fmoc-leucinal 3 (60.7 mg, 0.18 mmol, 1.4 equiv), and NaCNBH₃

(15.8 mg, 0.25 mmol, 2 equiv). Flash chromatography was conduct-

ed with hexane/EtOAc (3:1). From the fractions with R_f=0.5

(hexane/EtOAc, 2:1), compound 4 (40 mg, 57%) was isolated as

1

a yellow oil: H NMR (300 MHz, CDCl₃): d=0.93 (d, J=5.5 Hz, 3H),

0.96 (d, J=5.5 Hz, 3H), 1.42 (m, 1H), 1.65 (m, 2H), 3.28–3.17 (m,

2H), 3.98 (brs, 1H), 4.20 (m, 1H), 4.53–4.43 (m, 2H), 4.73 (d, J=

8.3 Hz, 1H), 6.68 (d, J=7.8 Hz, 1H), 6.99 (s, 1H), 7.20–7.42 (m, 9H),

7.49 (d, J=7.3 Hz, 2H), 7.57 (d, J=7.8 Hz, 2H), 7.75 ppm (t, J=

7.3 Hz, 3H); ¹³C NMR (75 MHz, CDCl₃): d=21.0, 22.0, 23.2, 24.6,

42.3, 47.2, 49.5, 60.4, 112.9, 113.3, 117.7, 119.7, 119.9 (2C), 120.0,

124.5, 124.9, 127.0 (2C), 127.6 (2C), 127.8, 128.5 (2C), 129.3, 131.2,

135.6, 139.7, 141.2 (2C), 141.3 (2C), 148.1, 153.3, 156.7, 160.6 ppm;

HRMS (electrospray ionization, positive mode): m/z calcd for

C₃₆H₃₅N₂O₄: 559.2591 [M+H]⁺; found: 559.2597; HPLC purity:

100% (UV, t_R=3.50 min).

(9H-Fluoren-9-yl)methyl-(S)-1-(3-(2H-chromen-2-one-3 yl)pheny-

lamino)-3-phenylpropyl-2-ylcarbamate (12): Reagents were 3-(3-

aminophenyl)-2H-chromen-2-one (1; 23.7 mg, 0.10 mmol, 1 equiv),

(9H-fluoren-9-yl)methyl-(S)-1-formyl-2-phenylethylcarbamate (30a;

43.8 mg, 0.13 mmol, 1.2 equiv), and NaCNBH₃(13.2 mg, 0.21 mmol,

2 equiv). Flash chromatography was conducted with hexane/EtOAc

(3:1). From the fractions with R_f=0.6 (hexane/EtOAc, 3:1), com-

AcSEVNL-CH₂NH₂(8): Fmoc-leucinal (860.4 mg, 2.55 equiv),

NaCNBH₃(160.3 mg, 2.55 equiv), and DMF were added to an

amino trityl resin (500 mg, f=1.7 mmolg^À1) placed in a 10 mL poly-

propylene syringe fitted with a polyethylene filter disk. The result-

ing slurry was stirred for 24 h at room temperature. When the reac-

ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemMedChem 2013, 8, 1041 – 1056 1051

CHEMMEDCHEM

FULL PAPERS

www.chemmedchem.org

pound 12 was isolated as a yellow oil (41.8 mg, 67%): ¹H NMR

(300 MHz, CDCl₃): d=2.93 (m, 2H), 3.30 (m, 2H), 4.48–4.36 (m, 3H),

4.80 (m, 1H), 6.62 (brd, J=7.5 Hz, 1H), 6.96 (brs, 1H), 7.06 (d, J=

7.5 Hz, 1H), 7.20–7.43 (m, 12H), 7.51 (m, 5H), 7.75 ppm (t, J=

7.5 Hz, 4H); ¹³C NMR (75 MHz, CDCl₃): d=38.8, 47.1, 52.1, 60.3,

66.6, 112.9, 113.2, 116.2, 117.7, 119.5, 119.8, 124.3, 124.9, 126.6,

126.9, 127.6, 127.7 (2C), 128.4, 128.6, 128.7, 128.8, 128.9, 129.1,

129.2, 131.1, 135.5, 136.3, 137.2, 131.1, 135.5, 136.3, 137.2, 139.7,

141.2 (2C), 143.6 (2C), 147.8, 153.3, 156.2, 160.5, 171.1 ppm; HRMS

(electrospray ionization, positive mode): m/z calcd for C₃₉H₃₃N₂O₄:

593.2435 [M+H]⁺; found: 593.2470; HPLC purity: 100% (UV, t_R=

3.76 min).

55.8, 66.7, 77.2, 109.6, 112.9, 113.2, 115.1, 117.2, 117.3, 117.7, 118.8,

119.7 (2C), 119.9, 124.8, 126.6, 126.9, 127.5 (2C), 128.5 (2C), 128.7

(2C), 129.1 (2C), 129.2, 135.7, 137.3, 139.6, 141.3, 143.8, 147.9,

153.7, 156.0, 156.3, 160.7, 162.5 ppm; HRMS (electrospray ioniza-

tion, positive mode): m/z calcd for C₄₀H₃₅N₂O₅: 623.2540 [M+H]⁺;

found: 623.2558; HPLC purity: 100% (UV, t_R=2.84 min).

(9H-Fluoren-9-yl)methyl-(S)-1-(3-(7-methoxy-2H-chromen-2-one-

3-yl)phenylamino)-3-phenylpropyl-2-ylcarbamate (16): Reagents

were

3-(3-aminophenyl)-7-methoxy-2H-chromen-2-one

(10,

48.1 mg, 0.18 mmol, 1 equiv), (9H-fluoren-9-yl)methyl-(S)-1-formyl-

2-phenylethylcarbamate (30a, 83.2 mg, 0.22 mmol, 1.2 equiv), and

NaCNBH₃(14.4 mg, 0.37 mmol, 2 equiv). Flash chromatography was

conducted with hexane/EtOAc (3:1). From the fractions with R_f=

0.5 (hexane/EtOAc, 3:1), compound 16 (75 mg, 64%) was isolated

(S)-1-(3-(2H-Chromen-2-one-3-yl)phenylamino)-tert-butyl-3-(3,5-

difluorophenylpropyl)-2-ylcarbamate (13): Reagents were 3-(3-

aminophenyl)-2H-chromen-2-one (1; 19.0 mg, 0.08 mmol, 1 equiv),

tert-butyl-(S)-2-(3,5-difluorophenyl)-1-formylethylcarbamate (30b;

39.1 mg, 0.10 mmol, 1.2 equiv), and NaCNBH₃(10.5 mg, 0.17 mmol,

2 equiv). Flash chromatography was conducted with hexane/EtOAc

(4:1 to 2:1). From the fractions with R_f=0.3 (hexane/EtOAc, 4:1),

compound 13 (30 mg, 70%) was isolated as a pale oil: ¹H NMR

(300 MHz, CDCl₃): d=1.41 (s, 9H), 2.89 (m, 2H), 3.20 (m, 1H), 3.30

(dd, J=12.7, 4.3 Hz, 1H), 4.07 (m, 1H), 4.54 (brs, 1H), 5.30 (brs,

1H), 6.65 (dt, J=8.4, 2.2 Hz, 2H), 6.76 (d, J=6.5 Hz, 2H), 6.98 (m,

1H), 7.02 (d, J=7.6 Hz, 1H), 7.22–7.32 (m, 2H), 7.37 (d, J=8.4 Hz,

1

as a yellow oil: H NMR (300 MHz, CDCl₃): d=2.92 (brs, 1H), 3.26

(m, 2H), 3.89 (s, 3H), 4.16 (m, 2H), 4.40 (m, 2H), 4.85 (m, 2H), 6.59

(brs, 1H), 6.84–7.01 (m, 4H), 7.13–7.55 (m, 14H), 8.02 (s, 1H),

7.73 ppm (m, 2H); ¹³C NMR (75 MHz, CDCl₃): d=31.4, 36.4, 47.2,

55.8, 66.6, 77.2, 100.3, 112.6, 112.9, 113.4, 115.3, 117.7, 119.9 (3C),

125.0, 126.7, 127.0, 127.6 (2C), 128.6 (2C), 128.7 (2C), 129.2 (2C),

129.3, 135.9, 137.3, 139.8, 141.3, 143.8, 153.7, 155.2, 156.3, 160.8,

162.5 ppm; HRMS (electrospray ionization, positive mode): m/z

calcd for C₄₀H₃₅N₂O₅: 623.2540 [M+H]⁺; found: 623.2563; HPLC

purity: 100% (UV, t_R=2.84 min).

1H), 7.53 (t, J=8.0 Hz, 2H), 7.82 ppm (s, 1H); ¹³C NMR (75 MHz,

(9H-Fluoren-9-yl)-methyl-(S)-1-(4-(2H-chromen-2-one-3-yl)phen-

ylamino)-3-phenylpropyl-2-ylcarbamate (17): Reagents were 3-(4-

aminophenyl)-2H-chromen-2-one (11, 35.6 mg, 0.15 mmol, 1 equiv),

(9H-fluoren-9-yl)methyl-(S)-1-formyl-2-phenylethylcarbamate (30a;

66.8 mg, 0.18 mmol, 1.2 equiv), and NaCNBH₃(11.2 mg, 0.29 mmol,

2 equiv). Flash chromatography was conducted with hexane/EtOAc

(3:1). From the fractions with R_f=0.4 (hexane/EtOAc, 3:1), com-

pound 17 (58.0 mg, 66%) was isolated as a yellow oil: ¹H NMR

(300 MHz, CDCl₃): d=2.97 (m, 2H), 3.26 (brs, 1H), 3.96 (m, 1H),

4.17 (m, 2H), 4.42 (m, 2H), 4.78 (m, 1H), 6.62 (d, J=6.7 Hz, 1H),

6.74 (d, J=7.8 Hz, 2H), 7.20 (brs, 1H), 7.23–7.41 (m, 8H), 7.45–7.58

(m, 7H), 7.65 (s, 1H), 7.72–7.76 ppm (m, 3H); ¹³C NMR (75 MHz,

CDCl₃): d=31.5, 35.8, 47.3, 52.0, 66.6, 77.3, 102.5, 112.5, 116.3,

120.0, 124.2, 124.9, 126.8, 127.0, 127.2, 127.5, 127.7, 127.8 (2C),

128.1, 128.7 (2C), 128.8, 128.9 (2C), 129.1, 129.2, 129.7 (2C), 130.5,

134.5, 136.9, 137.1, 141.3, 143.8, 153.1, 153.6, 154.0, 161.0 ppm;

HRMS (electrospray ionization, positive mode): m/z calcd for

C₃₉H₃₃N₂O₄: 593.2435 [M+H]⁺; found: 593.2421; HPLC purity:

100% (UV, t_R=5.17 min).

2

CDCl₃): d=28.3 (3C), 43.4, 47.7, 53.5, 77.2, 102.2, 112.0 (d, J_C,F

=

2

23.0 Hz), 113.0 (d, J_C,F=23 Hz), 116.4 (2C), 119.7, 124.4 (2C), 127.1,

127.8 (2C), 128.6, 129.4, 131.2, 135.7, 137.6, 138.3, 139.8, 141.7,

153.5 (2C), 155.4 ppm; HRMS (electrospray ionization, positive

mode): m/z calcd for C₂₉H₂₉F₂N₂O₄: 507.2090 [M+H]⁺; found:

507.1925; HPLC purity: 100% (UV, t_R=2.70 min).

(S)-1-(3-(2H-Chromen-2-one-3-yl)phenylamino)-tert-butyl-3-(3,5-

biphenylpropyl)-2-ylcarbamate (14): Reagents were 3-(3-amino-

phenyl)-2H-chromen-2-one (1; 23.7 mg, 0.10 mmol, 1 equiv), 2-bi-

phenyl-tert-butyl-(S)-1-formylethylcarbamate

(30c;

39.2 mg,

0.12 mmol, 1.2 equiv), and NaCNBH₃(13.2 mg, 0.20 mmol, 2 equiv).

Flash chromatography was conducted with hexane/EtOAc (4:1 to

2:1). From the fractions with R_f=0.3 (hexane/EtOAc, 3:1), com-

pound 14 (34.5 mg, 60%) was isolated as a white powder after lyo-

philization: ¹H NMR (300 MHz, CDCl₃): d=1.41 (brs, 9H), 2.92 (d,

J=6.0 Hz, 2H), 3.36 (m, 1H), 3.45 (dd, J=12.7, 4.4 Hz, 1H), 4.19 (m,

1H), 4.87 (brs, 2H), 6.97 (s, 1H), 7.24–7.36 (m, 7H), 7.41 (t, J=

7.5 Hz, 3H), 7.50–7.56 (m, 6H), 7.82 ppm (s, 1H); ¹³C NMR (75 MHz,

CDCl₃): d=28.1 (2C), 45.4, 48.6, 53.5, 77.0, 116.3, 116.5, 119.0, 119.4,

123.2, 123.4, 124.4, 124.7, 126.8 (2C), 127.1, 127.3, 127.9, 128.1,

128.2, 128.6 (2C), 129.5, 131.4, 132.2, 134.5, 136.0, 136.1, 139.7,

140.5, 141.1, 153.4, 153.6, 160.3 ppm; HRMS (electrospray ioniza-

tion, positive mode): m/z calcd for C₃₅H₃₅N₂O₄: 547.2591 [M+H]⁺;

found: 547.2601; HPLC purity: 100% (UV, t_R=5.20 min).

General procedure for Fmoc deprotection: synthesis of deriva-

tives 5, 18, and 21–23: Under an N₂atmosphere, the appropriate

Fmoc derivative (1 equiv) was treated with a 20% piperidine in

DMF solution for 30 min at room temperature. After that, the reac-

tion mixture was diluted with CH₂Cl₂(50 mL) and washed sequen-

tially with brine (2ꢄ30 mL) and H₂O (2ꢄ20 mL). The organic

phases were dried over Na₂SO₄and filtered. The solvent was re-

moved under reduced pressure. The residue was purified by flash

column chromatography with the solvents indicated in each case.

(9H-Fluoren-9-yl)methyl-(S)-1-(3-(6-methoxy-2H-chromen-2-one-

3-yl)phenylamino)-3-phenylpropyl-2-ylcarbamate (15): Reagents

were 3-(3-aminophenyl)-6-methoxy-2H-chromen-2-one (9; 29.4 mg,

0.11 mmol, 1 equiv), (9H-fluoren-9-yl)methyl (S)-1-formyl-2-phenyle-

thylcarbamate (30a; 48.3 mg, 0.13 mmol, 1.2 equiv), and NaCNBH₃

(8.50 mg, 0.22 mmol, 2 equiv). Flash chromatography was conduct-

ed with hexane/EtOAc (3:1). From the fractions with R_f=0.5

(hexane/EtOAc, 3:1), compound 15 (45 mg, 64%) was isolated as

3-(3-((S)-2-Amino-4-methylpentylamino)phenyl)-2H-chromen-2-

one (5): Reagents were ((9H-fluoren-9-yl)methyl-(S)-1-(3-(2H-chro-

men-2-one-3-yl)phenylamino)-4-methylpentan-2-yl-carbamate (4;

40 mg, 0.07 mmol, 1 equiv) and 20% piperidine in DMF (750 mL).

Flash chromatography was conducted with hexane/EtOAc (3:1) to

CH₂Cl₂/CH₃OH/30% NH_3(aq)(8:1:0.1). From the fractions with R_f=0.6

(CH₂Cl₂/CH₃OH/30% NH_3(aq), 8:1:0.2), amine derivative 5 was isolat-

ed as a yellow oil (16 mg, 66%): ¹H NMR (300 MHz, CD₃OD): d=

0.94 (d, J=6.5 Hz, 3H), 0.97 (d, J=6.5 Hz, 3H), 1.37 (m, 2H), 1.64

1

a yellow oil: H NMR (300 MHz, CDCl₃): d=2.98 (brs, 1H), 3.28 (m,

2H), 3.83 (s, 3H), 4.16 (m, 2H), 4.39 (m, 2H), 4.84 (m, 1H), 6.62 (m,

1H), 6.93 (dd, J=8.5, 3.0 Hz, 2H), 7.19–7.30 (m, 10H), 7.38 (t, J=

7.4 Hz, 3H), 7.75 (d, J=7.4 Hz, 2H), 7.74 (d, J=8.5 Hz, 4H),

8.02 ppm (s, 1H); ¹³C NMR (75 MHz, CDCl₃): d=31.4, 36.47, 47.2,

ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemMedChem 2013, 8, 1041 – 1056 1052

CHEMMEDCHEM

FULL PAPERS

www.chemmedchem.org

(m, 2H), 1.80 (hept, J=6.5 Hz, 1H), 3.00 (m, 1H), 3.13 (m, 1H), 3.22

(dd, J=12.8, 4.9 Hz, 1H), 6.70 (dd, J=8.0, 1.9 Hz, 1H), 6.94 (dd, J=

8.0, 1.9 Hz, 1H), 7.01 (t, J=1.9 Hz, 1H), 7.18 (t, J=8.0 Hz, 1H), 7.30–

7.36 (m, 2H), 7.57 (dt, J=7.5, 1.5 Hz, 1H), 7.67 (d, J=7.5 Hz, 1H),

8.00 ppm (s, 1H); ¹³C NMR (75 MHz, CD₃OD): d=20.6, 21.6, 23.8,

42.8, 47.9, 48.8, 112.0, 112.4, 115.0, 116.4, 119.2, 123.8, 127.4, 127.7,

128.1, 130.6, 134.9, 139.5, 148.1, 152.7, 160.4 ppm; HRMS (electro-

spray ionization, positive mode): m/z calcd for C₂₁H₂₅N₂O₂: 337.1911

[M+H]⁺; found: 337.1943; HPLC purity: 100% (UV, t_R=2.58 min).

3-(4-((S)-2-Amino-3-phenylpropylamino)phenyl)-2H-chromen-2-

one (23): Reagents were ((9H-fluoren-9-yl)methyl-(S)-1-(4-(2H-chro-

men-2-one-3-yl)phenylamino)-3-phenylpropyl-2-yl-carbamate (17;

30.0 mg, 0.05 mmol) and 20% piperidine in DMF solution (500 mL).

Flash chromatography was conducted with CH₂Cl₂/CH₃OH/30%

NH_3(aq)(from 20:1:0 to 8:1:0.1). From the fractions with R_f=0.2

(CH₂Cl₂/CH₃OH/30% NH_3(aq), 10:1:0.2), compound 23 (13.8 mg,

74%) was isolated as a yellow solid after lyophilization: ¹H NMR

(300 MHz, CDCl₃): d=1.88 (brs, 3H), 2.64 (dd, J=13.2, 7.8 Hz, 1H),

2.99 (m, 1H), 3.04 (dd, J=13.2, 8.6 Hz, 1H), 3.31 (d, J=9.0 Hz, 2H),

6.66 (d, J=8.4 Hz, 2H), 7.21–7.35 (m, 7H), 7.44–7.51 (m, 2H), 7.59

(d, J=7.2 Hz, 2H), 7.70 ppm (s, 1H); ¹³C NMR (75 MHz, CDCl₃): d=

41.8, 48.7, 52.2, 112.6 (2C), 116.3, 120.2, 123.4, 124.3, 126.7, 127.5,

128.2, 128.7 (2C), 129.3 (2C), 129.7 (2C), 130.5, 136.9, 138.0, 148.7,

153.1, 161.0 ppm; HRMS (electrospray ionization, positive mode):

m/z calcd for C₂₄H₂₃N₂O₂: 371.1754 [M+H]⁺; found: 371.1801;

HPLC purity: 100% (UV, t_R=1.95 min).

3-(3-((S)-2-Amino-3-phenylpropylamino)phenyl)-2H-chromen-2-

one (18): Reagents were ((9H-fluoren-9-yl)methyl-(S)-1-(3-(2H-chro-

men-2-one-3-yl)phenylamino)-3-phenylpropyl-2-yl-carbamate (12;

32.0 mg, 0.05 mmol) and 20% piperidine in DMF solution (500 mL).

Flash chromatography was conducted with CH₂Cl₂/CH₃OH/30%

NH_3(aq)(from 20:1:0 to 8:1:0.1). From the fractions with R_f=0.3

(CH₂Cl₂/CH₃OH/aqueous 30% NH₃, 10:1:0.2), 18 (19 mg, 95%) was

isolated as a yellow solid after lyophilization: IR (DMSO): n˜ =2360.0,

2341.9, 1717.7, 1601.8, 1507.1, 1489.8, 1456.8, 1331.8, 1204.0,

3-(3-((S)-2-Amino-AcSEVN-4-methylpentylamino)phenyl)-2H-

chromen-2-one (2): Under an N₂atmosphere, a mixture of 3-(3-

((S)-2-amino-4-methylpentylamino)phenyl)-2H-chromen-2-one (5;

6.73 mg, 0.02 mmol), AcS(tBu)E(OtBu)VN(trt)-OH (6; 20.9 mg,

0.02 mmol), HOBt (4.82 mg, 0.04 mmol), and DIC (5.5 mL, 4.50 mg,

0.04 mmol) in CH₂Cl₂(20 mL) was treated with DIPEA (8.9 mL,

6.15 mg, 0.047 mmol) at room temperature overnight. After that,

the reaction mixture was diluted with CH₂Cl₂(50 mL) and washed

with 10% aq NaHCO₃solution (3ꢄ30 mL), brine (3ꢄ30 mL), and

H₂O (3ꢄ30 mL). The organic phase was dried over Na₂SO₄and fil-

tered, then the solvent was removed under reduced pressure. The

resulting residue was purified by flash chromatography on a silica

gel column with the mixture CH₂Cl₂/MeOH (16:1) as an eluent.

From the fractions with R_f=0.6 (CH₂Cl₂/MeOH, 16:1), the fully pro-

tected peptide intermediate (22.7 mg, 82%) was isolated as

a white solid after lyophilization: HRMS (electrospray ionization,

positive mode): m/z calcd for C₆₇H₈₄N₇O₁₁: 1162.6223 [M+H]⁺;

found: 1162.6218; HPLC purity: 99% (UV, t_R=3.60 min). After that,

the intermediate (18 mg, 0.015 mmol) was treated with a mixture

of CH₂Cl₂/TFA/EDT (25:6:1; 1 mL) for 2 h at room temperature. The

mixture was then treated with cool Et₂O and the precipitate was

filtered off and washed three times with Et₂O. Compound 2 was

isolated as a white powder after lyophilization (7 mg, 56%): HRMS

(electrospray ionization, positive mode): m/z calcd for C₄₀H₅₄N₇O₁₁:

808.3876 [M+H]⁺; found: 808.3867; HPLC purity: 99% (UV, t_R=

3.02 min).

1110.1, 986.5, 923.2, 787.8, 752.9, 695.5 cm^À^{1 1}H NMR (300 MHz,

;

CDCl₃): d=2.62 (dd, J=8.3, 13.7 Hz, 1H), 2.91 (1H, m), 3.03 (dd, J=

8.5, 13.7 Hz, 1H), 3.31 (2H, m), 3.47 (brs, 3H), 6.65 (d, J=8.1 Hz,

1H), 6.98 (d, J=6.4 Hz, 2H), 7.20–7.37 (m, 8H), 7.52 (2H, m),

7.78 ppm (s, 1H); ¹³C NMR (75 MHz, CDCl₃): d=42.5, 49.7, 52.0,

113.1, 113.4, 116.3, 117.6, 119.7, 124.4, 126.5, 127.8, 128.5 (2C),

128.7, 129.2 (2C), 129.3, 131.2, 135.6, 138.5, 139.6, 148.3, 153.42,

160.6 ppm; HRMS (electrospray ionization, positive mode): m/z

calcd for C₂₄H₂₃N₂O₂: 371.1754 [M+H]⁺; found: 371.1763; HPLC

purity: 100% (UV, t_R=2.58 min).

3-(3-((S)-2-Amino-3-phenylpropylamino)phenyl)-6-methoxy-2H-

chromen-2-one (21): Reagents were ((9H-fluoren-9-yl)methyl-(S)-1-

(3-(6-methoxy-2H-chromen-2-one-3-yl)phenylamino)-3-phenylprop-

yl-2-yl-carbamate (15; 14.9 mg, 0.02 mmol) and 20% piperidine in

DMF solution (400 mL). Flash chromatography was conducted with

CH₂Cl₂/CH₃OH/30% NH_3(aq)(from 8:1:0 to 8:1:0.1). From the frac-

tions with R_f=0.2 (CH₂Cl₂/CH₃OH/30% NH_3(aq), 10:1:0.1), compound

21 (8 mg, 83%) was isolated as a yellow oil: ¹H NMR (300 MHz,

CDCl₃): d=2.50 (brs, 3H), 2.77 (dd, J=13.8, 7.8 Hz, 1H), 3.13 (dd,

J=13.8, 7.8 Hz, 1H), 3.29–3.41 (m, 2H), 3.49 (m, 1H), 3.86 (s, 3H),

6.60 (d, J=7.7 Hz, 1H), 6.95 (m, 2H), 7.03–7.11 (m, 2H), 7.16–7.30

(m, 6H), 7.73 (d, J=8.0 Hz, 1H), 8.02 ppm (s, 1H); ¹³C NMR

(75 MHz, CDCl₃): d=42.4, 49.6, 52.2, 55.8, 100.4, 112.7, 113.0, 113.1,

113.4, 117.5, 125.2, 126.5, 128.6 (2C), 128.7, 129.2 (2C), 135.9, 138.4,

139.8, 148.3, 153.7, 155.2, 160.8, 162.5 ppm; HRMS (electrospray

ionization, positive mode): m/z calcd for C₂₅H₂₅N₂O₂: 401.1860 [M+

H]⁺; found: 401.1870; HPLC purity: 100% (UV, t_R=1.96 min).

General procedure for Boc deprotection: synthesis of chromen-

2-ones 19 and 20: Under an N₂atmosphere, the appropriate tert-

butylcarbamate derivatives 10b or 10c (1 equiv) were treated with

4m HCl/dioxane solution for 2 h at room temperature. After that,

the solvents were removed under reduced pressure, the resulting

residue was precipitated in cool Et₂O, and the product was isolated

by filtration as the hydrochloride form.

3-(3-((S)-2-Amino-3-phenylpropylamino)phenyl)-7-methoxy-2H-

chromen-2-one (22): Reagents were ((9H-fluoren-9-yl)methyl-(S)-1-

(3-(7-methoxy-2H-chromen-2-one-3-yl)phenylamino)-3-phenylprop-

yl-2-yl-carbamate 16 (37.4 mg, 0.06 mmol) and 20% piperidine in

DMF solution (500 mL). Flash chromatography was conducted with

CH₂Cl₂/CH₃OH/30% NH_3(aq)(from 8:1:0 to 8:1:0.1). From the frac-

tions with R_f=0.2 (CH₂Cl₂/CH₃OH/30% NH_3(aq), 10:1:0.1), compound

22 (15 mg, 65%) was isolated as a yellow oil: ¹H NMR (300 MHz,

CDCl₃): d=2.46 (brs, 3H), 2.66 (dd, J=14.0, 7.7 Hz, 1H), 2.99 (m,

1H), 3.06 (dd, J=14.0, 7.8 Hz, 1H), 3.32 (m, 2H), 3.89 (s, 3H), 6.62

(d, J=8.1 Hz, 1H), 6.85–6.88 (m, 2H), 6.96 (m, 2H), 7.18–7.33 (m,

6H), 7.42 (d, J=9.2 Hz, 1H), 7.74 ppm (s, 1H); ¹³C NMR (75 MHz,

CDCl₃): d=42.4, 49.6, 52.2, 55.8, 100.4, 112.7, 113.0, 113.1, 113.4,

117.5, 125.2, 126.5, 128.6 (2C), 128.7, 129.2 (2C), 135.9, 138.4,

139.8, 148.3, 153.7, 155.2, 160.8, 162.5 ppm; HRMS (electrospray

ionization, positive mode): m/z calcd for C₂₅H₂₅N₂O₂: 401.1860 [M+

H]⁺; found: 401.1862; HPLC purity: 100% (UV, t_R=1.96 min).

3-(3-((S)-2-Amino-3-(3,5-difluorophenyl)propylamino)phenyl)-2H-

chromen-2-one (19): Reagents were 3-(3-((S)-2-amino-tert-butylcar-

bamato-3-(3,5-difluorophenyl)propylamino)phenyl)-2H-chromen-2-

one (13; 10.1 mg, 0.02 mg, 1 equiv) and 4m HCl/dioxane (0.8 mL).

Compound 19 (6.80 mg, 85%) was obtained as a yellow powder:

¹H NMR (300 MHz, CDCl₃): d=3.04 (d, J=7.6 Hz, 2H), 3.29 (m, 2H),

3.68 (m, 1H), 3.85 (brs, 1H), 6.87–6.89 (m, 3H), 7.21–6.99 (m, 4H),

7.37 (dt, J=8.3, 1.3 Hz, 1H), 7.43 (d, J=8.3 Hz, 1H), 7.62 (dt, J=8.3,

1.4 Hz, 1H), 7.79 (dd, J=7.6, 1.3 Hz, 1H), 8.18 (s, 1H), 8.43 ppm

(brs, 2H); ¹³C NMR (75 MHz, CDCl₃): d=43.6, 50.9, 60.0, 102.3,

112.3, 112.5, 112.8, 113.1, 115.8, 117.1, 119.4 (2C), 124.5, 127.4 (2C),

128.5, 128.8, 131.6, 135.3, 140.8, 147.6, 152.9, 159.5 ppm; HRMS

ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemMedChem 2013, 8, 1041 – 1056 1053

CHEMMEDCHEM

FULL PAPERS

www.chemmedchem.org

(electrospray ionization, positive mode): m/z calcd for C₂₄H₂₁F₂N₂O₂:

407.1566 [M+H]⁺; found: 407.1577; HPLC purity: 100% (UV, t_R=

1.96 min).

3-(3-((S)-2-Acetylamino-3-(3,5-difluorophenyl)propylamino)phen-

yl)-2H-chromen-2-one (25): Reagents were 3-(3-((S)-2-amino-3-(3,5-

difluorophenyl)propylamino)phenyl)-2H-chromen-2-one

(19;

10.9 mg, 0.03 mmol) and Ac₂O (2.5 mL, 0.027 mmol). Flash chroma-

tography was conducted with CH₂Cl₂/hexane (10:1). From the frac-

tions with R_f=0.4 (CH₂Cl₂/hexane, 10:1), compound 25 (9.5 mg,

77%) was isolated as a yellow oil: ¹H NMR (300 MHz, CDCl₃): d=

1.95 (s, 3H), 2.90 (d, J=7.5 Hz, 2H), 3.29 (m, 1H), 3.65 (m, 1H), 4.43

(m, 1H), 5.67 (d, J=7.6 Hz, 1H), 6.75–6.78 (m, 3H), 6.97–7.02 (m,

2H), 7.23 (d, J=8.8 Hz, 1H), 7.30–7.41 (m, 4H), 7.55 (d, J=7.6 Hz,

1H), 7.84 ppm (s, 1H); ¹³C NMR (75 MHz, CDCl₃): d=20.5, 43.3, 46.7,

50.3, 112.5 (2C), 114.3, 115.2, 116.5, 118.9, 124.0, 126.9, 128.0, 128.3,

128.7, 131.0, 134.9, 139.6, 140.3, 143.1, 147.2, 152.3, 159.1, 160.2 (d,

¹J_C,F=206 Hz, 2C), 171.4 ppm; HRMS (electrospray ionization, posi-

tive mode): m/z calcd for C₂₆H₂₃F₂N₂O₃: 449.4687 [M+H]⁺; found:

449.1673; HPLC purity: 100% (UV, t_R=2.43 min).

3-(3-((S)-2-Amino-3-biphenylpropylamino)phenyl)-2H-chromen-

2-one (20): Reagents were (S)-1-(3-(2H-chromen-2-one-3-yl)phenyl-

amino)-tert-butyl-3-(3,5-biphenylpropyl)-2-yl-carbamate

(14;

8.20 mg, 0.01 mmol, 1 equiv) and 4m HCl/dioxane (0.8 mL). Com-

pound 20 was obtained as a white powder after lyophilization

(5.87 mg, 90%): ¹H NMR (300 MHz, [D₆]DMSO): d=2.94 (dd, J=

13.3, 8.1 Hz, 1H), 3.11 (dd, J=13.3, 5.7 Hz, 1H), 3.28 (m, 2H), 6.63

(d, J=7.9 Hz, 1H), 3.60 (m, 2H), 6.90 (d, J=7.9 Hz, 1H), 7.18–7.12

(m, 1H), 7.47–7.30 (m, 8H), 7.66–7.55 (m, 5H), 7.75 (dd, J=7.9,

1.3 Hz, 1H), 8.15 (s, 1H), 8.39 ppm (brs, 2H); ¹³C NMR (75 MHz,

[D₆]DMSO): d=38.0, 44.3, 53.6, 117.6, 121.5, 126.4, 128.3 (2C),

128.4, 128.9, 129.0, 129.1, 130.1, 130.3 (2C), 130.4, 131.0, 131.3,

131.4, 133.3, 136.2, 136.6, 137.9, 142.1, 142.2, 144.4, 143.5, 155.2,

155.6, 162.8 ppm; HRMS (electrospray ionization, positive mode):

m/z calcd for C₃₀H₂₇N₂O₂: 447.2067 [M+H]⁺; found: 447.2082;

HPLC purity: 100% (UV, t_R=2.06 min).

3-(3-((S)-2-Acetylamino-3-biphenylpropylamino)phenyl)-2H-chro-

men-2-one (26): Reagents were 3-(3-((S)-2-amino-3-biphenylpropy-

lamino)phenyl)-2H-chromen-2-one (20; 9.82 mg, 0.02 mmol) and

Ac₂O (2.1 mL, 0.022 mmol). Flash chromatography was conducted

with CH₂Cl₂/hexane (10:1). From the fractions with R_f=0.6 (CH₂Cl₂/

hexane, 10:1), compound 26 (8.5 mg, 79%) was isolated as

a yellow oil: ¹H NMR (300 MHz, [D₆]DMSO): d=1.84 (s, 1H), 2.88

(dd, J=13.7, 8.1 Hz, 1H), 3.01 (dd, J=13.7, 5.4 Hz, 1H), 3.51 (m,

1H), 3.77 (dd, J=14.6, 4.2 Hz, 1H), 4.04 (dd, J=14.6, 7.6 Hz, 1H),

7.19 (d, J=7.9 Hz, 2H), 7.31–7.42 (m, 5H), 7.46 (t, J=7.20 Hz, 2H),

7.55 (t, J=7.20 Hz, 5H), 7.65 (t, J=7.9 Hz, 2H), 7.75 (t, J=7.8 Hz,

2H), 8.18 (brs, 1H), 8.31 ppm (s, 1H); ¹³C NMR (75 MHz, [D₆]DMSO):

d=22.9, 44.4, 49.8, 50.9, 115.9, 119.3, 124.7, 125.8, 126.4 (2C),

126.5, 126.7 (2C), 127.3, 128.0, 128.6, 128.8 (2C), 128.9, 129.5, 129.7

(2C), 129.9, 132.0, 135.1, 138.6, 139.6, 141.2, 142.6, 153.0, 159.6,

171.0 ppm; HRMS (electrospray ionization, positive mode): m/z

calcd for C₃₂H₂₉N₂O₃: 489.2173 [M+H]⁺; found: 489.2209; HPLC

purity: 100% (UV, t_R=2.03 min).

General procedure for the synthesis of acetyl derivatives 7 and

24–26: Under an N₂atmosphere, a solution of the amino deriva-

tives 5 or 18–20 (1 equiv) in CH₂Cl₂(2 mL) was treated with Ac₂O

(1 equiv) in the absence of base for 4 h at room temperature. After

that, the reaction mixture was diluted with CH₂Cl₂(30 mL) and

washed sequentially with brine (3ꢄ20 mL) and H₂O (3ꢄ20 mL).

The organic phase was dried over Na₂SO₄and filtered, then the sol-

vent was removed under reduced pressure. The residue was puri-

fied by silica gel flash column chromatography with solvent mix-

tures indicated in each case as eluents.

3-(3-((S)-2-Acetylamino-4-methylpentylamino)phenyl)-2H-chro-

men-2-one (7): Reagents were 3-(3-((S)-2-amino-4-methylpentyla-

mino)phenyl)-2H-chromen-2-one (5; 5.72 mg, 0.02 mmol) and Ac₂O

(1.6 mL, 0.02 mmol). Flash chromatography was conducted with

CH₂Cl₂/MeOH (8:0.2). From the fractions with R_f=0.4 (CH₂Cl₂/

MeOH, 8:0.2), compound 7 (10 mg, 87%) was isolated as a yellow

oil: ¹H NMR (300 MHz, CDCl₃): d=0.87 (t, J=6.0 Hz, 6H), 1.37 (m,

2H), 1.94 (s, 3H), 1.63 (m, 1H), 3.00 (m, 1H), 3.13 (dd, J=13.8,

3.8 Hz, 1H), 4.27 (m, 1H), 5,90 (d, J=8.9 Hz, 1H), 7.29–7.40 (m, 4H),

7.50–7.60 (m, 3H), 7.63 (d, J=7.8 Hz, 1H), 7.81 (d, J=7.8 Hz, 1H),

8.05 ppm (s, 1H); ¹³C NMR (75 MHz, CDCl₃): d=22.3, 22.7, 23.6,

24.9, 42.5, 47.3, 52.5, 116.5, 119.6, 124.6, 128.3 (2C), 128.5, 130.1,

131.8, 134.9, 136.5, 140.9, 142.9, 153.6, 154.2, 160.4, 170.3 ppm;

HRMS (electrospray ionization, positive mode): m/z calcd for

C₂₃H₂₇N₂O₃: 379.2016 [M+H]⁺; found: 379.2011; HPLC purity:

100% (UV, t_R=min).

3-(3-((S)-Amino-tert-butylcarbamate-3-phenylpropylamino)phen-

yl)-2H-chromen-2-one (27): Under an N₂atmosphere, 3-(3-((S)-2-

amino-3-phenylpropylamino)phenyl)-2H-chromen-2-one

(18;

10.0 mg, 0.03 mmol) was treated with di-tert-butyldicarbonate

(6.33 mg, 0.03 mmol, 1 equiv) in CH₂Cl₂(2 mL) for 4 h at room tem-

perature. The reaction was diluted with CH₂Cl₂(40 mL) and washed

with saturated aq. NaCl (3ꢄ30 mL) and H₂O (3ꢄ30 mL). The organ-

ic phase was dried over Na₂SO₄and filtered, then the solvent was

removed under reduced pressure. The residue was purified by

flash chromatography on a silica gel column by using CH₂Cl₂/

MeOH (4:0.2) as the eluent mixture. From the fractions with R_f=

0.7 (CH₂Cl₂/MeOH, 4:0.2), compound 27 (11 mg, 86%) was isolated

1

as a yellow oil: H NMR (300 MHz, CDCl₃): d=1.41 (brs, 9H), 2.88

3-(3-((S)-2-Acetylamino-3-phenylpropylamino)phenyl)-2H-chro-

men-2-one (24): Reagents were 3-(3-((S)-2-amino-3-phenylpropyla-

mino)phenyl)-2H-chromen-2-one (18; 10.0 mg, 0.03 mmol) and

Ac₂O (2.5 mL, 0.03 mmol). Flash chromatography was conducted

with CH₂Cl₂/MeOH (8:0.2). From the fractions with R_f=0.4 (CH₂Cl₂/

MeOH, 8:0.2), compound 24 (10 mg, 87%) was isolated as a yellow

oil: ¹H NMR (300 MHz, CDCl₃): d=1.90 (s, 3H), 2.91 (d, J=6.8 Hz,

2H), 3.20 (dd, J=12.6, 7.4 Hz, 1H), 3.30 (m, 1H), 4.40 (m, 1H), 5.62

(d, J=5.6 Hz, 1H), 6.62 (dd, J=8.2, 2.2 Hz, 1H), 6.97 (d, J=7.7 Hz,

2H), 7.19–7.24 (m, 4H), 7.28–7.38 (m, 5H), 7.54 (d, J=7.7 Hz, 2H),

7.81 ppm (s, 1H); ¹³C NMR (75 MHz, CDCl₃): d=23.5, 38.8, 47.6,

50.4, 112.8, 113.3, 116.4, 117.6, 119.7, 124.4, 126.7, 127.8, 128.5,

128.7 (2C), 129.1 (2C), 129.4, 131.4, 135.6, 137.4, 139.8, 148.1,

153.4, 160.7, 170.8 ppm; HRMS (electrospray ionization, positive

mode): m/z calcd for C₂₆H₂₅N₂O₃: 413.1865 [M+H]⁺; found:

413.1879; HPLC purity: 100% (UV, t_R=2.39 min).

(brd, J=5.9 Hz, 2H), 3.13 (m, 1H), 3.29 (dd, J=12.7, 5.5 Hz, 1H),

4.10 (m, 1H), 4.58 (br, 1H), 6.61 (dd, J=7.9, 1.9 Hz, 1H), 6.93 (brs,

1H), 6.99 (brd, J=7.9 Hz, 1H), 7.02 (d, J=7.5 Hz, 1H), 7.20–7.37 (m,

8H), 7.54 (d, J=7.9 Hz, 1H), 7.81 ppm (s, 1H); ¹³C NMR (75 MHz,

CDCl₃): d=28.3 (3C), 40.6, 46.8, 47.6, 77.2, 112.9, 113.2, 116.4, 117.7,

119.7, 124.3, 126.6, 127.9, 128.6, 128.7 (2C), 129.1 (2C), 129.3,

131.3, 135.6, 137.5, 139.7, 146.8, 153.4, 155.9, 160.6 ppm; HRMS

(electrospray ionization, positive mode): m/z calcd for C₂₉H₃₁N₂O₄:

471.2278 [M+H]⁺; found: 471.2333; HPLC purity: 100% (UV, t_R=

2.71 min).

3-(3-((S)-Amino-benzoylcarbamate-3-phenylpropylamino)phen-

yl)-2H-chromen-2-one (28): Under an N₂atmosphere, a solution of

3-(3-((S)-2-amino-3-phenylpropylamino)phenyl)-2H-chromen-2-one

(18; 10.2 mg, 0.03 mmol) in CH₂Cl₂(2 mL) was treated with benzyl

chloroformate (3.8 mL, 0.03 mmol, 1 equiv) for 4 h at room temper-

ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemMedChem 2013, 8, 1041 – 1056 1054

CHEMMEDCHEM

FULL PAPERS

www.chemmedchem.org

ature. After that, the reaction mixture was diluted with CH₂Cl₂

(50 mL) and washed with H₂O (3ꢄ25 mL). The organic phase was

dried over Na₂SO₄and filtered off, then the solvent was removed

under reduced pressure. The residue was purified by flash chroma-

tography on silica gel by employing CH₂Cl₂/MeOH (4:0.2). From the

fractions with R_f=0.4 (CH₂Cl₂/MeOH, 4:0.2) compound 28 (9 mg,

66%) was isolated as a yellow oil: ¹H NMR (300 MHz, CDCl₃): d=

2.81 (d, J=6.3 Hz, 2H), 3.31 (m, 1H), 3.57 (m, 1H), 4.00 (m, 1H),

4.16 (brs, 1H), 5.03 (s, 2H), 7.09 (d, J=6.8 Hz, 2H), 7.18–7.42 (m,

13H), 7.49–7.52 (m, 3H), 7.69 (d, J=6.8 Hz, 1H), 7.78 ppm (s, 1H);

¹³C NMR (75 MHz, CDCl₃): d=51.3, 51.9, 52.9, 66.3, 116.3, 119.4,

124.0, 126.4, 127.5 (2C), 127.7, 127.9, 128.2 (3C), 128.3, 128.3 (3C),

128.6, 128.7, 128.8, 129.0, 129.1 (2C), 131.4, 135.5, 136.7, 140.3,

153.4, 155.9, 160.2 ppm; HRMS (electrospray ionization, positive

mode): m/z calcd for C₃₂H₂₉N₂O₄: 505.2121 [M+H]⁺; found:

505.2167; HPLC purity: 100% (UV, t_R=2.68 min).

cence intensities in the presence and absence of inhibitor were

compared, and the percent inhibition due to the presence of

tested compounds was calculated. The background signal, mea-

sured in control wells containing all of the reagents except BACE-1,

was subtracted for each reaction mixture. The relative percent in-

hibition due to the presence of eight increasing concentrations of

test compounds was calculated from the equation: 100À(IF_i/

IF₀ꢄ100), in which IF_iand IF₀are the fluorescence intensities ob-

tained for BACE-1 in the presence and absence of inhibitor, respec-

tively. The inhibition curves were obtained by plotting the relative

inhibition or activity (%) versus the logarithm of concentration of

the inhibitor. The regression parameters were determined and the

IC₅₀value was extrapolated (GraphPad Prism 4.0, GraphPad Soft-

ware Inc.; dose–response inhibition, log[I] versus normalized re-

sponse).

Docking studies

Biological evaluation

All docking studies were carried out by using the Surflex docking

program within the SYBYL-X1.3 program. As the receptor, the X-ray

structure of b-secretase complexed with a statine peptide inhibitor,

available in the Protein Data Bank (PDB ID: 1FKN), was used. The li-

gands were prepared by using the ligand preparation module of

the SYBYL-X1.3 program. The ligand 3D input files (2, 5, 7, and 13–

28) were generated with the SYBYL-X1.3 program and optimized

by energy minimization with the Conjugate Gradients method and

the Tripos force field until the convergence criterion of 0.05 kcal(

mol^À1ꢅ^À1) was reached; 20 docked poses were obtained and ex-

amined visually for their interactions with the active-site residues

through hydrogen and aromatic bonding interactions. The most

probable docking conformations were chosen.

BACE-1 kinetics: The increase of fluorescence during the substrate

cleavage was monitored by using a Tecan Safire spectrofluorimeter

(working at l_exc328 nm and l_emi398 nm wavelengths) in black

with clear bottom microtiter plates (Corning 3711, 384 wells). Typi-

cally, 100 mm sodium acetate buffer (pH 4.0; 9 mL) with 0.001%

Triton X-100 and BACE-1 full protein (His*Tag, human, recombi-

nant, NSO cells, purchased from Calbiochem (PF 125); 10 mL,

25 ng) were pre-incubated for 5 min at room temperature. Finally,

the reaction was initiated by addition of the FRET substrate (I–IV;

1 mL) in DMSO. The final substrate concentration range was 1–

100 mm (final concentration) and the final DMSO concentration

range was 5–10% for all assays. The enzymatic activity was mea-

sured by monitoring the increase of fluorescence intensity at the

emission maximum of 398 nm. The initial velocity, measured as

fluorescence intensity per time (V_f), was calculated from the slope

during the linear phase of cleavage, usually during the first 20 min.

Initial velocities were converted from units of fluorescence (V_f) into

concentration per unit time (V_c). Initial velocity data were fitted to

the Michaelis–Menten equation by using GraphPad Prism 4.0

(GraphPad Software Inc.) to calculate the V_maxand K_Mvalues.

Acknowledgements

The authors thank the European Union for the Postdoctoral

Intra-European Fellowship (IEF) Marie Curie to M.I.F.-B. in the 7th

Framework (Call: FP7-PEOPLE-IEF-2008).

BACE-1 inhibition assay: BACE-1 full protein (His*Tag, human, re-

combinant, NSO cells) was purchased from Calbiochem (PF 125)

and the rhodamine-derivative substrate, which contains the pep-

tide quencher sequence, RhoRVNLDAEFK (Panvera peptide, P2985),

was acquired from Invitrogen (Milan, Italy). Sodium acetate and

DMSO were obtained from common commercial suppliers. Purified

water from a Mili-RX system (Millipore, Milford, USA) was used to

prepare buffers and standard solutions. Spectrofluorimetric analy-

ses were carried out on Tecan Safire spectrofluorimeter (working at

l_exc544 and l_emi590 nm wavelengths) by using black with clear

bottom microtiter plates (Corning 3711, 384 wells). Stock solutions

of tested compounds were prepared at 10 mm in DMSO and dilut-

ed with 100 mm sodium acetate buffer (pH 4) containing 0.001%

Triton X-100. For each reaction, BACE-1 enzyme (10 mL; 18.6 nm,

final concentration) was incubated with the tested compound

(5 mL) for 60 min. The reaction was then started by addition of the

peptidic FRET substrate (Panvera peptide; 1 mL; 0.25 mm, final con-

centration). The final volume in each reaction is 20 mL. The mixture

was incubated at 288C for 60 min. To stop the reaction, BACE-

1 stock solution (20 mL; sodium acetate, 2.5m) was added to each

well. The fluorimetric assay was followed by monitoring the in-

crease of the fluorescence signal at 590 nm with time. The DMSO

concentration in the final reaction volume was maintained at 5%

to guarantee no significant loss of enzymatic activity. The fluores-

Keywords: aspartic proteases

inhibitors · peptides

· drug discovery · FRET ·

[1] a) M. Tobias, L. C. Yeh, E. Johnson, Aust. N. Z. J. Psychiatry 2008 , 42 ,

82 8 –8 36; b) C. Dufouil, A. Alpꢂrovitch, Rev. Prat. 2005, 55, 1869–1878.

[2] H. Fillit, J. Hill, Am. J. Geriatr. Pharmacother. 2005 , 3 , 3 9 –4 9.

[3] J. T. Jarrett, E. P. Berger, P. T. Lansbury, Biochemistry 1993 , 32 , 469 3 –

4697.

[4] J. Hardy, Curr. Alzheimer Res. 2006, 3, 71–73.

[5] D. Leung, G. Abbenante, D. P. Fairlie, J. Med. Chem. 2000 , 43 , 30 5 –3 41.

[6] T. B. Durham, T. A. Shepherd, Curr. Opin. Drug Discovery Dev. 2006, 9,

776–791.

[7] B. De Strooper, W. Annaert, P. Cupers, P. Saftig, K. Craessaerts, J. S.

Mumm, E. H. Schroeter, V. Schrijvers, M. S. Wolfe, W. J. Ray, A. Goate,

R. A. Kopan, Nature 1999 , 398 , 51 8 –5 22.

[8] D. W. Klaver, M. C. Wilce, H. Cui, A. C. Hung, R. Gasperini, L. Foa, D. H.

Small, Biol. Chem. 2010, 391, 849–859.

[9] X. Luo, R. Yan, Int. J. Clin. Exp. Pathol. 2010, 3, 618–628.

[10] A. K. Ghosh, N. Kumaragurubaran, J. Tang, Curr. Top. Med. Chem. 2005,

5, 1609–1622.

[11] A. K. Ghosh, N. Kumaragurubaran, L. Hong, S. S. Kulkarni, X. Xu, W.

Chang, V. Weerasena, R. Turner, G. Koelsch, G. Bilcer, J. Tang, J. Med.

Chem. 2007 , 50 , 239 9 –2 407.

[12] M. C. Maillard, R. K. Hom, T. E. Benson, J. B. Moon, S. Mamo, M. Bienkow-

ski, A. G. Tomasselli, D. D. Woods, D. B. Prince, D. J. Paddock, T. L.

ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemMedChem 2013, 8, 1041 – 1056 1055

CHEMMEDCHEM

FULL PAPERS

www.chemmedchem.org

Emmons, J. A. Tucker, M. S. Dappen, L. Brogley, E. D. Thorsett, N. Jewett,

S. Sinha, V. John, J. Med. Chem. 2007 , 50 , 77 6 –7 81.

[34] J. Marcinkeviciene, Y. Luo, N. R. Graciani, A. P. Combs, R. A. Copeland, J.

Biol. Chem. 2001 , 276 , 2379 0 –2 3794.

[13] a) M. S. Malamas, J. Erdei, I. Gunawan, J. Turner, Y. Hu, E. Wagner, K. Fan,

R. Chopra, A. Olland, J. Bard, S. Jacobsen, R. L. Magolda, M. Pangalos,

A. J. Robichaud, J. Med. Chem. 2010 , 53 , 114 6 –1 15 8; b) M. S. Malamas, J.

Erdei, I. Gunawan, K. Barnes, M. Johnson, Y. Hui, J. Turner, Y. Hu, E.

Wagner, K. Fan, A. Olland, J. Bard, A. J. Robichaud, J. Med. Chem. 2009 ,

52 , 631 4 –6 323.

[14] Z. Zhu, Z. Y. Sun, Y. Ye, J. Voigt, C. Strickland, E. M. Smith, J. Cumming, L.

Wang, J. Wong, Y. S. Wang, D. F. Wyss, X. Chen, R. Kuvelkar, M. E. Ken-

nedy, L. Favreau, E. Parker, B. A. McKittrick, A. Stamford, M. Czarniecki,

W. Greenlee, J. C. Hunter, J. Med. Chem. 2010 , 53 , 95 1 –9 65.

[15] E. W. Baxter, K. A. Conway, L. Kennis, F. Bischoff, M. H. Mercken, H. L.

Winter, C. H. Reynolds, B. A. Tounge, C. Luo, M. K. Scott, Y. Huang, M.

Braeken, S. M. Pieters, D. J. Berthelot, S. Masure, W. D. Bruinzeel, A. D.

Jordan, M. H. Parker, R. E. Boyd, J. Qu, R. S. Alexander, D. E. Brenneman,

A. B. Reitz, J. Med. Chem. 2007 , 50 , 426 1 –4 264.

[16] W. H. Huang, R. Sheng, Y. Z. Hu, Curr. Med. Chem. 2009, 16, 1806–1820.

[17] Y. S. Wang, C. Strickland, J. H. Voigt, M. E. Kennedy, B. M. Beyer, M. M.

Senior, E. M. Smith, T. L. Nechuta, V. S. Madison, M. Czarniecki, B. A.

McKittrick, A. W. Stamford, E. M. Parker, J. C. Hunter, W. J. Greenlee, D. F.

Wyss, J. Med. Chem. 2010 , 53 , 94 2 –9 50.

[18] R. Godemann, J. Madden, J. Krꢆmer, M. Smith, U. Fritz, T. Hesterkamp, J.

Barker, S. Hçppner, D. Hallett, A. Cesura, A. Ebneth, J. Kemp, Biochem-

istry 2009 , 48 , 1074 3 –1 0751.

[35] P. Colantonio, L. Leboffe, A. Bolli, M. Marino, P. Ascenzi, G. Luisi, Bio-

chem. Biophys. Res. Commun. 2008 , 377 , 75 7 –7 62.

[36] N. Micale, A. P. Kozikowski, R. Ettari, S. Grasso, M. Zappala, J. J. Jeong, A.

Kumar, M. Hanspal, A. H. Chishti, J. Med. Chem. 2006 , 49 , 306 4 –3 067.

[37] J. C. Kappel, G. Barany, J. Pept. Sci. 2005 , 11 , 52 5 –5 35.

[38] E. Sarubbi, P. F. Seneci, M. R. Angelastro, N. P. Peet, M. Denaro, K. Islam,

FEBS Lett. 1993 , 319 , 25 3 –2 56.

[39] S. I. Al-Gharabli, S. T. Shah, S. Weik, M. F. Schmidt, J. R. Mesters, D. Kuhn,

G. Klebe, R. Hilgenfeld, J. Rademann, ChemBioChem 2006 , 7 , 104 8 –

1055.

[40] a) M. Lisurek, B. Rupp, J. Wichard, M. Neuenschwander, J. P. v. Kries, R.

Frank, J. Rademann, R. Kꢇhne, Mol. Diversity 2010 , 14 , 40 1 –4 08; b) Y.

Cao, T. Jiang, T. Girke, Bioinformatics 2008 , 24 , i36 6 –i 374.

[41] L. Piazzi, A. Cavalli, F. Colizzi, F. Belluti, M. Bartolini, F. Mancini, M. Reca-

natini, V. Andrisano, A. Rampa, Bioorg. Med. Chem. Lett. 2008 , 18 , 42 3 –

426.

[42] C. Garino, T. Tomita, N. Pietrancosta, Y. Laras, R. Rosas, G. Herbette, B.

Maigret, G. Quꢂlꢂver, T. Iwatsubo, J. L. Kraus, J. Med. Chem. 2006 , 49 ,

427 5 –4 285.

[43] S. Marumoto, M. Miyazawa, Bioorg. Med. Chem. 2012 , 20 , 78 4 –7 88.

[44] Y. H. Choi, G. H. Yon, K. S. Hong, D. S. Yoo, C. W. Choi, W. K. Park, J. Y.

Kong, Y. S. Kim, S. Y. Ryu, Planta Med. 2008 , 74 , 140 5 –1 408.

[45] Y. H. Choi, M. Y. Yoo, C. W. Choi, M. R. Cha, G. H. Yon, D. Y. Kwon, Y. S.

Kim, W. K. Park, S. Y. Ryu, Planta Med. 2009 , 75 , 53 7 –5 40.

[46] G. Kemmer, S. Keller, Nat. Protoc. 2010 , 5 , 26 7 –2 81.

[47] W. Lin, L. Yuan, Z. Cao, Y. Feng, J. Song, Angew. Chem. 2010 , 122 , 38 5 –

389; Angew Chem. Int. Ed. 2010 , 49 , 37 5 –3 79.

[19] W. Yang, R. Fucini, B. T. Fahr, M. Randal, K. E. Lind, M. B. Lam, W. Lu, Y.

Lu, D. R. Cary, M. J. Romanowski, D. Colussi, B. Pietrak, T. J. Allison, J. M.

Munshi, D. M. Penny, P. Pham, J. Sun, A. E. Thomas, J. M. Wilkinson, J. W.

Jacobs, R. S. McDowell, M. D. Balliger, Biochemistry 2009 , 48 , 448 8 –

4496.

[48] SYBYL-X1.3: Tripos Inc., St. Louis, MO 63144 (USA).

[20] P. D. Edwards, J. S. Albert, M. Sylvester, D. Aharony, D. Andisik, O. Calla-

ghan, J. B. Campbell, R. A. Carr, G. Chessari, M. Congreve, M. Frederick-

son, R. H.-A. Folmer, S. Geschwindner, G. Koether, K. Kolmodin, J. Krumr-

ine, R. C. Mauger, C. W. Murray, L. L. Olsson, S. Patel, N. Spear, G. Tian, J.

Med. Chem. 2007 , 50 , 591 2 –5 925.

[49] A. N. Jain, J. Med. Chem. 2003 , 46 , 49 9 –5 11.

[50] A. N. Jain, J. Comput.-Aided Mol. Des. 2007 , 21 , 28 1 –3 06.

[51] A. N. Jain, J. Comput.-Aided Mol. Des. 2009 , 23 , 35 5 –3 74.

[52] G. Velmourougane, M. B. Harbut, S. Dalal, S. McGowan, C. A. Oellig, N.

Meinhardt, J. C. Whisstock, M. Klemba, D. C. Greenbaum, J. Med. Chem.

2011 , 54 , 165 5 –1 666.

[21] M. Congreve, G. Chessari, D. Tisi, A. J. Woodhead, J. Med. Chem. 2008 ,

51 , 366 1 –3 680.

[53] V. Limongelli, L. Marinelli, S. Cosconati, H. A. Braun, B. Schmidt, E. Novel-

lino, ChemMedChem 2007 , 2 , 66 7 –6 78.

[22] P. J. Hajduk, J. Greer, Nat. Rev. Drug Discovery 2007 , 6 , 21 1 –2 19.

[23] C. W. Murray, D. C. Rees, Nat. Chem. 2009 , 1 , 18 7 –1 92.

[24] C. W. Murray, T. L. Blundell, Curr. Opin. Struct. Biol. 2010 , 20 , 49 7 –5 07.

[25] a) J. Barker, S. Courtney, T. Hesterkamp, D. Ullmann, M. Whittaker, Exp.

Opin. Drug Discovery 2006 , 1 , 22 5 –2 36; b) T. Hesterkamp, J. Barker, A.

Davenport, M. Whittaker, Curr. Top. Med. Chem. 2007, 7, 1582–1591.

[26] M. F. Schmidt, J. Rademann, Trends Biotechnol. 2009 , 27 , 51 2 –5 21.

[27] a) M. F. Schmidt, A. Isidro-Llobet, M. Lisurek, A. El-Dahshan, J. Tan, R. Hil-

genfeld, J. Rademann, Angew. Chem. 2008 , 120 , 331 9 –3 323; Angew.

Chem. Int. Ed. 2008 , 47 , 327 5 –3 278; b) M. F. Schmidt, A. El-Dahshan, S.

Keller, J. Rademann, Angew. Chem. 2009 , 121 , 646 4 –6 467; Angew.

Chem. Int. Ed. 2009 , 48 , 634 6 –6 349; c) M. F. Schmidt, M. Groves, J. Rade-

mann, ChemBioChem 2011 , 12 , 264 0 –2 646.

[54] J. N. Freskos, Y. M. Fobian, T. E. Benson, M. J. Bienkowski, D. L. Brown,

T. L. Emmos, R. Heintz, A. LaBorde, J. J. McDonald, B. V. Mischke, J. M.

Molyneaux, J. B. Moon, P. B. Mullins, D. B. Prince, D. J. Pakkock, A. G.

Tomaselli, G. Winterrowd, Bioorg. Med. Chem. Lett. 2007 , 17 , 7 3 –7 7.

[55] J. N. Freskos, Y. M. Fobian, T. E. Benson, J. B. Moon, M. J. Bienkowski,

D. L. Brown, T. L. Emmons, R. Heintz, A. Laborde, J. J. McDonald, B. V.

Mischke, J. M. Molyneaux, P. B. Mullins, D. B. Prince, D. J. Paddock, A. G.

Tomasselli, G. Winterrowd, Bioorg. Med. Chem. Lett. 2007 , 17 , 7 8 –8 1.

[56] S. Weik, T. Luksch, A. Evers, J. Bçttcher, C. A. Sotriffer, A. Hasilik, H. G.

Lçffler, G. Klebe, J. Rademann, ChemMedChem 2006 , 1 , 44 5 –4 57.

[57] a) K. Bornemann, H. Nar, M. Bauer, S. Handschuh, M. Kostka, C. Dorner-

Ciossek, W. Stransky, K. Fuchs, C. Eickmeier, S. Peters, K. Klinder, J. Rade-

mann, S. Weik, PCT Int. Appl. WO2005113582 A1, 2005; b) C. Eickmeier,

K. Fuchs, S. Peters, C. Dorner-Crossek, N. Heine, S. Handschuh, K. Klin-

der, M. Kostka, PCT Int. Appl. WO2006103038 A1, 2006.

[28] L. Zhu, S. George, M. F. Schmidt, S. I. Al-Gharabli, J. Rademann, R. Hil-

genfeld, Antiviral Res. 2011 , 92 , 20 4 –2 12.

[29] a) S. Cosgrove, L. Rogers, C. M. Hewage, J. P. G. Malthouse, Biochemistry

2007 , 46 , 1120 5 –1 121 5; b) L. Toulokhonova, W. J. Metzler, M. R. Witmer,

R. A. Copeland, J. Marcinkeviciene, J. Biol. Chem. 2003 , 278 , 458 2 –4 589.

[30] F. Grꢇninger-Leitch, D. Schlatter, E. Kꢇng, P. Nelbçck, H. Dçbeli, J. Biol.

Chem. 2002, 277, 4687–4693.

[58] L. Piazzi, A. Cavalli, F. Belluti, A. Bisi, S. Gobbi, S. Rizzo, M. Bartolini, V.

Andrisano, M. Recanatini, A. Rampa, J. Med. Chem. 2007 , 50 , 425 0 –

4254.

[59] Y.-C. Cheng, W. H. Prusoff, Biochem. Pharmacol. 1973, 22, 3099–3108.

[31] J. Ermolieff, J. A. Loy, G. Koelsch, J. Tang, Biochemistry 2000 , 39 , 1245 0 –

12456.

[32] V. Porcari, L. Magnoni, G. C. Terstappen, W. Fecke, Assay Drug Dev. Tech-

nol. 2005 , 3 , 28 7 –2 97.

[33] L. Hong, G. Koelsch, X. Lin, S. Wu, S. Terzyan, A. K. Ghosh, X. C. Zhang, J.

Tang, Science 2000 , 290 , 15 0 –1 53.

Received: February 20, 2013

Revised: April 25, 2013

Published online on June 11, 2013

ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemMedChem 2013, 8, 1041 – 1056 1056

Article Doi

Alzheimer's disease: Identification and development of β-secretase (BACE-1) binding fragments and inhibitors by dynamic ligation screening (DLS)

DOI: 10.1002/cmdc.201300078

Source and publish data:

Authors:

Article abstract of DOI:10.1002/cmdc.201300078

Full text of DOI:10.1002/cmdc.201300078

Products guided by the article

R&D Labs maybe for 1447762-81-3

Relevant to this article

Hot Product