Synthesis and biological evaluation of direct thrombin inhibitors bearing 4-(piperidin-1-yl)pyridine at the P1 position with potent anticoagulant activity

Article

pubs.acs.org/jmc

Synthesis and Biological Evaluation of Direct Thrombin Inhibitors

Bearing 4‑(Piperidin-1-yl)pyridine at the P1 Position with Potent

Anticoagulant Activity

Modesto de Candia,^†Filomena Fiorella,^†,⊥Gianfranco Lopopolo,^†,▽Andrea Carotti,^‡

Maria Rosaria Romano,^†Marcello Diego Lograno,^†Sophie Martel,^§Pierre-Alain Carrupt,^§

Benny D. Belviso,^∥Rocco Caliandro,^∥and Cosimo Altomare*^,†

^†Dipartimento di Farmacia-Scienze del Farmaco, University of Bari “Aldo Moro”, Via E. Orabona 4, 70125 Bari, Italy

^‡Dipartimento di Chimica e Tecnologia del Farmaco, University of Perugia, Via del Liceo 1, 60123 Perugia, Italy

^§School of Pharmaceutical Sciences, University of Geneva, University of Lausanne, Quai Ernest-Ansermet 30, CH-1211 Geneva 4,

Switzerland

^∥Istituto di Cristallograﬁa, Consiglio Nazionale delle Ricerche, Via Amendola 122/o, 70126 Bari, Italy

S

* Supporting Information

ABSTRACT: The design and synthesis of a new class of

nonpeptide direct thrombin inhibitors, built on the structure of

1-(pyridin-4-yl)piperidine-4-carboxamide, are described. Start-

ing from a strongly basic 1-amidinopiperidine derivative (6)

showing poor thrombin (fIIa) and factor Xa (fXa) inhibition

activities, anti-fIIa activity and artiﬁcial membrane permeability

were considerably improved by optimizing the basic P1 and

the X-substituted phenyl P4 binding moieties. Structure−

activity relationship studies, usefully complemented with

molecular modeling results, led us to identify compound 13b, which showed excellent fIIa inhibition (K_i= 6 nM), weak

anti-Xa activity (K_i= 5.64 μM), and remarkable selectivity over other serine proteases (e.g., trypsin). Compound 13b showed in

vitro anticoagulant activity in the low micromolar range and signiﬁcant membrane permeability. In mice (ex vivo), 13b

demonstrated anticoagulant eﬀects at 2 h after oral dosing (100 mg·kg⁻¹), with a signiﬁcant 43% prolongation of the activated

partial thromboplastin time (aPTT), over controls (P < 0.05).

inhibition of antithrombin III (ATIII), require parenteral

administration, which makes their use outside the hospital

diﬃcult. Moreover, UFHs, whose anticoagulant responses need

monitoring, are not active against ﬁbrin-bound fIIa, have

unpredictable PK proﬁles, and can trigger formation of

antibodies, causing heparin-induced thrombocytopenia

(HIT).^2,5Other parenteral anticoagulants include the direct

thrombin inhibitors (DTIs), such as the peptides hirudin and

bivalirudin, and argatroban, which is the only synthetic small-

molecule DTI currently in use for intravenous administration.⁶

In the search for safer antithrombotic drugs, enzymes

(factors) of the blood coagulation cascade, such as fIIa, fXa,

and fVIIa, have been identiﬁed as drug targets.⁷Thrombin is a

trypsin-like serine protease that catalyzes the conversion of

soluble ﬁbrinogen into insoluble ﬁbrin in the ﬁnal step of clot

formation and primary hemostasis.^8,9The thrombin active site

features the catalytic triad, formed by Asp102, His57, and

Ser195, and four distinct substrate binding sites (S1−S4). The

S1 site is a major determinant of substrate speciﬁcity; it is a

INTRODUCTION

■

Thrombotic disorders, including deep venous thrombosis

(DVT), venous thromboembolism (VTE), pulmonary embo-

lism (PE), myocardial infarction, unstable angina, and stroke,

are major causes of morbidity and mortality worldwide, with a

higher incidence in developed countries.¹Current antithrom-

botic therapies mostly rely on traditional anticoagulants,

including vitamin K antagonists and heparins [both unfraction-

ated (UFHs) and low-molecular-weight heparins (LMWHs)].²

In spite of their proven clinical eﬃcacy, these agents have

limitations that restrict their usefulness in therapy.¹Vitamin K

antagonists, such as warfarin, interfering with biosynthesis of

coagulation factors (II, VII, IX, X), proteins C and S, have long

been the only approved oral anticoagulant treatments,³but

their use is often associated with considerable shortcomings,

such as drug and food interactions, highly variable pharmaco-

kinetics (PK) and pharmacodynamics (PD) proﬁles, and intra-

and interpatient variability in drug response. Warfarin-treated

patients need continuous monitoring of clinical coagulation

parameters and frequent dose adjustment to avoid bleeding

complications.⁴UFHs, LMWHs, and the indirect fXa inhibitor

fondaparinux, whose anticoagulant activity is mediated by the

Received: August 1, 2013

A

dx.doi.org/10.1021/jm401169a | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

Figure 1. Structures of approved oral direct thrombin inhibitors.

Figure 2. Isonipecotamide-based antithrombotic agents and molecular modiﬁcation of compound 6 for the SAR study.

deep pocket, mainly hydrophobic, containing the Asp189

residue at the bottom, which shows a preference for the Arg

residue in the protein substrate sequence. The proximal S2

pocket in the insertion loop 60A−D (Tyr60A, Pro60B,

Pro60C, and Trp60D) accommodates small-size hydrophobic

residues in substrate sequence (e.g., Pro). In the ﬂat S3 site,

ligands (both substrates and inhibitors) may bind in an

antiparallel hydrogen bonding pattern usually involving Gly216

in fIIa. The large hydrophobic S4 site, also called D-pocket (D

stands for distal from the catalytic Ser195) or aryl binding site

because of its preference for large and aromatic moieties in low-

molecular weight ligands, is lined by the hydrophobic residues

Leu99, Ile174, and Trp215.¹⁰⁻¹²

In the past decade, extensive eﬀorts have been made to

develop direct active-site small-molecule DTIs,^2,3but despite

the large number of discovered potent compounds, only a few

DTIs are clinically available. Among them, only one synthetic

DTI, argatroban (not shown), is currently in use for

intravenous administration in patients with thrombosis and

HIT.¹³The ﬁrst oral DTI, ximelagatran (2, Figure 1), a double

prodrug converted in vivo to the active parent compound

melagatran (1),¹⁴was approved for the treatment of VTE in

Europe; however, it was withdrawn from the market in 2006

because of its hepatotoxicity.¹⁵The most recently approved

DTI was dabigatran etexilate (4),¹⁶which is an orally available

double prodrug of dabigatran (3) indicated for the prevention

of VTE in patients undergoing total hip replacement and total

knee replacement, as well as prevention of stroke in patients

with atrial ﬁbrillation.¹⁷

development of a new DTI lacking a highly basic P1 moiety,

with good oral absorption, is still a challenge.

We previously described (iso)nipecotamide-based inhibitors

of platelet aggregation induced by adenosine 5′-diphosphate

(ADP).¹⁸Subsequently, some of them were suitably modiﬁed,

providing compounds endowed with potential dual activity,

that is, both inhibition of ADP-induced platelet aggregation and

fXa inhibitory activities.¹⁹During the development of these

compounds with dual function against thrombosis, the

introduction of the highly polar amidine group on the

piperidine nitrogen of the isonipecotamide derivative 5 (Figure

2), which is a moderate antiplatelet agent (IC₅₀= 68 μM)

devoid of activity against the blood coagulation factors fIIa and

fXa, yielded compound 6, which exhibited some inhibition of

the main blood coagulation factors with a preference for

thrombin (K_i= 20.8 μM) while losing antiplatelet activity (IC₅₀

= 768 μM).

As suggested by our early docking calculation performed with

Glide,²⁰compound 6 can bind into the catalytic site of bovine

thrombin (PDB code 1UVT)²¹through the following three

main interactions (Figure 3): (i) a salt bridge and a bifurcated

hydrogen bonding between amidinium and carboxylate groups

of Asp189 in the S1 pocket; (ii) H-bond of the amide NH to

the backbone carbonyl group of Gly216 (S3); (iii) hydrophobic

binding and CH···π interactions with Trp215 of the 4-F-phenyl

group within the S4 site. The small hydrophobic S2 pocket

appeared to be not involved in any interaction with the ligand.

In order to obtain new orally bioavailable fIIa/fXa inhibitors,

we modiﬁed the structure of compound 6, ﬁrst exploring the

eﬀects of replacement of the P1 amidine group with less polar

basic moieties on both the enzymes’ inhibition potency and

membrane permeability, as assessed by a parallel artiﬁcial

membrane permeability assay (PAMPA),^22,23and then trying

to optimize physicochemical properties and positions of the

substituents on the P4 phenyl group (Figure 2). The eﬀects of

Both melagatran and dabigatran bear a highly basic

benzamidine group that interacts with Asp189 in the S1

pocket. Several classes of DTIs employ an amidine group as P1

moiety, but despite their eﬃcacy they suﬀer from poor oral

bioavailability. As with dabigatran etexilate, the prodrug

approach can be a solution to this problem. Nevertheless, the

B

dx.doi.org/10.1021/jm401169a | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

available ethyl isonipecotate, which was reﬂuxed in EtOH and

K₂CO₃with 2-chloropyrimidine (10a) or 2-bromopyridine

(10b)²⁴or 4-chloronitrobenzene (10d),²⁵followed by

hydrolysis of the ethyl esters with ethanolic NaOH. Compound

10c was instead prepared in satisfactory yield by a one-pot

reaction between ethyl isonipecotate, 4-chloropyridine hydro-

chloride, and triethylamine (TEA) in an EtOH/water mixture

in a sealed tube.²⁶The amide coupling between 10a−d and 3-

[(4-ﬂuorobenzyl)oxy]aniline, with TBTU/DIPEA as the

coupling reagents, ﬁnally produced compounds 11, 12, 13a,

and 14. The 4-nitrophenyl derivative 14 was then reduced to

the corresponding aminophenyl derivative 15 with hydrazine

and Raney Ni.

The intermediate 10c was further coupled with diverse 3-

substituted aniline derivatives, producing m-benzyloxyphenyl 1-

(pyridin-4-yl)isonipecotamides which diﬀer in the physico-

chemical features and positions of R⁴-substituents (Table 2), X-

Y linker, and amide group modiﬁcations (Table 3). The aniline

intermediates I−IV and 17 were synthesized through known

methods as shown in Scheme 3.

Figure 3. Schematic representation of the key interactions between

compound 6 and thrombin, according to molecular modeling; key

residues in the S1−S4 binding sites are shown in red.

The R⁴-substituted 3-(benzyloxy)anilines of series I were

prepared by benzylation of 3-nitrophenol, followed by

reduction of the nitro group with hydrazine and Raney Ni in

reﬂuxing MeOH. The 3-F derivative was reacted with ethyl

formate, followed by reduction with LiAlH₄, to give the

secondary amine derivative 17 in high yield. The OH group of

3-nitrophenol was alkylated via a Mitsunobu reaction;

subsequent hydrogenation over 10% Pd/C in EtOH provided

the phenylethoxy derivatives IIa and IIb. The aniline

intermediates IIIa,b were prepared by benzylation of 4(or 3)-

ﬂuorophenol with 1-(bromomethyl)-3-nitrobenzene, followed

by reduction of the NO₂group. The commercially available 3-

nitrobenzaldehyde was converted to the aniline intermediates

IVa,b by a Wittig reaction and subsequent catalytic hydro-

genation.

some changes in the X-Y linker and amide group (N-

methylation and reduction to amine) were also investigated.

In this work, we report the design and synthesis of N-[3-

(benzyloxy)phenyl]isonipecotamide derivatives and analogues,

most of them bearing the pyridin-4-yl group as the P1 moiety;

their evaluation as inhibitors of fIIa, fXa, and other related

serine proteases; and in vitro and ex vivo clotting assays. The

structure−activity relationship (SAR) study is supported by

molecular modeling and molecular dynamics (MD) results.

CHEMISTRY

■

The R¹-substituted 4-ﬂuorobenzyloxyphenyl isonipecotamide

derivatives were synthesized as shown in Schemes 1 (7−9) and

2 (11−13a, 15).

The coupling between 10c and the aniline intermediates I−

IV was performed by use of TBTU/DIPEA as the coupling

reagents in dry N,N-dimethylformamide (DMF) to prepare a

number of m-benzyloxyphenyl 1-(pyridin-4-yl)piperidine-4-

carboxamide derivatives bearing diverse R⁴-substituents (13,

18−26) and X-Y linkers (32−34), as shown in Scheme 4.

To verify the importance of the amide group, we investigated

the eﬀects of two simple modiﬁcations in 13b (i.e., the most

potent thrombin inhibitor of this series): (i) amide reduction

(29) and (ii) NH methylation (31). Compound 29 was

obtained in more than 90% yield by reduction of 13b with

BH₃−tetrahydrofuran (THF) complex. Unexpectedly, the

coupling of 10c with 17 produced 31 in very poor yield,

which required laborious chromatographic puriﬁcation. Com-

pound 31 was then more eﬃciently synthesized by arylation

with 4-chloropyridine of the amide intermediate 30, which was

obtained by coupling the Boc-protected isonipecotic acid and

the N-methylaniline 17, followed by the removal of the Boc

group with HCl.

Experimental details on synthesis and analytical data for

intermediates and ﬁnal products can be found in the

Experimental Section and Supporting Information.

While initial eﬀorts to crystallize the most active compounds

in complex with thrombin were unsuccessful, the crystal

structure of the free ligand 13b, as a representative example,

was determined by X-ray diﬀraction. The structure of 13b is

shown in Figure 4 (crystal data and details on crystal packing

are given in Supporting Information).

Scheme 1. Synthetic Methods for Preparation of R¹-Bearing

N-[3-(4-Fluorobenzyloxy)phenyl]piperidine-4-carboxamide

a

Derivatives 7−9

a

Reagents and conditions: (a) CH₃C(=NH)OC₂H₅·HCl, TEA, EtOH,

reﬂux, 6 h; (b) Me₂CO, Na(CN)BH₃, MeOH, rt, overnight; (c) 4-

(bromomethyl)pyridine hydrobromide, K₂CO₃, dry DMF, rt, 48 h.

The acetamidine-bearing compound 7 was synthesized by

reﬂuxing the already reported compound 5¹⁹with ethyl

acetimidate hydrochloride in EtOH. The 1-isopropylpiperidine

derivative 8 was synthesized by one-pot reductive alkylation,

with acetone and Na(CN)BH₄. Compound 9 was synthesized

in almost quantitative yield by reacting 5 with 4-bromome-

thylpyridine hydrobromide.

The preparation of 1-aza-heteroaryl-substituted piperidine-4-

carboxamide derivatives 11−13a and 15, as shown in Scheme 2,

proved to be more eﬃcient than N¹-arylation of compound 5.

The 1-aza-heteroaryl-substituted isonipecotic acid intermedi-

ates (10a−d) were prepared starting from commercially

C

dx.doi.org/10.1021/jm401169a | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

Scheme 2. Synthetic Methods for Preparation of N¹-Aryl Derivatives of N-[3-(4-Fluorobenzyloxy)phenyl]piperidine-4-

a

carboxamide, 11−15

a

Reagents and conditions: (a) absolute EtOH, K₂CO₃, reﬂux 4−40 h; (b) for 10c, 4-chloropyridine hydrochloride, TEA, EtOH/water (1:3 v/v),

sealed tube, 140 °C, 5 days; (c) NaOH, EtOH, reﬂux, overnight; (d) TBTU, DIPEA, dry DMF, rt, 72 h; (e) N₂H₄·H₂O, Raney Ni, MeOH, reﬂux, 1

h.

α-chymotrypsin (αCT). The K_ivalues are listed in Tables 1−3

and 6.

Scheme 3. General Synthetic Methods for Preparation of 3-

Substituted Aniline Intermediates I−IV and 17

a

Replacement of the P1 Amidine Group with Less

Polar Basic Moieties. We began by exploring the eﬀects on

fIIa/fXa inhibition potency and membrane permeability, as a

main property aﬀecting oral bioavailability, due to the

replacement of the highly polar basic amidine group as P1

moiety in compound 6. The enzymes’ inhibition constants (K_i,

micromolar), experimental ionization constants (pK_a), and

permeability data (log P_e) are reported in Table 1.

Replacement of the amidine group in 6 with ethanimine (7),

and even more with the isopropyl group (8), resulted in a

decrease of inhibition potency, whereas among the 1-aza-

heteroaryl-substituted compounds only the 1-pyridin-4-yl (13a)

displayed a substantial (100-fold) increase in potency against

fIIa (K_i= 0.211 μM) compared to 6 (K_i= 20.8 μM). The

comparison between the inhibition data and pK_avalues of the

1-aza-heteroaryl compounds supports the importance of the

protonation degree and position of nitrogen within the aryl R¹

group. For the most potent inhibitor 13a, only one pK_avalue

(8.9) was detected that corresponds to protonation of the

pyridine nitrogen. Indeed, according to reported values and

expert systems’ calculations, the piperidine nitrogen is much

less basic (ACDLabs calculated pK_aca. −4.8), because of its

aromatic conjugation with pyridin-4-yl group.²⁷The same

applies to compounds 11 and 12. On the basis of the

experimental pK_a, the pyridine nitrogen in 13a should exist

predominantly (ca. 89%) in protonated form at the assay pH

(8.0) and be in a suitable position to form a salt bridge to

Asp189 (S1), as shown by X-ray structures of thrombin in

complex with other 4-pyridyl-containing fIIa inhibitors.^28,29

The 1-pyrid-2-yl isomer 12, while having a basicity similar to

that of 13a, most likely loses potency because of the unsuitable

a

Reagents and conditions: (a) X = OH, (R⁴-substituted)benzyl

bromide, K₂CO₃, dry DMF, rt, overnight; (b) N₂H₄·H₂O, Raney Ni,

MeOH, reﬂux, 1−4 h; (c) ethyl formate, reﬂux, 10 h; (d) LiAlH₄, dry

THF, rt, overnight; (e) X = OH, 4(or 3)-(ﬂuoro)phenethyl alcohol,

PPh₃, DIAD, dry THF, rt, 48 h; (f) H₂, 10% Pd/C, EtOH, rt, 8 h; (g)

X = CH₂Br, 4(or 3)-ﬂuorophenol, K₂CO₃, dry DMF, rt, overnight; (h)

X = CHO, 4(or 3)-(ﬂuorobenzyl)triphenylphosphonium bromide,

DBU, rt, 24 h.

RESULTS AND DISCUSSION

■

Inhibition constant (K_i) values for the newly synthesized

compounds toward bovine fIIa and human fXa were measured

via a chromogenic assay. The most active compounds were also

tested for their inhibition potencies toward puriﬁed human fIIa

and some related serine proteases, such as human recombinant

fVIIa, human leukocyte elastase (LE), bovine trypsin (Try), and

D

dx.doi.org/10.1021/jm401169a | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

Scheme 4. Synthetic Methods for Preparation of Meta-Substituted N-Phenyl-1-(pyridin-4-yl)piperidine-4-carboxamide

a

Derivatives 13 and 18−34

a

Reagents and conditions: (a) TBTU, DIPEA, dry DMF, rt, 72 h; (b) BH₃-THF, rt, 24 h; (c) DCC, HOBt, dry THF, rt, 48 h; (d) HCl gas, CHCl₃;

(e) 4-chloropyridine hydrochloride, absolute EtOH, TEA, reﬂux, 48 h.

The experimental eﬀective permeability data at pH 6.8 are

spread over a range of 2 log units. According to reported

criteria,³¹compounds 13a and 15 can be considered as highly

permeable (P_e> 1 × 10⁻⁶cm·s⁻¹), whereas 5 and 6 have low

permeability (0.1 × 10⁻⁶≤ P_e< 1 × 10⁻⁶cm·s⁻¹). Two

compounds, namely 9 and 11, showed no permeability, most

likely due to their slower diﬀusion through the unstirred water

layer at the HDM/water interface. As a matter of fact, the

amounts of 9 and 11 in the acceptor compartment were below

the limits of the quantiﬁcation via UV spectrophotometry and

Figure 4. ORTEP drawing of 13b crystal structure. C, O, N, H, and F

P_ecould not be determined. In line with expectations, the

permeability of 13a, and of the weaker fIIa/fXa inhibitors 5 and

6 as well, slightly increased with decreasing acidity (log P_e

values at pH 8; Table 1), whereas the less basic compounds 9

and 11 remained not permeable and 15 precipitated at the

HDM/water interface. Reported permeability−pH proﬁles

showed that the apparent pK_aof weak organic bases in

HDM-PAMPA can be shifted toward lower pH.³⁰However, the

limited property space examined does not allow us to fully

understand the relationships between structure and membrane

permeation of the investigated compounds. We can just

highlight that, for the permeable compounds at pH 6.8, the

permeability increases as a function of the solvent-accessible

surface area (SASA) descriptor calculated in silico on the ligand

conformers in the top-ranked poses found by docking

calculations (see below), using a probe with 1.4 Å²radius, by

the module QikProp v 3.4 in Schrodinger Suite 2011. Indeed,

the compounds with medium permeability, 5 (log P_e= −6.2)

and 6 (log P_e= −6.0), have SASAs of 653 and 700 Å²,

respectively, whereas compounds 13a (log P_e= −4.7) and 15

(log P_e= −4.0), showing high permeability, have SASAs of 756

and 780 Å², respectively. No clear relationship was observed

with polar surface area (PSA) and other surface descriptors.

Compound 13a, which combines the highest fIIa/fXa

inhibition potency with the highest artiﬁcial membrane

permeability, was chosen as starting point for further

optimization of the P4 moiety.

atoms are shown in gray, red, light blue, white, and yellow,

respectively. Ellipsoids represent the reﬁned thermal factors. The O

atom of methanol used as the crystallization solvent appears in the

asymmetric unit of compound 13b, whereas the C atom has not been

unambiguously detected due to the poor electron density around the

O atom. The 13b crystal structure is stabilized by π−π stacking

intermolecular interactions between the 3-F-Ph moieties and between

the pyridyl groups; also, a H-bond network between the N atom of

pyridine, the amide CO and NH of two symmetry-related molecules,

and the O atom of a MeOH molecule is involved in the formation of

13b crystal packing.

location of pyridine nitrogen. Due to their lower basicity, and

unfavorable nitrogen position as well, the 1-pyrimidin-2-yl (11)

and 1-pyridin-4-ylmethyl (9) derivatives are orders of

magnitude less potent than 13a. The same arguments can be

used for explaining the lower potency of 15. Overall, even

though there is no direct relationship with anti-fIIa activity, the

ionization state proved to be important in enhancing the

binding aﬃnity to the active site’s S1 pocket, as the data for the

1-aza-heteroaryl-substituted compounds showed (Table 1).

The eﬀective permeability (P_e) of the majority of R¹-

substituted isonipecotanilide-based fIIa inhibitors was measured

by a hexadecane membrane (HDM) PAMPA. The HDM-

PAMPA P_evalues, which were shown to correlate with

gastrointestinal (GI) absorption in humans for a number of

drugs,³⁰were taken as estimates of the abilities of our fIIa/fXa

inhibitors to passively cross the GI tract, and then to have

potential oral bioavailability, at the early stage of our molecular

optimization study.

Eﬀects on fIIa/fXa Binding Aﬃnities of R⁴Substitu-

ents at the Distal Phenyl Group. Starting from compound

E

dx.doi.org/10.1021/jm401169a | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

Table 1. Eﬀects of Replacement of the P1 Amidine Group on Thrombin and Factor Xa Inhibition Potency and Artiﬁcial

a

Membrane Permeability (PAMPA)

a

Inhibition constants against bovine thrombin (fIIa) and human activated factor X (fXa) are means of three duplicate determinations (SEM < 5% of

the mean). Compounds were tested at a maximum concentration of 250 μM. ^bExperimental pK_avalues were determined by potentiometric titration

(Sirius GLpKa). Calculated pK_avalues with ACD/Laboratories software (v. 12.0). Log of the eﬀective permeability coeﬃcient, P_e(cm·s⁻¹), as

c

d

assessed by a parallel artiﬁcial membrane permeability assay (HDM-PAMPA) targeting gastrointestinal absorption (testosterone, log P_e= −3.4, was

e

used as reference compound); np = not permeable. Precipitation of solute at the hexadecane/water interface.

13a, we investigated the eﬀects on fIIa/fXa inhibition potency

of a number of R⁴substituents, which diﬀer in electronic

properties, lipophilicity, bulkiness, and hydrogen-bonding

capacity, at the para, meta, and ortho positions of the P4

phenyl moiety (Table 2).

the halogen, given that the meta-ﬂuorinated derivative 13b

displayed 26- and 3500-fold increase in potency over the

bulkier 3-Cl (19b) and 3-Br (20b) compounds, respectively.

Evidence of the role of F-bonding (i.e., ﬂuorophilicity) in

enhancing the binding aﬃnity to thrombin and other related

serine proteases of inhibitors bearing ﬂuorinated aromatic

moieties as P4 fragments has been already reported.^19,32,33In

the case of our inhibitors, once the most common protein−

ligand interactions involving ﬂuorinated P4 groups have been

considered and the modeling results have been taken into

account (see below), we believe that the highly polar C−F

bond in meta position is suitably located for close contacts with

the backbone carbonyls of amino acid residues inside the S4

pocket (e.g., Thr98, Glu97), thereby allowing the P4 moiety to

establish therein C−F···C=O dipole−dipole interactions, which

enhance the binding aﬃnity and inhibitory potency.³²

With only three exceptions (18, 20b, and 21b), all the R⁴-

substituted analogues showed fIIa selectivity. Importantly, the

meta-F compound (13b), while improving selectivity toward

fIIa over fXa by a factor of about 1000, showed K_iof 6 nM, with

nearly 35-fold increase over the para-F isomer 13a and 10-fold

increase over the ortho-F isomer 13c. Introduction of a second

ﬂuorine, such as in 13d (2,4-F₂) and 13e (3,5-F₂), did not

increase the potency compared to monoﬂuorinated com-

pounds; only the 3,5-F₂derivative 13e maintained a potency

close to that of 2-F congener 13c but 10 times less active than

the most potent compound 13b. The replacement of F atom

with Cl and Br, at either meta or para position, resulted in a

lowering of the inhibition potency against fIIa (and fXa as well).

With the monohalogenated compounds, it clearly appears that

meta substitution (13b, 19b, and 20b), compared to para

substitution (13a, 19a, and 20a), is more sensitive to the size of

Besides the halogen substituents, we explored a number of

other groups diﬀering in electronic, lipophilic, steric, and H-

bonding properties (Table 2). None of the substituents

examined, at any position on the P4 phenyl group, showed

inhibitory potency higher than that of the 3-F-Ph derivative

F

dx.doi.org/10.1021/jm401169a | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

Table 2. Eﬀect of R⁴Substituents on Binding Aﬃnity to fIIa

and fXa of m-Benzyloxyphenyl 1-(Pyridin-4-yl)piperidine-4-

carboxamides

a

K_i(μM)

compd

R⁴

fIIa

fXa

7.46

18

H

39.5

13a

13b

13c

13d

13e

19a

19b

20a

20b

21a

21b

22a

22b

22c

23a

24a

24b

25a

26a

26b

26c

26d

26e

26f

4-F

0.211

0.006

0.064

4.64

1.10

5.64

6.83

8.80

3-F

2-F

2,4-F₂

3,5-F₂

4-Cl

3-Cl

4-Br

3-Br

4-NO₂

3-NO₂

0.063

0.260

0.156

2.03

27.0

29.6

>250

37.1

6.31

Figure 5. Enzymes’ inhibition potency data of positional isomers at P4

3-ﬂuorobenzyloxy moiety of compound 13b.

20.9

inhibitory eﬀects at the highest concentration tested (250 μM),

and the para isomer 28 proved to be 3 orders of magnitude less

potent than 13b against both enzymes, which clearly showed

that the meta-substituted isomer may be able to achieve a more

correct orientation of the moieties interacting with the fIIa

active site’s pockets.

Exploration of Modiﬁcations of the Amide Group and

X-Y Linker. The removal of H-bond donor and acceptor in the

central amide group of 13b, through its reduction to amine

(29) and N-methylation (31), led to a loss of activity toward

fIIa but not fXa (which remained low), supporting the

importance of H-bonds between the ligand and the backbone

CO and/or NH of Gly216 in the S3 ﬂat site (Table 3). To our

5.85

6.00

6.73

7.73

26.8

4-SO₂CH₃

3-SO₂CH₃

2-SO₂CH₃

4-CF₃

0.124

15.9

15.6

12.3

35.1

48.5

25.0

36.1

12.2

11.1

3.40

4-NH₂

10.0

3-NH₂

13.3

4-CH₂N(CH₃)₂

4-OCH₃

8.19

2.03

3-OCH₃

1.15

7.23

16.4

0.606

126

3.72

2-OCH₃

1.30

3,4-(OCH₃)₂

3,5-(OCH₃)₂

3,4-OCH₂O-

0.031

0.083

0.053

Table 3. Eﬀects of Modiﬁcations of the Amide Group and X-

Y Linker on Binding Aﬃnity to fIIa and fXa

a

Inhibition constants against bovine thrombin (fIIa) and human

a

activated factor X (fXa) are means of three duplicate determinations

(SEM < 5% of the mean). Compounds were tested at a maximum

concentration of 250 μM.

K_i(μM)

amide

compd

modiﬁcation

X-Y linker

R³

fIIa

0.006

>250

4.46

0.488

26.8

18.2

3.72

12.4

fXa

5.64

3.83

23.6

13b

29

CONH

CH₂NH

CON(CH₃)

CONH

OCH₂

3-F

4-F

3-F

4-F

3-F

4-F

3-F

OCH₂

13b. Moreover, no signiﬁcant correlation (quantitative SAR)

was found between pK_ivalues and lipophilicity, electronic, or

steric parameters, alone or in linear/nonlinear combinations,

suggesting that the R⁴substituents, irrespective of their

electron-withdrawing/donating properties and/or hydrophilic/

hydrophobic nature but in part depending on their bulkiness

and/or H-bonding capacity, may induce conformational

changes in the inhibitors, which then bind to the enzyme

with diﬀerent poses, more or less productive in terms of

potency.

31

OCH₂

32a

32b

33a

33b

34a

34b

OCH₂CH₂

CH₂O

0.133

2.14

0.774

CH₂O

>250

37.7

44.1

CH₂CH₂

a

Inhibition constants against bovine thrombin (fIIa) and human

activated factor X (fXa) are means of three duplicate determinations

(SEM < 5% of the mean). Compounds were tested at a maximum

concentration of 250 μM.

Among the investigated R⁴substituents, the electron-

withdrawing and H-bond acceptor substituent SO₂CH₃in the

para position (22a) achieved nanomolar fIIa potency (124 nM)

in the chromogenic assay. Notably, introducing two methoxy

substituents on the P4 phenyl group with a 3,4-substitution

pattern (26d) enhanced the potency with a K_ivalue in the low

nanomolar range, however, with a marked decrease of

selectivity over fXa. The almost equipotent 3,5-(OCH₃)₂

(26e) and 3,4-OCH₂O- (26f) analogues instead proved to be

more selective than 26d against fIIa. Actually, the piperonyl

fragment had already successfully been employed to bind the

thrombin S4 pocket, because it can strengthen binding by

establishing a H-bond with the hydroxyl group of Tyr60A in

the S2 subsite.³⁴

surprise, the X-Y linker appeared as a key factor for fIIa

inhibition potency, and OCH₂proved to be, among those

investigated in this study and a number of other more polar and

rigid fragments (unpublished results), the linker that aﬀorded

the highest thrombin inhibition potency and selectivity. As a

matter of fact, regardless of the position of F on the P4 phenyl

group, either elongation (32a,b) or isosteric replacement

(33a,b and 34a,b) of the OCH₂linker led to a loss in fIIa

potency of at least 2 orders of magnitude, whereas in only two

cases (32a and 33a) was a substantial improvement of fXa

inhibition potency observed.

The 3-F-benzyloxy ortho and para positional isomers (27

and 28, respectively) of the most active inhibitor 13b were also

tested (Figure 5). The ortho isomer 27 did not show signiﬁcant

We paid attention to variations in physicochemical properties

(e.g., basicity, lipophilicity) possibly aﬀecting the thrombin

inhibition potency of the isosteric 3-F-Ph derivatives 13b

G

dx.doi.org/10.1021/jm401169a | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

(excellent inhibitor) and 33b and 34b (weak inhibitors), but

the experimental measurements of pK_a(8.9, 8.8, and 8.7 for

13b, 33b, and 34b, respectively) and 1-octanol/water log D at

pH 7.4 (3.14, 3.02, and 3.58 for 13b, 33b, and 34b,

respectively) revealed no noteworthy variation in the ionization

state and lipophilicity, which may be somehow related to the

enzyme inhibition potency.

that of 13b (Figure 6B). They are closely superimposed in the

molecular region encompassing the P4 pyridine and

isonipecotamide moieties, with the amide NH forming a H-

bond with Gly216, but they diﬀer in folding of the P4 moiety,

which is well accommodated for 13b and ﬁts poorly or not at

all into the S4 hydrophobic pocket for 33b and 34b,

respectively.

Binding Mode Evaluation by Molecular Modeling. To

understand the presumable causes of the observed loss of

aﬃnity, we investigated the possible binding modes within the

binding site of thrombin of the most active inhibitor 13b

compared with inactive compounds 29 and 31 and weak

inhibitors 33b and 34b. The X-ray crystal structure of bovine

thrombin in complex with the known 4-aminopyridine-based

inhibitor BM14.1248 (PDB code 1UVT)²¹was retrieved from

the RCSB Protein Data Bank and used in docking calculations

with Glide software,²⁰which had been proven to perform better

than other automated docking programs with thrombin.³⁵All

ﬂexible docking calculations were performed in the extra

precision (XP) mode of Glide.^20cThe choice of the top-ranked

pose for each ligand was made using the model energy score

(Emodel), that combines Glide score, nonbonded interaction

energy, and excess internal energy of the generated ligand

conformation. As can be seen in Table 4, the Emodel score

For 34b, probably because of the greater ﬂexibility of the

CH₂CH₂linker compared to OCH₂(13b) and CH₂O (33b),

the phenethyl P4 group appears more folded, positioned farther

from the S4 site and pointed in the direction of the S2 pocket,

without any clashing with Tyr60A and Trp60D. This S2 pocket

is formed only in thrombin, whereas it is lacking in other

related serine proteases, including fXa and the digestive

enzymes chymotrypsin and trypsin.

On the basis of the Glide docking poses, it is reasonable to

assume that the improvement in fIIa binding aﬃnity of our

compounds is related to the extent of the hydrophobic contact

surface of the P4 side chain inside the S4 pocket and the H-

bond between ligand NHCO and Gly216, as the anchoring

binding of the P1 pyridine NH⁺and in the S1 pocket (Asp189)

is a constant feature in both very active (13b) and weak

inhibitors (33b and 34b). Our computational models compare

well enough with the data published a few years ago that

showed for a series of fIIa inhibitors a mutual cooperative

increase of hydrophobic interactions in the S4 pocket and H-

bond involving Gly216, whose global strength is higher than

that expected by simple additivity.³⁸For instance, in that study

Muley et al.³⁸proved that an NH₂group in the inhibitor that

engage H-bond with the Gly216 backbone carbonyl oxygen

may improve a close contact of the P4 moiety inside the S4

pocket by more than 59% (as assessed by the enhancement of

binding aﬃnity per square angstrom of hydrophobic contact

surface in the S4 pocket) over inhibitors lacking this H-bond-

donating group.

Table 4. Inhibition Constants and Calculated Interaction

Energies of the Best Docked Ligand Structures on Thrombin

compd

pK_i

Emodel (kcal/mol)

13b

29

8.22

−86.46

−73.37

−69.30

−75.01

−79.21

<3.6

4.74

4.91

31

33b

34b

In order to further investigate how the CONH···Gly216 H-

bond and P4−S4 hydrophobic contact in our three isosteric

inhibitors 13b, 33b, and 34b can interact with each other,

molecular dynamics (MD) simulations were performed on the

inhibitors (in their top-ranked docking poses) in complex with

thrombin (details on MD procedure are given in the

Experimental Section). An analysis of the MD trajectory over

2 ns was performed, and the average distances relative to key

interactions engaging the inhibitors’ groups and the protein

residues/binding pockets are summarized in Table 5.

The MD simulation data along the trajectory suggest that the

hydrogen-bonding salt bridge formed by the pyridinium NH⁺

with the Asp189 COO⁻in the S1 pocket is almost permanently

maintained in all three inhibitors; the average distances

between NH⁺and O1/O2 of the Asp189 COO⁻oscillate just

between 2.84 and 3.24 Å with standard deviations that in most

cases do not signiﬁcantly change. The H-bond NH···OC-

Gly216 (S3) is maintained along the trajectory only for 13b. In

parallel, the time-averaged mean hydrophobic P4 side chain (3-

F-Ph) distance to the S4 pocket is the shortest one (4.89 Å in

13b versus 6.76 and 6.85 Å in 33b and 34b, respectively), with

the lowest standard deviation. Overall, these MD simulation

data seem to indicate that the shorter and stronger H-bond

formed in 13b by the amide NH with the Gly216 backbone CO

most likely restricts the 3-F-benzyl movement inside the

hydrophobic S4 binding site, thereby allowing this P4 side

chain to spend more time in closer contacts (enhanced van der

Waals interactions) with the S4 pocket surface.

correlated reasonably well with the binding aﬃnity, according

to previous studies.^36,37In our case, given the high similarity of

the analyzed ligands, we were quite conﬁdent in using this

scoring function to pick out the top-scored docking poses for

visual inspection.

As shown in Figure 6, the most potent inhibitor 13b, in the

top-ranked solution (panels A and C), binds to the enzyme

pockets through three main interactions: (i) salt bridge,

strengthened by a H-bond, between the pyridine NH⁺and

the COO⁻of Asp189, at the bottom of the enzyme S1 pocket;

(ii) interaction of the meta-F-Ph group in the large hydro-

phobic S4 site; (iii) H-bond between the amide NH in 13b and

the Gly216 backbone CO (S3). No interaction with the small

hydrophobic S2 pocket of the enzyme was revealed by the

Glide search.

The N-methyl derivative 31 (Figure 6D) shows a complete

reversal of the docking relative to the binding pose of 13b: the

P4 moiety (1-pyridin-4-ylpiperidine), which binds deep in the

S1 pocket in the model of 13b, occupies in part the S4

hydrophobic/aryl binding site, with the pyridinium group

pointing out of the protein surface toward the solvent; the

meta-F-Ph group interacts with the S1 pocket; and the H-bond

between the amide group and Gly216 residue is lost. For

compound 29, Glide suggested two almost equally ranked

docking poses, the ﬁrst one resembling that of 13b and the

other being similar to that adopted by 31.

According to the Glide solutions, compounds 33b and 34b

adopt similar orientations but not overlapping conformations to

H

dx.doi.org/10.1021/jm401169a | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

Figure 6. Glide top-ranked poses of 1-(pyridin-4-yl)piperidine-4-carboxamide-based inhibitors into the binding site of thrombin (PDB code 1UVT).

The protein is shown in ribbon (panels A and B) or accessible surface representation (panels C and D), and the ligands and potentially interacting

residues are shown in stick format (blue, nitrogen; red, oxygen; cyan, ﬂuorine; carbon atoms, diﬀerent colors; yellow dashed lines, H-bonds). (A)

13b (green carbons); (B) overlay of 13b (green carbons), 33b (orange carbons), and 34b (yellow carbons); (C) 13b (green carbons); (D) 31

(purple carbons). The ﬁgures were drawn by PyMOL.

inhibition constants close to 7 μM for chymotrypsin, about 11

μM for human leukocyte elastase, and higher than 250 μM for

trypsin (Table 6). It is roughly 1000-fold less potent on fXa and

essentially inactive (K_i> 250 μM) on human recombinant fVIIa

involved in the blood coagulation cascade. Other 1-pyridin-4-yl

isonipecotamide derivatives chosen among the most potent fIIa

inhibitors in the series were also assayed and found to be as

selective as compound 13b.

Table 5. Average Distances in Isosteric Inhibitors 13b, 33b,

and 34b

a

time-averaged distance SD (Å)

13b

33b

34b

S1 Pocket: H-bond Pyridinium NH⁺···Asp189

NH⁺···O1(O)C-Asp189

NH⁺···O2(O)C-Asp189

3.10 0.31

2.84 0.26

2.98 0.36

2.94 0.24

2.90 0.37

3.24 0.54

S3 Pocket: H-bond CONH···Gly216

Compounds 13a (R⁴= 4-F) and 13b (R⁴= 3-F), which diﬀer

very much in aﬃnity for bovine fIIa, were equipotent in

inhibiting human wild-type thrombin. A striking diﬀerence in

binding aﬃnity to human and bovine fIIa was found not only

for compound 13a, which binds 15-fold better to human than

to bovine fIIa, but also for the 3,4-(OCH₃)₂derivative 26d, that

on the contrary binds 24-fold better to bovine than to human

fIIa. These diﬀerences cannot be explained by Glide modeling

calculations, which, when performed for investigating the

docking of 13a, 13b, and 26d to bovine fIIa (i.e., 1UVT) and

two representative human fIIa structures (PDB codes 1JWT³⁹

and 2BDY⁴⁰), provided almost similar binding modes and

CO···HN-Gly216

NH···OC-Gly216

5.61 1.10

2.41 1.00

4.29 0.59

6.41 0.56

7.53 1.65

4.56 0.89

b

S4 Pocket: 3-F-Ph Centroid−S4 Pocket Center

4.89 0.50 6.76 1.23

6.85 1.26

a

b

Recorded over 2 ns along the MD trajectory. Distance between the

centroid of the P4 m-ﬂuorophenyl group and the S4 pocket (calculated

as the middle point between Leu99, Ile174, and Trp215 residues).

Selectivity over Other Serine Proteases and Antico-

agulant Activity. The most potent fIIa inhibitor 13b proved

to be selective over a number of other serine proteases, with

I

dx.doi.org/10.1021/jm401169a | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

Table 6. Inhibition Data against Blood Coagulation Factors and Other Serine Proteases, in Vitro Anticoagulant Activities, and

Physicochemical Parameters of the Most Potent Thrombin Inhibitors Described in This Study

a

K_i(μM)

13a

0.211

13b

0.006

13c

0.064

13e

0.063

0.031

27.0

>250

31.4

>250

22a

0.124

26d

0.031

26e

26f

0.053

bfIIa

hfIIa

hfXa

rfVIIa

bαCT

bTry

hLE

0.083

0.014

1.10

0.016

5.64

0.090

6.83

0.755

0.606

7.73

>250

126

33.8

3.72

>250

71.6

>250

68.7

>250

6.60

>250

11.3

>250

71.6

27.0

>250

71.5

27.0

>250

Other Parameters

13a

13b

13c

13e

22a

26d

26e

26f

b

aPTT₂(μM)

59.0

8.9

6.60

9.0

19.8

9.2

26.3

9.2

98.2

9.1

154

9.2

200

8.9

>200

9.1

c

pK_a

log P^N

d

5.1

5.0

4.9

5.2

4.9

e

log P_e

−4.7

−4.8

−4.6

−5.0

−4.9

−4.5

−4.6

np

a

Inhibition constants against bovine and human thrombin, human activated factor X, human recombinant activated fVII, bovine α-chymotrypsin and

trypsin, and human leukocyte elastase are means of three duplicate determinations (SEM < 5% of the mean). Compounds were tested at a maximum

b

concentration of 250 μM. Concentrations of test compounds required to produce a doubling of the activated partial thromboplastin time (aPTT)

c

of uninhibited clotting time in human plasma. The maximum concentration tested was 500 μM. Experimental pK_avalues determined by

d

potentiometric titration (Sirius GLpKa). 1-Octanol/water partition coeﬃcient of neutral form, as assessed by hydrophilic interaction liquid

e

chromatography (HILIC; deﬁnitions and method details in Experimental Section). Log of eﬀective permeability coeﬃcients, P_e(cm·s⁻¹) as assessed

by a parallel artiﬁcial membrane permeability assay (PAMPA) targeting gastrointestinal absorption (testosterone, log P_e= −3.4, was used as

reference compound); np = not permeable.

scores. On the other hand, the human thrombin 1JWT and

2BDY structures, compared to the bovine thrombin 1UVT,

have more than 80% sequence identity and 100% identity in the

inhibitors’ binding sites (6 Å distance from the ligands’

structures). It is likely that the observed diﬀerences in binding

aﬃnity can be ascribed to a diﬀerent degree of purity of the

enzymes used in the assays and thus probably to dissimilar

interactions of the ligands with other components in the

enzymes’ samples.

phospholipids of the examined compounds, notably those

bearing methoxy and methylenedioxy substituents on the P4

phenyl ring, can cause a decrease in their anticoagulant activity.

The lipophilicity of the compounds in Table 6 was assessed

by hydrophilic interaction liquid chromatography (HILIC),

which is recognized to be suitable for accurate measurements of

1-octanol/water partition coeﬃcients (log P^N) of the neutral

form of basic compounds,⁴³by a method already reported for

similar compounds (details are given in the Experimental

Section).⁴⁴The log P values calculated from HILIC retention

data demonstrate that these basic compounds are nearly

isolipophilic in their neutral form (they span a range of just 0.3

log unit), whereas on the other hand suggest that lipophilicity

has no signiﬁcant eﬀect either on their fIIa/fXa inhibition or on

their anticoagulant potency.

The eﬀective permeability coeﬃcients (log P_e) as measured

by PAMPA (Table 6) showed that, with the remarkable

exception of the 3,4-OCH₂O- derivative (26f), which proved to

be not permeable (a result that corresponds to the loss of

activity in the aPTT clotting assay), all the most active

antithrombin compounds cross the artiﬁcial membranes (log P_e

values ranging from −4.5 to −5), with the most active fIIa

inhibitor 13b showing signiﬁcant permeability (log P_e= −4.8).

As all the permeable compounds show ﬁnite P_evalues

(including those reported in Table 1: log P^N= 3.7, 4.3, and

4.7 for compounds 5, 6, and 15, respectively), with the

exception of 15 (R¹= 4-NH₂-Ph), there is a trend of linear

correlation between log P_eand log P^N(n = 9, r²= 0.738; ρ =

1.1). In addition to PAMPA, the in vitro metabolic stability on

microsomes remains to be investigated for a better druglikeness

assessment of these anti-fIIa compounds.

Anticoagulant properties were assessed in vitro through the

activated partial thromboplastin time (aPTT) clotting assay in

pooled human plasma, and the anticoagulant potency was

expressed as the concentration of inhibitor required to double

the uninhibited clotting time (aPTT₂in Table 6). Compound

13b showed the best anticoagulant properties (aPTT₂= 6.6

μM), whereas compound 13a, albeit being an equipotent

inhibitor on hfIIa, showed a signiﬁcantly higher aPTT₂value

(59 μM). The diﬀerences in aPTT₂values of the compounds in

Table 6 are diﬃcult to understand on the basis of either

inherent fIIa/fXa inhibitory potencies or measured physico-

chemical properties. For example, compound 26f (fIIa K_i= 53

nM) proved to be more than 10-fold less active in the aPTT

clotting assay than 26d (fIIa K_i= 31 nM), although they have

similar thrombin inhibition potency, basicity, and lipophilicity.

In early studies on potent noncovalent DTIs, it had been shown

that the increase in plasma protein binding had detrimental

eﬀects on aPTT₂values and antithrombotic activities of DTIs,⁴¹

and more recently Remko,⁴²on the basis of a detailed analysis

of novel anticoagulant agents, concluded that compounds with

high binding aﬃnity to plasma components (e.g., proteins and

phospholipids), albeit being potent fIIa/fXa inhibitors,

generally show low eﬀects on clotting time prolongation in

plasma. In our case, the drop of activity in the aPTT assay,

whose data were uncorrelated with the relative fIIa/fXa

inhibition potency and/or lipophilicity, may indicate that the

increase in binding aﬃnity to plasma proteins and/or

Finally, for an early assessment of the oral anticoagulant

eﬀectiveness of compound 13b, we examined its ex vivo activity

after oral dosing in mice. Figure 7 shows that 13b, administered

via oral gavage at 100 mg·kg⁻¹dose, signiﬁcantly prolonged

plasma aPTT 1 h (28.9 0.4 s) and 2 h (30.4 3.3 s) after

J

dx.doi.org/10.1021/jm401169a | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

(100 mg·kg⁻¹). These data indicate that compound 13b, in

addition to being worthy of further pharmacological inves-

tigation, including a full ADMET characterization and in vivo

antithrombotic proﬁling in comparison with the reference

dabigatran etexilate, may be considered as a lead for further

optimization of potent DTIs and fIIa/fXa dual inhibitors, as

novel eﬃcacious oral anticoagulant drugs safer than the existing

ones.

EXPERIMENTAL SECTION

■

Chemistry. Melting points were determined by using the capillary

method on a Stuart Scientiﬁc SMP3 electrothermal apparatus and are

uncorrected. Elemental analyses (C, H, N) were performed on a Euro

EA3000 analyzer (Eurovector, Milan, Italy) by the Analytical

Laboratory Service of the Department of Pharmacy−Drug Sciences

of the University of Bari (Italy), and the results agreed to within

0.40% of theoretical values. Mass spectra were recorded on an

Agilent gas chromatograph−mass spectrometer GC-MS 6890-5973. IR

spectra were recorded via KBr disks on a Perkin-Elmer Spectrum One

Fourier transform infrared spectrophotometer (Perkin-Elmer Ltd.,

Buckinghamshire, U.K.), and the most signiﬁcant absorption bands are

listed. ¹H NMR spectra were recorded at 300 MHz on a Varian

Mercury 300 instrument. Chemical shifts are expressed in δ and the

coupling constants J are in hertz (Hz). The following abbreviations are

used: s, singlet; d, doublet; dd, doublet−doublet; t, triplet; m,

multiplet. Signals due to NH and OH protons were located by

deuterium exchange with D₂O.

Chromatographic separations were performed on silica gel 60 for

column chromatography (Merck 70−230 mesh, or alternatively 15−40

mesh for ﬂash chromatography). Unless otherwise stated, starting

materials, and all chemicals and solvents as well, were purchased from

Sigma−Aldrich. Several compounds were synthesized according to

known procedures with slight modiﬁcations; their melting points and

spectral data were in full agreement with those reported in literature,

and no eﬀort was made at this stage to optimize the yields.

Figure 7. Ex vivo anticoagulant activity of compound 13b (100 mg·

kg⁻¹) after oral dosing in mice. Animals were administered via oral

gavage with either vehicle (control) or test compound, and the

activated partial thromboplastin time (aPTT) was measured 1 and 2 h

after administration; aPTT values are expressed as means SEM of

duplicate determinations in three mice per group (*P < 0.05 in

Dunnett’s test).

treatment versus controls (21.2

4.9 s), both displaying

signiﬁcant clotting time prolongation (P < 0.05 in Dunnett’s

test). The mean percent aPTT prolongation was quantiﬁed as

36% and 43% at 1 and 2 h postadministration, respectively.

CONCLUSIONS

■

While many research groups, in both academia and industry,

have attempted to develop new orally bioavailable DTIs, with

wide therapeutic windows and ﬁxed-dose administration, which

could be safer and easier to use compared with warfarin, only

the double prodrug dabigatran etexilate (4)¹⁶has been

approved for prevention of stroke in patients with atrial

ﬁbrillation.^17cIn this study, we report new isonipecotamide-

based compounds that are eﬀective direct thrombin inhibitors

(DTIs), showing in vitro and ex vivo anticoagulant activity in

plasma. Starting from the simple N-[3-(4-ﬂuorobenzyloxy)-

phenyl]isonipecotamide bearing the amidine group at the

piperidine N1 (6), that showed modest anti-fIIa and fXa

potencies and poor membrane permeability (PAMPA), both

enzymes’ activities were signiﬁcantly improved, with a net

increase in the binding aﬃnity toward fIIa. Structure−activity

relationship studies, usefully complemented and supported by

comparative molecular docking calculations and molecular

dynamics simulations on some inhibitors in complex with

thrombin, allowed us to expand the knowledge of the

determinants for fIIa aﬃnity and selectivity by N¹- and R⁴-

substituted-(benzyloxyphenyl)isonipecotamides, and led to

identifying a number of new potent fIIa inhibitors with artiﬁcial

membrane permeability (Table 6). Replacement of the 1-

amidino group with 1-pyridin-4-yl group as the P1 moiety and,

in particular, the displacement of ﬂuorine from para to meta

position of the P4 phenyl group, aﬀorded compound 13b,

which showed excellent antithrombin activity (K_i= 6 nM) in

the same range of dabigatran,¹⁶weak fXa inhibition (K_i= 5.64

μM), and remarkable selectivity versus a panel of serine

proteases (>1000-fold lowest ratio). According to molecular

modeling, 13b closely ﬁts in the fIIa active site’s pockets S1, S3,

and S4 (and not S2): the pyridinium group binds deep in the

S1 pocket (Asp189), the amide NH forms a strong H-bond

with the backbone CO of Gly216 in the ﬂat S3 site, and the P4

3-F-Ph moiety binds to the hydrophobic S4 pocket. As

expected by the PAMPA estimates and in vitro clotting assay

in pooled human plasma, the most potent fIIa inhibitor 13b,

compared with controls, showed a 43% prolongation of clotting

time in an ex vivo aPTT assay in mice at 2 h after oral dosing

1-Ethanimidoyl-N-{3-[(4-ﬂuorobenzyl)oxy]phenyl}-

piperidine-4-carboxamide Hydrochloride (7). To a solution of

365 mg (1 mmol) of 5·HCl in 10 mL of absolute EtOH were added

741 mg (6 mmol) of ethyl acetimidate hydrochloride and 0.98 mL (7

mmol) of triethylamine, and the reaction mixture was reﬂuxed for 6 h.

After cooling and solvent removal, the residue was dissolved in 30 mL

of EtOAc and the organic phase washed with 3 × 10 mL of brine,

dried over Na₂SO₄, ﬁltered, and concentrated. The crude oil was taken

up in 10 mL of chloroform and treated with HCl gas. After solvent

removal, 260 mg (64% yield) of 7 was obtained as a yellow oil. IR

1

(ﬁlm) 3415, 1670, 1606, 1223, 1154, 1934, 825, 775 cm⁻¹. H NMR

(300 MHz, DMSO-d₆) δ 10.18 (s, 1H), 9.30 (s, 1H), 8.73 (s, 1H),

7.56−7.37 (m, 3H), 7.27−7.07 (m, 4H), 6.67 (d, J = 7.0 Hz, 1H), 5.02

(s, 2H), 4.13 (d, J = 14 Hz, 1H), 3.92 (d, J = 14 Hz, 1H), (d, J = 12

Hz, 1H), 3.16 (d, J = 12 Hz, 1H), 2.73 (m, 1H), 2.28 (s, 3H), 1.92 (d,

J = 10.5 Hz, 2H), 1.80−1.60 (m, 2H). ESIMS m/z 370 (MH⁺). Anal.

Calcd for C₂₁H₂₅N₃O₂F × HCl: C, 57.07; H, 6.61; N, 9.51. Found: C,

57.32; H, 6.49; N, 9.33.

N-{3-[(4-Fluorobenzyl)oxy]phenyl}-1-isopropylpiperidine-4-

carboxamide Hydrochloride (8). To a solution of 300 mg (0.82

mmol) of 5·HCl in 10 mL of MeOH and 5 mL of acetone was added

portionwise 103 mg (1.64 mmol) of Na(CN)BH₃, and the reaction

mixture was stirred at room temperature overnight. After solvent

removal, the residue was dissolved in 50 mL of EtOAc and the organic

phase was washed three times with 5% NaHCO₃solution and brine,

dried over Na₂SO₄, ﬁltered, and concentrated. The crude oil was taken

up in 10 mL of chloroform and treated with 5 mL of 1.25 M HCl

solution in MeOH. After solvent removal, a solid residue was obtained,

which was recrystallized from EtOAc/EtOH, and 110 mg (36% yield)

of 8 was obtained as a brown solid, mp 145.3−148 °C. IR (KBr) 3223,

1

2944, 1646, 1606, 1226, 1154, 1012, 839, 776 cm⁻¹. H NMR (300

MHz, CDCl₃) δ 7.47 (t, J = 1.0 Hz, 1H), 7.42−7.36 (m, 2H), 7.21 (br

s, 1H), 7.20 (t, J = 8 Hz, 2H), 7.06 (t, J = 8.5 Hz, 2H), 6.91 (dd, J₁=

8.0 Hz, J₂= 1.5 Hz, 1H), 6.70 (dd, J₁= 7.5 Hz, J₂= 1.5 Hz, 1H), 5.02

K

dx.doi.org/10.1021/jm401169a | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

(s, 2H), 2.99 (d, J = 12 Hz, 2H), 2.85 (heptet, J = 8.0 Hz, 1H), 2.35−

2.15 (m, 2H), 2.05−1.95 (m, 2H), 1.07 (d, J = 8.0 Hz, 6H). ESIMS m/

z 371 (MH⁺). Anal. Calcd for C₂₂H₂₈N₂O₂F × HCl: C, 64.94; H, 6.94;

N, 6.88. Found: C, 65.12; H, 6.49; N, 6.93.

N-{3-[(4-Fluorobenzyl)oxy]phenyl}-1-(pyridin-4-ylmethyl)-

piperidine-4-carboxamide (9). To a solution of 200 mg (0.55

mmol) of 5·HCl in 10 mL of DMF were added 149 mg (1.08 mmol)

of K₂CO₃and 91 mg (0.36 mmol) of 4-bromomethylpyridine

hydrobromide, and the reaction mixture was stirred at room

temperature for 48 h and then poured into ice−water. The resulting

precipitate was collected, washed with Et₂O, and then ﬁltered to

provide the title compound as a pale brown solid (147 mg, 97% yield),

mp 155−158 °C. IR (KBr) 3225, 1648, 1439, 1224, 1154, 1010 cm⁻¹.

¹H NMR (300 MHz, DMSO-d₆) δ 9.82 (s, 1H), 8.48 (d, J = 5.8 Hz,

2H), 7.49−7.46 (m, 2H), 7.44 (s, 1H), 7.31 (d, J = 5.8 Hz, 2H), 7.22−

7.08 (m, 4H), 6.65 (d, J = 9.0 Hz, 1H), 5.01 (s, 2H), 3.48 (s, 2H),

2.83−2.80 (m, 2H), 2.33−2.25 (m, 1H), 1.99 (d, J = 9.0 Hz, 1H), 1.95

(d, J = 8.5 Hz, 1H), 1.74−1.62 (m, 4H). ESIMS m/z 420 (MH⁺).

Anal. Calcd for C₂₅H₂₆N₃O₂F: C, 71.58; H, 6.25; N, 10.02. Found: C,

71.25; H, 6.49; N, 9.93.

General Procedure for Synthesis of Substituted N-

[(Benzyloxy)phenyl]-1-arylpiperidine-carboxamide Deriva-

tives (11−13, 18−29, and 31−34). Yields and spectroscopic data

of compounds that have been prepared following the same route are

reported in Supporting Information. The synthesis of N-{3-[(4-

ﬂuorobenzyl)oxy]phenyl}-1-pyridin-2-ylpiperidine-4-carboxamide 13a

is reported here as an example.

To a solution of 500 mg (2.42 mmol) of 1-(pyridin-2-yl)piperidine-

4-carboxylic acid 10c in 10 mL of dry DMF were added 777 mg (2.42

mmol) of TBTU, 1.686 mL (9.68 mmol) of DIPEA, and, after 30 min,

475 mg (2.91 mmol) of 3-[(4-ﬂuorobenzyl)oxy]aniline.¹⁹The reaction

mixture was stirred at room temperature for 72 h and then poured on

ice. The resulting precipitate was ﬁltered, collected, and washed with

Et₂O to yield the title compound (when necessary, the crude products

were recrystallized from acetone and MeOH) as a pale brown solid,

100 mg (45% yield), mp 144−146 °C. IR (KBr) 3232, 1649, 1596,

1489, 1439, 1226, 1207, 1154, 1011 cm⁻¹. ¹H NMR (300 MHz,

DMSO-d₆) δ 9.89 (s, 1H), 8.08 (d, J = 3 Hz, 1H), 7.51−7.39 (m, 4H),

7.22−7.12 (m, 4H), 6.82 (d, J = 8.5 Hz, 1H), 6.65 (d, J = 7.0 Hz, 1H),

6.60−6.56 (m, 1H), 5.01 (s, 2H), 4.33 (d, J = 13 Hz, 2H), 2.83 (d, J =

12 Hz, 1H), 2.79 (d, J = 12 Hz, 1H), 2.61−2.53 (m, 1H), 1.82−1.78

(m, 2H), 1.63−1.52 (m, 2H). ESIMS m/z 406 (MH⁺). Anal. Calcd for

C₂₄H₂₄N₃O₂F: C, 71.09; H, 5.97; N, 10.36. Found: C, 71.12; H, 6.29;

N, 10.25.

1-(4-Aminophenyl)-N-{3-[(4-ﬂuorobenzyl)oxy]phenyl}-

piperidine-4-carboxamide (15). To a solution of 100 mg (0.22

mmol) of 14 (its synthesis is described in the Supporting Information)

in 20 mL of MeOH were added 0.1 mL of hydrazine hydrate and 0.1

mL of activated Raney Ni suspension. The reaction mixture was

reﬂuxed for 0.5 h until disappearance (as monitored by thin-layer

chromatography, TLC) of the starting material and then cooled and

ﬁltered through a Celite pad. The ﬁltrate was concentrated under

reduced pressure to provide 60 mg (65% yield) of 15 as a brown solid,

mp 159−161 °C. IR (KBr) 3435, 1671, 1261, 1153, 1022, 800 cm⁻¹.

¹H NMR (300 MHz, acetone-d₆) δ 9.12 (br s, 1H), 7.60 (s, 1H), 7.54

room temperature. After cooling to 0 °C, the reaction was quenched

by addition of 10 mL of saturated Na₂SO₄solution, ﬁltered, and

washed with Et₂O. The organic phase was dried over Na₂SO₄, ﬁltered,

and concentrated under reduced pressure to yield 1.57 g (84% yield)

of 3-(3-ﬂuorobenzyloxy)-N-methylbenzenamine 17 as an oil. IR (ﬁlm)

3414, 1618, 1592, 1514, 1497, 1448, 1261, 1194, 1166, 1037, 932, 868,

1

828, 774, 686 cm⁻¹. H NMR (300 MHz, CDCl₃) δ 7.37−7.27 (m,

1H), 7.25−7.14 (m, 2H), 7.09 (t, J = 8.0 Hz, 1H), 7.00 (t, J = 8.2 Hz,

1H), 6.34−6.23 (m, 3H), 5.03 (s, 2H), 3.77 (br s, 1H), 2.82 (s, 3H).

N-{3-[(3-Fluorophenoxy)methyl]phenyl}-N-[(1-pyridin-4-yl-

piperidin-4-yl)methyl]amine (29). To a 0 °C cooled solution of

1.566 g (3.86 mmol) of compound 13b in 20 mL of THF was added

27 mL of a solution of 1 M borane−THF complex, and the reaction

mixture was stirred at room temperature for 24 h. After quenching by

addition of 6 N HCl (20 mL), water (20 mL), and MeOH (40 mL),

stirring was prolonged for 24 h. The resulting solution was then

concentrated under reduced pressure and the aqueous solution was

basiﬁed to pH 12 with 4 N NaOH and extracted three times with

EtOAc (20 mL). The combined organic layers were dried, ﬁltered, and

concentrated under reduced pressure to provide 1.42 g of 29 (94%

yield) as a brown solid, mp 159−161 °C. IR (KBr) 3246, 1616, 1600,

1543, 1513, 1194, 1054, 990, 804, 771, 752, 699 cm⁻¹. ¹H NMR (300

MHz, DMSO-d₆) δ 8.09 (d, J = 5.0 Hz, 2H), 7.44−7.37 (m, 1H),

7.25−7.20 (m, 2H), 7.12 (t, J = 9.0 Hz, 1H), 6.92 (t, J = 8.5 Hz, 1H),

6.78 (d, J = 5.0 Hz, 2H), 6.18−6.12 (m, 3H), 5.69 (t, J = 6.0 Hz, 1H),

5.02 (s, 2H), 3.91 (d, J = 13 Hz, 2H), 2.87 (t, J = 6.0 Hz, 2H), 2.78 (d,

J = 13 Hz, 1H), 2.74 (m, J = 13 Hz, 1H), 1.78 (d, J = 11 Hz, 2H),

1.44−1.39 (m, 1H), 1.21−1.09 (m, 2H). ESIMS m/z 420 (MH⁺).

Anal. Calcd for C₂₄H₂₆N₃OF: C, 73.63; H, 6.69; N, 10.73. Found: C,

73.51; H, 6.75; N, 10.93.

N-[3-(3-Fluorobenzyloxy)phenyl]-N-methylpiperidine-4-car-

boxamide (30). To a solution of 1.71 g (7.48 mmol) of 1-(tert-

butoxycarbonyl)piperidine-4-carboxylic acid¹⁹in 25 mL of THF were

added 0.92 g (6.80 mmol) of 1-hydroxybenzotriazole hydrate (HOBt),

1.403 g (6.80 mmol) of N,N′-dicyclohexylcarbodiimide (DCC), and,

after 30 min, 1.57 g (6.80 mmol) of 3-(3-ﬂuorobenzyloxy)-N-

methylbenzenamine 17, dissolved in 5 mL of dry THF. The reaction

mixture was stirred for 48 h at room temperature, and then DCU was

ﬁltered oﬀ and the ﬁltrate was concentrated to dryness. The oil residue

was dissolved in 100 mL of EtOAc, and the organic phase was washed

sequentially with 3 × 20 mL of saturated NaHCO₃, 1 N HCl, and

brine, dried, ﬁltered, and concentrated under reduced pressure. The

obtained oil residue was chromatographed on silica gel (mobile phase

EtOAc/40−60 petroleum ether, 3:7 v/v), to provide the N-Boc-

protected derivative. After removal of the Boc protecting group, by

bubbling HCl gas through a cooled CHCl₃solution, the obtained HCl

salt was poured into water and the solution was brought to pH 12 by

adding 4 N NaOH. The aqueous basic suspension was extracted three

times with EtOAc, and the combined organic layers were dried,

ﬁltered, and concentrated to provide 1.22 g (52% yield) of 30, as a

brown oil. ¹H NMR (300 MHz, CDCl₃) δ 7.39−7.27 (m, 2H), 7.20−

7.13 (m, 2H), 7.05−6.91 (m, 2H), 6.80−6.74 (m, 2H), 5.09 (s, 2H),

3.22 (s, 3H), 2.99 (d, J = 13 Hz, 2H), 2.36−2.25 (m, 3H), 1.73−1.62

(m, 3H), 1.52−1.48 (m, 2H).

N-{3-[(3-Fluorophenoxy)methyl]phenyl}-N-methyl-1-pyri-

din-4-ylpiperidine-4-carboxamide (31). A mixture, containing 366

mg (1.07 mmol) of 30, 161 mg (1.07 mmol) of 4-chloropyridine

hydrochloride,and 0.45 mL (3.21 mmol) of TEA in 4 mL of absolute

EtOH, was stirred for 2 days at 140 °C in a sealed tube. After cooling

to room temperature, the mixture was concentrated under reduced

pressure and the oil residue was dissolved in 30 mL of EtOAc and

washed twice with 20 mL of 2 N NaOH. The organic phase was dried,

ﬁltered, and concentrated under reduced pressure to provide 373 mg

of 31 (83% yield), as a dark brown oil. IR (liquid) 1652, 1592, 1226,

1028, 989, 873, 784 cm⁻¹. ¹H NMR (300 MHz, DMSO-d₆) δ 8.09 (d,

J = 6.0 Hz, 2H), 7.47−7.25 (m, 4H), 7.18−7.11 (m, 1H), 7.04−6.85

(m, 3H), 6.72 (d, J = 6.0 Hz, 2H), 5.17 (s, 2H), 3.80 (d, J = 13 Hz,

2H), 3.31 (s, 3H), 2.73−2.78 (m, 1H), 2.42−2.54 (m, 1H), 1.96−2.06

(m, 1H), 1.64−1.34 (m, 4H). ESIMS m/z 420 (MH⁺). Anal. Calcd for

(dd, J₁= 5.5 Hz, J₂= 3.0 Hz, 2H), 7.20−7.13 (m, 4H), 6.76 (d, J = 9.0

Hz, 2H), 6.72−6.68 (m, 1H), 6.59 (d, J = 9.0 Hz, 2H), 5.08 (s, 2H),

3.48−3.43 (m, 2H), 2.81 (br s, 2H), 2.61−2.51 (m, 2H), 2.47−2.36

(m, 1H), 1.95−1.90 (m, 4H). ESIMS m/z 420 (MH⁺). Anal. Calcd for

C₂₅H₂₆N₃O₂F × H₂O: C, 68.63; H, 6.45; N, 9.60. Found: C, 68.80; H,

6.49; N, 9.73.

3-(3-Fluorobenzyloxy)-N-methylbenzenamine (17). A solu-

tion containing 1.29 g (5.96 mmol) of 3-[(4-ﬂuorobenzyl)oxy]aniline

in 10 mL of ethyl formate was reﬂuxed for 20 h. The mixture was

cooled to room temperature and concentrated under reduced pressure

to provide 1.43 g (98% yield) of N-[3-(3-ﬂuorobenzyloxy)phenyl]-

formamide 16. Then 1.88 g (7.67 mmol) of the crude product 16 was

dissolved in 10 mL of dry THF and added dropwise to a cooled

suspension of 0.436 g (11.50 mmol) of LiAlH₄in 30 mL of dry THF.

The reaction mixture was stirred for 1 h at 0 °C and then for 20 h at

L

dx.doi.org/10.1021/jm401169a | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

C₂₅H₂₆N₃O₂F × H₂O: C, 68.63; H, 6.45; N, 9.60. Found: C, 68.72; H,

6.59; N, 9.83.

permeability coeﬃcients P_e(centimeters per second), in a 96-well

microtiter ﬁlter plates, on polycarbonate ﬁlter of 3 μm pore size, 10

μm thickness, and 5−20% porosity (Millipore AG, Volketswil,

Switzerland), according to the procedure of Wohnsland and Faller,³⁰

and testosterone (log P_e= −3.4) was used as the reference

compound.⁴⁹

X-ray Crystal Structure Analysis of 13b. Crystals of 13b

suitable for X-ray structure determination were obtained by slowly

adding dichloromethane to a methanolic solution of the compound.

The crystal structure was determined from X-ray single-crystal

diﬀraction data. The crystal of 13b was mounted on a glass ﬁber

and diﬀraction measurements were carried out on a Bruker-Nonius

Kappa charge-coupled device (CCD) diﬀractometer equipped with a

CCD area detector, a Mo ﬁne focus X-ray tube (λ = 0.710 73 Å), and a

0.35 mm capillary optical collimator to improve radiation intensity and

reduce beam divergence. Crystallographic parameters and data

collection information are reported in Supporting Information (Tables

S1 and S2).

Each compound, dissolved in DMSO (concentration range 10⁻²−

10⁻³M, according to solubility and UV detection limits), was tested at

least in triplicate at pH 6.8 (and pH 8.0 as well) in iso-pH conditions

(i.e., the same pH in donor and acceptor compartments). Each well

was coated with 15 μL of hexadecane (5% in hexane solution) for at

least 20 min to completely evaporate the hexane, whereas the Teﬂon

acceptor plate was hydrated with 280 μL of pH 6.8 phosphate buﬀer

containing 5% DMSO. The donor plate was placed upon and ﬁlled

with 280 μL of pH 6.8 phosphate buﬀer containing 5% DMSO test

compound solution, and the resulting chamber was incubated at room

temperature under constant shaking (150 rpm) for 5 h, when it was

carefully disassembled and both acceptor and donor compartment

solutions were transferred into UV-quartz plates and absorptions were

measured (PowerWave, Bio-Tek Instruments, Inc.) at the compound’s

λ_max. Hexadecane membrane stability was tested at the end of

incubation time by electrical resistance measurements, by use of an

electrometer system for PAMPA assays (EVOMX and MULTI96,

World Precision Instruments, Sarasota, FL). The eﬀective permeability

P_ewas determined from the ratio C_A(t)/C_D(t₀) between UV

absorbances in the acceptor compartment at time t and in the

donor well at time t₀. The permeability equation used includes a term

related to membrane compartment:⁵⁰

The structure was solved by direct methods with SIR2011⁴⁵and

reﬁned by full matrix least-squares on F²with SHELXL-97.⁴⁶Friedel

pairs were merged before the ﬁnal reﬁnement. The positions of the H

atoms bonded to N atoms were obtained from the diﬀerence Fourier

map and were reﬁned. The remaining H atoms were placed in

geometrically calculated positions and were reﬁned by use of a riding

3

model, with C−H = 0.96 and 0.93 Å for C_spH and C_arH, respectively.

The thermal factors of non-H atoms were reﬁned anisotropically, while

those of H atoms were considered isotropic and linked to the

equivalent isotropic thermal factors of the carrier non-H atom: U_iso

(H) = 1.2U_eq(non-H). The reﬁned structure of 13b, with ellipsoids

representing the reﬁned thermal factors, generated with ORTEP-3,⁴⁷is

shown in Figure 4. The reﬁnement parameters (Table S3) and details

on crystal packing are reported in Supporting Information.

Potentiometric Determination of Ionization Constants.

Experimental ionization constants (pK_a) were determined by

potentiometric titration on a Sirius GLpKa instrument, and data

were treated by ReﬁnementPro software (Sirius Analitical Ltd., Forest

Row, East Sussex, U.K.). An appropriate amount of each compound

was dissolved in 20 mL of ionic strength adjusted water (0.15 M KCl

solution) to achieve a ﬁnal sample concentration around 5 × 10⁻⁴M.

All the titrations were recorded in the pH range between 1.8 and 12.2

at 25 0.5 °C, with standardized 0.5 M HCl and 0.5 M carbonate-free

KOH as titrating agents. Experiments were carried out under a slow

nitrogen ﬂow to avoid CO₂absorption under high-pH conditions.⁴⁸

All the measurements were carried out in triplicate, and precise pK_a

values were calculated from the Bjerrum diﬀerential curve as the pH

value required to have 50% of dissociated form of test compound. For

compounds with low aqueous solubility, methanol, ranging from 20%

to 60% (v/v), was added as the cosolvent and the aqueous pK_awas

then extrapolated by use of the Yasuda−Shedlovsky equation.

⎡

⎤

⎥

⎛

⎜

⎞

⎟

⎠

⎛

⎜

⎝

⎞

2.303V_D

V_A

V_A+ V_DC_A(t)

⎢

P = −

log 1 −

⎟

e

⎢

⎣

⎥

⎦

A(t − τ ) V + V

V (1 − R) C (t )

⎝

⎠

lag

A

D

D 0

where A is the accessible ﬁlter area (0.24 cm²) multiplied by nominal

porosity of 13% according to manufacturer, t is the incubation time

(seconds), V_Aand V_Dare the volumes (cubic centimeters) in the

acceptor and donor wells, R is the retention factor, deﬁned as the mole

fraction of compound lost in the membrane and in microplates (i.e.,

ﬁlters and plate materials), and τ_lagis the steady-state time (seconds),

which is the time needed for the permeant’s concentration gradient to

become stabilized; because in PAMPA experiments τ_lagfor saturation

of the membrane is very shorter than the total permeation time, it is

usually neglected.⁵¹The retention factor R is deﬁned as

C_D(t)

C_D(t₀)

V_AC_A(t)

V_DC_D(t₀)

R = 1 −

−

Determination of Lipophilicity Parameters by Hydrophilic

Interaction Liquid Chromatography. Retention measurements

were performed at 23 °C and a ﬂow rate of 1.0 mL·min⁻¹by using as

the stationary phase a ZIC-pHILIC column (sulfoalkylbetaine phase

on a polymeric support, 10 cm × 4.6 mm, 5 μm) from SeQuant

Since permeation can be attributed to the neutral species alone,

particularly for compounds with un-ionized fraction f_ui> 0.1, the

eﬀective permeability coeﬃcient P_eis divided by the un-ionized

fraction f_ui, calculated for compounds with a single acidic or basic

group according to the dissociation constant (pK_a) and the pH in each

compartment, as follows:

(Umea, Sweden). All the experiments were performed on a Merck

̊

Hitachi EliteLaChrom liquid chromatograph (Merck, Darmstadt,

Germany, and Hitachi Instruments, Inc., San Jose, CA) equipped

with L-2200 auto sampler, L-2130 pump, L-7614 degasser, and L-2400

UV detector, operating at 254 nm for all compounds. The

chromatographic system was controlled by a EzChrom Elite System

Manager software version 3.1.7 (Merck Hitachi). Phoebus software 1.0

(Sedere, Centre Analyze, Orleans, France) was used to prepare buﬀer

(triﬂuoroacetic acid/ammonium) at pH 2 and ionic strength (I) equal

to 100 mM.

1

f_ui

=

1 + 10^g

where g = (pH − pK_a) for acids and g = (pK_a− pH) for bases.

Molecular Modeling. All calculations were performed with

programs included in the Schrodinger Suite 2011.⁵⁶The X-ray crystal

structure of bovine thrombin in complex with the 4-aminopyridine-

based inhibitor BM14.1248 (PDB code 1UVT)²¹was retrieved and

used as the target in docking calculations. After being prepared with

the standard options of the Protein Preparation Wizard protocol, the

docking grid was calculated by placing the coordinates of the center of

mass of the cocrystallized ligand as the center of a cubic box, having a

According to Bard et al.,⁴³the diﬀerence parameter (Δlog k₀₋₉₅

)

between two isocratic log k values of the basic compounds in their

cationic formsnamely, log k₀, which is the log k obtained in 100%

aqueous buﬀer (pH 2) mobile phase, and log k₉₅, which is the isocratic

log k measured in 95% (v/v) acetonitrile−aqueous buﬀer (pH 2)

mobile phaseallows the partition coeﬃcient of their neutral form

(log P^N) to be calculated for basic compounds through the following

validated correlation equation:⁴³log P^N= 1.07 Δlog k₀₋₉₅+ 0.61.

Permeability Measurements. Parallel artiﬁcial membrane

permeability assay (PAMPA) was used to determine eﬀective

side length of 20 Å. All the possible protonation states at pH 7.0

2

were assigned to the inhibitor molecules by use of the LigPrep module

version 2.5 (default options). All docking calculations were performed

in the extra precision (XP) mode of Glide version 5.7,²⁰and either

GlideScore scoring function or the composite Emodel score were used

to rank the poses of each ligand. The Emodel scoring function was

M

dx.doi.org/10.1021/jm401169a | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

ﬁnally preferred to select the top-ranked poses to be visualized, due to

its better correlation with the experimental aﬃnity data. The surface

area descriptors (SASA, PSA, etc.) of each ligand in the top-ranked

pose were calculated by use of the QikProp 3.4 module available in

Schrodinger Suite 2011.

determined, and the concentrations of the inhibitors required to

diminish the control velocity by 50% (IC₅₀) were calculated by

nonlinear (sigmoidal) regression. At least three independent IC₅₀

values were determined to calculate inhibition constants (K_i) by use of

the Cheng−Prusoﬀ equation.⁵⁵

In Vitro Plasma Clotting Time Assays. Clotting time of the test

compounds, namely activated partial thromboplastin time (aPTT),

was measured on a coagulometer (Behnk Electronic, Norderstedt,

Germany) and compared with those from human control plasma.

Pooled lyophilized human plasma (100 μL; Futura Systems, Formello,

Rome, Italy) was incubated for 3 min at 37 °C with test compound

solution (10 μL) or solvent (DMSO maximum of 1%), followed by

the addition of aPTT reagent (100 μL) and 0.025 M CaCl₂(100 μL;

Futura Systems) to trigger clot formation. Each measurement was

performed in triplicate, and the concentration of test compound that

caused 2-fold prolongation of the basal clotting times (aPTT₂) was

calculated from each individual concentration−response curve.

Ex Vivo Anticoagulant Assays in Mice. Ex vivo anticoagulant

assays (aPTT₂) were performed in agreement with the Italian Law on

Animal Care 116/1992 and EEC/609/86, and all eﬀorts were made to

minimize the number of animals used. CD1 male mice (Harlan

Laboratories) weighing 25−35 g were used. The mice were housed

under standard laboratory conditions with open access to standard

food and tap water. After overnight fasting, mice were anesthetized by

intraperitoneal injection with urethane (1.2 g·kg⁻¹), and test

compound 13b, suspended in a 0.5% methylcellulose solution, was

administered orally at concentration 100 mg·kg⁻¹by use of a gastric

tube (FTP-20−30, InstechLabs). At 1 and 2 h after test compound

administration, blood (0.2 mL) was collected from the inferior vena

cava into syringes containing sodium citrate (3.8% v/v). Platelet-poor

plasma was then prepared by centrifugation for 20 min at 1200g to

measure aPTT₂. Data (mean SEM) were compared with those of

the vehicle group. One-way analysis of variance (ANOVA) and

multiple comparison tests (Dunnett’s test) were performed by use of a

standard statistical package. P values less than 0.05 were considered

statistically signiﬁcant.

Molecular Dynamics Simulations. Complexes between the

protein and the ligands 13b, 33b, and 34b, in the top-ranked poses

as identiﬁed according the Emodel scoring function, were stored and

used for MD simulations, performed with the Desmond package

version 3.0.⁵²The protein−ligand complexes were ﬁrst solvated with

TIP3P water molecules, 0.15 M NaCl, and neutralized with proper

counterions by use of the System Builder module of Desmond. A

triclinic box, with each side at a minimum distance of 10 Å from any

atom of the complex, was used. For the ligands the OPLS 2005 force

ﬁeld in a canonical ensemble (NVT) was used; all the remaining

default settings have been left. The dynamic protocol was set in

subsequent steps: (i) a relaxing minimization, followed by (ii) a

progressive increase of temperature to 300 K in 140 ps, and ﬁnally (iii)

a production run of 2 ns. Frames of the simulation were recorded

every 5 ps. The 2 ns trajectories were analyzed by use of the simulation

event analysis tool, as implemented in the Desmond package, whereas

VMD⁵³was used for visualization purposes.

Inhibition Assays for Thrombin, Factor Xa, and Other Serine

Proteases. The test compounds were assayed in vitro for their

inhibitory activity toward thrombin, fXa, and other serine proteases, by

determining the hydrolysis rates of the synthetic chromogenic

substrates monitored at 405 nm. Enzymes and substrates were used

as follows (ﬁnal concentrations): 0.41 unit·mL⁻¹bovine thrombin

from Sigma−Aldrich (Milan, Italy) and 50 μM S-2238 (^D-Phe-Pip-

Arg-p-NA) from Chromogenix AB Instrumentation Laboratories; 0.5

nM puriﬁed human thrombin and 30 μM S-2238; 4 nM human factor

Xa and 0.04 μM S-2765 (Z-^D-Arg-Gly-Arg-p-NA) from Chromogenix

AB-Instrumentation Laboratories (Milan, Italy); 30 nM human

recombinant factor VIIa (reconstituted with 0.2 mL of deionized

water) and 1 mM Spectrozyme fVIIa from American Diagnostics

(Stamford, CT); 0.4 μg·mL⁻¹bovine α-chymotrypsin and 185 μM N-

succinyl-Ala-Ala-Pro-Phe-p-NA from Sigma−Aldrich; 2 nM bovine

pancreas trypsin from Calbiochem (Darmstadt, Germany) and 4 mM

S-2238; 5 nM human neutrophil leukocyte elastase and 660 μM N-

(methoxysuccinyl)-Ala-Ala-Pro-Val-p-NA from Sigma−Aldrich. Puri-

ﬁed human thrombin was obtained as previously reported.⁵⁴Wild-type

prothrombin was activated by Taipan snake venom (in 50 mM Tris-

HCl, 150 mM NaCl, and 2 mM CaCl₂), and the generated thrombin

was puriﬁed by cation-exchange HPLC. Sodium dodecyl sulfate−

polyacrylamide gel electrophoresisE of pooled chromatographic peaks

was carried out on 4−20% gradient gels under both reducing (5% β-

mercaptoethanol) and nonreducing conditions. The thrombin

concentration was measured spectrophotometrically at 280 nm, by

use of an extinction coeﬃcient (0.1%) equal to 1.83. The active-site

titration was carried out spectrophotometrically, and the puriﬁed

enzyme was immediately aliquoted and frozen at −80 °C until use. All

buﬀer salts were purchased from Sigma−Aldrich. Enzyme solutions

were incubated with DMSO solutions of the test inhibitors (DMSO

did not exceed 1%) in various concentrations (0.1−100 nM or 0.1−

100 μM), before the respective chromogenic substrates were added to

initiate the enzyme kinetics. Kinetic studies were performed at pH 8.0

(10 mM Tris buﬀer, 150 mM NaCl, and 0.1% PEG6000) for bovine

and human fIIa and fXa and at pH 7.5 (50 mM Tris buﬀer and 50 mM

CaCl₂) for α-chymotrypsin, recombinant human fVIIa, pancreas

bovine trypsin, and human leukocyte elastase. Kinetics for human

thrombin, recombinant human fVIIa, pancreas bovine trypsin, and

human leukocyte elastase (200 mM Tris buﬀer, 150 mM NaCl, and

0.1% PEG6000, pH 7.5), were performed by monitoring the

absorbance increase at 405 nm and 25 °C with a microplate Bio-

Rad spectrophotometer. The enzyme solution (50 μL) and buﬀer (50

μL) were mixed with 2 μL of DMSO solution containing the test

compound or DMSO alone as the control and incubated (15 min for

fVIIa, 30 min for trypsin, and 10 min for elastase). Reactions were

initiated by adding 100 μL of substrate solutions, and the increase in

absorbance was monitored for 5 min. Initial velocities were

ASSOCIATED CONTENT

* Supporting Information

■

S

Analytical data (IR, NMR, MS) and elemental analyses (C, H,

N) for all newly characterized intermediates and test

compounds; additional text, three tables, and two ﬁgures with

X-ray crystal data for 13b. This material is available free of

charge via the Internet at http://pubs.acs.org. The crystal

structure of compound 13b has been deposited in the

Cambridge Crystallographic Data Centre (CCDC); http://

www.ccdc.cam.ac.uk (deposition code 959986).

AUTHOR INFORMATION

Corresponding Author

*Phone +39-080-5442781; fax +39-080-5442230; e-mail

cosimodamiano.altomare@uniba.it.

■

Present Addresses

^⊥(F.F.) Farmalabor s.r.l., Via Pozzillo, II Trav. Z.I., 76012

Canosa di Puglia, Italy.

^▽(G.L.) Drug Design & Discovery, Aptuit, via A. Fleming,

37135 Verona, Italy.

Notes

The authors declare no competing ﬁnancial interest.

ACKNOWLEDGMENTS

■

We are grateful to Professor Raimondo De Cristofaro (Catholic

University School of Medicine, Rome, Italy), for his helpful

collaboration in biological data interpretation, as well as for

assistance in the inhibition assays with human serine proteases,

and we thank Dr. Brunella Maria Aresta (Istituto di

N

dx.doi.org/10.1021/jm401169a | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

heparin-antithrombin but susceptible to inactivation by antithrombin-

independent inhibitors. Circulation 1998, 97, 544−552.

(13) Dhillon, S. Argatroban: A review of its use in the management of

heparin-induced thrombocytopenia. Am. J. Cardiovasc. Drugs 2009, 9,

261−282.

Cristallograﬁa, Consiglio Nazionale delle Ricerche, Bari, Italy)

for her contribution in crystallographic analysis. Financial

support by the Italian Ministry for Education Universities and

Research (MIUR, Rome, Italy; PRIN 2007, Grant

2007T9HTFB_003) is acknowledged.

(14) Sorbera, L. A.; Bayes, M. C. J.; Silvestre, J. Melagatran and

ximelagatran. Drugs Future 2001, 26, 1155−1170.

ABBREVIATIONS USED

(15) AstraZeneca decides to withdraw Exanta. http://www.

astrazeneca.com/media/latest-press-releases/2006/5217?itemId=

3891692, 2006.

■

αCT, α-chymotrypsin; ADMET, adsorption, distribution,

metabolism, excretion, and toxicity; ADP, adenosine 5′-

diphosphate; aPTT, activated partial thromboplastin time;

ATIII, antithrombin III; DIAD, diisopropyilazodicarboxylate;

DIPEA, N,N-diisopropylethylamine; DMSO, dimethyl sulf-

oxide; DTI, direct thrombin inhibitor; DVT, deep venous

thrombosis; ESIMS, electrospray ionization mass spectrometry;

fIIa, thrombin; fVIIa, activated factor VII; fXa, activated factor

X; HDM, hexadecane membrane; HILIC, hydrophilic inter-

action liquid chromatography; HIT, heparin-induced thrombo-

cytopenia; LE, human leukocyte elastase; LMWHs, low

molecular weight heparins; PAMPA, parallel artiﬁcial mem-

brane permeability assay; P_e, eﬀective permeability; PE,

pulmonary embolism; PSA, polar surface area; SASA, solvent-

accessible surface area; TBTU, O-(benzotriazol-1-yl)-

N,N,N′,N′-tetramethyluronium tetraﬂuoroborate; Try, trypsin;

UFHs, unfractionated heparins; VTE, venous thromboembo-

lism

(16) Hauel, N. H.; Nar, H.; Priepke, H.; Ries, U.; Stassen, J.-M.;

Wienen, W. Structure-based design of novel potent nonpeptide

thrombin inhibitors. J. Med. Chem. 2002, 45, 1757−1766.

(17) (a) Stangier, J.; Rathgen, K.; Stahle, H.; Gansser, D.; Roth, W.

The pharmacokinetics, pharmacodynamics and tolerability of dabiga-

tran etexilate, a new oral direct thrombin inhibitor, in healthy male

subjects. Br. J. Clin. Pharmacol. 2007, 64, 292−303. (b) Eriksson, B. I.;

Dahl, O. E.; Rosencher, N.; Kurth, A. A.; van Dijk, C. N.; Frostick, S.

P.; Prins, M. H.; Hettiarachchi, R.; Hantel, S.; Schnee, J.; Buller, H. R.

̈

Dabigatran etexilate versus enoxaparin for prevention of venous

thromboembolism after total hip replacement: a randomised, double-

blind, non-inferiority trial. Lancet 2007, 370, 949−956. (c) Larsen, T.

́

B.; Rasmussen, L. H.; Skjøth, F.; Due, K. M.; Callreus, T.; Rosenzweig,

M.; Lip, G. Y. Efficacy and safety of dabigatran etexilate and warfarin in

″real-world″ patients with atrial fibrillation: a prospective nationwide

cohort study. J. Am. Coll. Cardiol. 2013, 61, 2264−2273.

(18) (a) de Candia, M.; Summo, L.; Carrieri, A.; Altomare, C.;

Nardecchia, A.; Cellamare, S.; Carotti, A. Investigation of platelet

aggregation inhibitory activity by phenyl amides and esters of

piperidinecarboxylic acids. Bioorg. Med. Chem. 2003, 11, 1439−1450.

(b) De Marco, A.; de Candia, M.; Carotti, A.; Cellamare, S.; De

Candia, E.; Altomare, C. Lipophilicity-related inhibition of blood

platelet aggregation by nipecotic acid anilides. Eur. J. Pharm. Sci. 2004,

22, 153−164.

REFERENCES

■

(1) (a) Macmanaman, N. Triggers, targets and treatment for

thrombosis. Nature 2008, 451, 914−918. (b) Beckman, M. G.;

Hooper, W. C.; Critchley, S. E.; Ortel, T. L. Venous thromboemb-

olism: a public health concern. Am. J. Prev. Med. 2010, 38 (4S), S495−

S501. (c) Chaudhry, F. S.; Schneck, M. J.; Morales-Vidal, S.; Javaid, F.;

Ruland, S. Prevention of venous thromboembolism in patients with

hemorrhagic stroke. Top. Stroke Rehabil. 2013, 20, 108−115.

(2) Alban, S. Pharmacological strategies for inhibition of thrombin

activity. Curr. Pharm. Des. 2008, 14, 1152−1175.

(19) de Candia, M.; Liantonio, F.; Carotti, A.; De Cristofaro, R.;

Altomare, C. Fluorinated benzyloxyphenyl piperidine-4-carboxamides

with dual function against thrombosis: Inhibitors of factor Xa and

platelet aggregation. J. Med. Chem. 2009, 52, 1018−1028.

(20) (a) Friesner, R. A.; Banks, J. L.; Murphy, R. B.; Halgren, T. A.;

Klicic, J. J.; Mainz, D. T.; Repasky, M. P.; Knoll, E. H.; Shelley, M.;

Perry, J. K.; Shaw, D. E.; Francis, P.; Shenkin, P. S. Glide: A new

approach for rapid, accurate docking and scoring. 1. Method and

assessment of docking accuracy. J. Med. Chem. 2004, 47, 1739−1749.

(b) Halgren, T. A.; Murphy, R. B.; Friesner, R. A.; Beard, H. S.; Frye,

L. L.; Pollard, W. T.; Banks, J. L. Glide: A new approach for rapid,

accurate docking and scoring. 2. Enrichment factors in database

screening. J. Med. Chem. 2004, 47, 1750−1759. (c) Friesner, R. A.;

Murphy, R. B.; Repasky, M. P.; Frye, L. L.; Greenwood, J. R.; Halgren,

T. A.; Sanschagrin, P. C.; Mainz, D. T. Extra precision Glide: Docking

and scoring incorporating a model of hydrophobic enclosure for

protein-ligand complexes. J. Med. Chem. 2006, 49, 6177−6196.

(21) Engh, R. A.; Brandstetter, H.; Sucher, G.; Eichinger, A.;

Baumann, U.; Bode, W.; Huber, R.; Poll, T.; Rudolph, R.; Von der

Saal, W. Enzyme flexibility, solvent and weak interactions characterize

thrombin-ligand interactions: implications for drug design. Structure

1996, 4, 1353−1362.

(22) Faller, B. Artificial membrane assays to assess permeability. Curr.

Drug Metab. 2008, 9, 886−892.

(23) Ottaviani, G.; Martel, S.; Escarala, C.; Nicolle, E.; Carrupt, P.-A.

The PAMPA technique as a HTS tool for partition coefficients

determination in different solvent/water system. Eur. J. Pharm. Sci.

2008, 35, 68−75.

(24) Batt, A. R.; Heeney, C.; Marguerite, S. Diazepinyl(piperidin-4-

yl)methanone derivatives as vasopressin receptor V1a antagonists and

their preparation and use for the treatment of diseases. World Patent

WO2010097576 (A8), 2011.

(25) Johnson, P. S.; Ryckmans, T.; Bryans, J.; Beal, D. M.; Dacka, K.

N.; Feeder, N.; Harrison, A.; Lewis, M.; Mason, H. J.; Mills, J.;

Newman, J.; Pasquinet, C.; Rawson, D. J.; Roberts, L. R.; Russell, R.;

(3) Haas, S. New oral Xa and IIa inhibitors: updates on clinical trial

results. J. Thromb. Thrombolysis 2008, 25, 52−60.

(4) Hirsch, J.; Dalen, J. E.; Anderson, D. R.; Poller, L.; Bussey, H.;

Ansell, J.; Deykin, D. Oral anticoagulants: mechanism of action,

clinical effectiveness, and optimal therapeutic range. Chest 2001, 119,

8S−21S.

(5) Hirsch, J.; Raschke, R. Heparin and low-molecular-weight

heparin: the seventh ACCP conference on antithrombotic and

thrombolytic therapy. Chest 2004, 126, 188S−203S.

(6) (a) Schwarz, R. P. The preclinical and clinical pharmacology of

Novastan (Argatroban). In New Anticoagulants for the Cardiovascular

Patient; Pifarre, R., Ed.; Hanley and Belfus Inc.: Philadelphia, PA,

1997; pp 231−249. (b) Schwienhorst, A. Direct thrombin inhibitors: A

survey of recent developments. Cell. Mol. Life Sci. 2006, 63, 2773−

2791.

(7) Linusson, A.; Gottfries, J.; Olsson, T.; Ornskov, E.; Folestad, S.;

̀

Norden, B.; Wold, S. Statistical molecular design, parallel synthesis,

and biological evaluation of a library of thrombin inhibitors. J. Med.

Chem. 2001, 44, 3424−3439.

(8) Coughlin, S. R.; Vu, T.-K. H.; Hung, D. T.; Wheaton, V. I.

Characterization of a functional thrombin receptor. J. Clin. Invest.

1992, 89, 351−355.

́

(9) Van Obbergghen-Schilling, E.; Pouysseggur, J. Signaling pathways

of the thrombin receptor. Thromb. Haemostasis 1993, 70, 163−167.

(10) Weitz, J.; Crowther, M. Direct thrombin inhibitors. Thromb. Res.

2002, 106, V275−V284.

(11) Weitz, J. I. A novel approach to thrombin inhibition. Thromb.

Res. 2003, 109 (Suppl. 1), S17−S22.

(12) Weitz, J.; Leslie, B.; Hudoba, M. Thrombin binds to soluble

fibrin degradation products where it is protected from inhibition by

O

dx.doi.org/10.1021/jm401169a | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

Spark, D.; Stobie, A.; Underwood, T. J.; Ward, R.; Wheeler, S.

Discovery of PF-184563, a potent and selective V1a antagonist for the

treatment of dysmenorrhoea. The influence of compound flexibility on

microsomal stability. Bioorg. Med. Chem. Lett. 2011, 21, 5684−5687.

(26) Guendouz, F.; Jacquier, R.; Verducci, J. Polymer bound 4-

dialkylamino pyridines: Synthesis, characterization and catalytic

efficiency. Tetrahedron 1988, 44, 7095−7108.

(27) Cruege, F.; Girault, G.; Coustal, S.; Lascombe, J.; Rumpf, P.

Basic properties of dimethylamino derivatives of pyridine and

pyrimidine. Bull. Soc. Chim. France 1970, 11, 3889−3894.

(28) Von der Saal, W.; Kucznierz, R.; Leinert, H.; Engh, R. A.

Derivatives of 4-aminopyridine as selective thrombin inhibitors. Bioorg.

Med. Chem. Lett. 1997, 7, 1283−1288.

(29) Engh, R. A.; Brandstetter, H.; Sucher, G.; Eichinger, A.;

Baumann, U.; Bode, W.; Huber, R.; Poll, T.; Rudolph, R.; Von der

Saal, W. Enzyme flexibility, solvent and weak interactions characterize

thrombin-ligand interactions: implications for drug design. Structure

1996, 4, 1353−1362.

(30) Wohnsland, F.; Faller, B. High-throughput permeability pH

profile and high-throughput alkane/water log P with artificial

membranes. J. Med. Chem. 2001, 44, 923−930.

(31) Fischer, H.; Kansy, M.; Avdeef, A.; Senner, F. Permeation of

permanently positive charged molecules through artificial membranes:

Influence of physico-chemical properties. Eur. J. Pharm. Sci. 2007, 31,

32−42.

(32) Olsen, J. A.; Banner, D. W.; Seiler, P.; Obst-sandr, U.; D’Arcy,

A.; Stihle, M.; Mueller, K.; Diederich, F. A fluorine scan of thrombin

inhibitors to map the fluorophilicity/fluorophobicity of an enzyme

active site: Evidence for C-F···C=O interactions. Angew. Chem., Int. Ed.

2003, 42, 2507−2511.

ylalanine in the P3 position. Implications on oral bioavailability and

antithrombotic efficacy. J. Med. Chem. 1997, 40, 1565−1569.

(42) Remko, M. Molecular structure, lipophilicity, solubility,

absorption, and polar surface area of novel anticoagulant agents. J.

Mol. Struct. Theochem. 2009, 916, 76−85.

(43) Bard, B.; Carrupt, P.-A.; Martel, S. Lipophilicity of basic drugs

measured by hydrophilic interaction chromatography. J. Med. Chem.

2009, 52, 3416−3419.

(44) Lopopolo, G.; Fiorella, F.; de Candia, M.; Nicolotti, O.; Martel,

S.; Carrupt, P.-A.; Altomare, C. Biarylmethoxy isonipecotanilides as

potent and selective inhibitors of blood coagulation factor Xa. Eur. J.

Pharm. Sci. 2011, 42, 180−191.

(45) Burla, M. C.; Caliandro, R.; Camalli, M.; Carrozzini, B.;

Cascarano, G.l.; Giacovazzo, C.; Mallamo, M.; Mazzone, A.; Polidori,

G.; Spagna, R. SIR2011: A new package for crystal structure

determination and refinement. J. Appl. Crystallogr. 2012, 45, 357−361.

(46) Sheldrick, G. M. A short history of SHELX. Acta Crystallogr.

2008, A64, 112−122.

(47) Farrugia, L. J. ORTEP-3 for Windows: A version of ORTEP-III

with a graphical user interface (GUI). J. Appl. Crystallogr. 1997, 30,

565.

(48) (a) Avdeef, A. pH-metric log P. Part 1. Difference plots for

determining ion-pair octanol-water partition coefficients of multiprotic

substances. Quant. Struct.-Act. Relat. 1992, 11, 510−517. (b) Avdeef, A.

J. pH-Metric log P. II. Refinement of partition coefficients and

ionization constants of multiprotic substances. Pharm. Sci. 1993, 82,

183−190. (c) Avdeef, A.; Box, K. J.; Comer, J. E. A.; Hibbert, C.; Tam,

K. Y. pH-metric logP 10. Determination of liposomal membrane-water

partition coefficients of ionizable drugs. Pharm. Res. 1998, 15, 209−

215.

(49) Vecchia, B. E.; Bunge, A. Transdermal Drug Delivery; Marcel

(33) Hoffmann-Roeder, A.; Schweizer, E.; Egger, J.; Seiler, P.; Obst-

Sander, U.; Wagner, B.; Kansy, M.; Banner, D. W.; Diederich, F.

Mapping the fluorophilicity of a hydrophobic pocket: Synthesis and

biological evaluation of tricyclic thrombin inhibitors directing

fluorinated alkyl groups into P pocket. ChemMedChem 2006, 1,

1205−1215.

(34) Obst, U.; Banner, D. W.; Weber, L.; Diederich, F. Molecular

recognition at the thrombin active site: structure-based design and

synthesis of potent and selective thrombin inhibitors and the X-ray

crystals structure of two thrombin-inhibitor complexes. Chem. Biol.

1997, 4, 287−295.

Dekker: New York, 2003; pp 57−141.

(50) Ruell, J. A.; Tsinman, K. L.; Avdeef, A. PAMPA: A drug

absorption in vitro model 5. Unstirred water layer in iso-pH mapping

assays and pKaflux-optimized design (pOD-PAMPA). Eur. J. Pharm.

Sci. 2003, 20, 393−402.

(51) Avdeef, A.; Strafford, M.; Block, E.; Balogh, M. P.; Chambliss,

W.; Khan, I. Drug absorption in vitro model: filter-immobilized

artificial membranes 2. Studies of the permeability properties of

lactones in Piper methysticum Forst. Eur. J. Pharm. Sci. 2001, 14, 271−

280.

(52) Bowers, K. J.; Chow, E.; Xu, H.; Dror, R. O.; Eastwood, M. P.;

́

Gregersen, B. A.; Klepeis, J. L.; Kolossvary, I.; Moraes, M. A.;

Sacerdoti, F. D.; Salmon, J. K.; Shan, Y.; Shaw, D. E. Scalable

algorithms for molecular dynamics simulations on commodity clusters.

In Proceedings of the ACM/IEEE Conference on Supercomputing (SC06),

Tampa, Florida, November 11−17, 2006. DOI: 10.1145/

1188455.1188544.

(53) Humphrey, W.; Dalke, A.; Schulten, K. VDM: visual molecular

dynamics. J. Mol. Graphics 1996, 14, 33−38.

(35) Cummings, M. D.; DesJarlais, R. L.; Gibbs, A. C.; Mohan, V.;

Jaeger, E. P. Comparison of automated docking programs as virtual

screening tools. J. Med. Chem. 2005, 48, 962−976.

(36) Bytheway, I.; Cochran, S. Validation of molecular docking

calculations involving FGF-1 and FGF-2. J. Med. Chem. 2004, 47,

1683−1693.

(37) Enyedy, I. J.; Egan, W. J. Can we use docking and scoring for

hit-to-lead optimization? J. Comput.-Aided Mol. Des. 2008, 22, 161−

168.

(54) De Cristofaro, R.; Akhavan, S.; Altomare, C.; Carotti, A.;

Peyvandi, F.; Mannucci, P. M. A natural prothrombin mutant reveals

an unexpected influence of A-chain structure on the activity of human

alpha-thrombin. J. Biol. Chem. 2004, 279, 13035−13043.

(55) Cheng, Y.; Prusoff, W. H. Relation between the inhibition

constant K1) and the concentration of inhibitor which causes fifty per

cent inhibition (I₅₀) of an enzymic reaction. Biochem. Pharmacol. 1973,

22, 3099−3108.

(38) Muley, L.; Baum, B.; Smolinski, M.; Freindorf, M.; Heine, A.;

Klebe, G.; Hangauer, D. G. Enhancement of hydrophobic interactions

and hydrogen bond strength by cooperativity: Synthesis, modeling,

and molecular dynamics simulation of a congeneric series of thrombin

inhibitors. J. Med. Chem. 2010, 53, 2126−2135.

́ ́

(39) Levesque, S.; St-Denis, Y.; Bachand, B.; Preville, P.; Leblond, L.;

Winocour, P. D.; Edmunds, J. J.; Rubin, J. R.; Siddiqui, M. A. Novel

bicyclic lactam inhibitors of thrombin: potency and selectivity

optimization through P1 residues. Bioorg. Med. Chem. Lett. 2001, 11,

3161−3164.

(56) Schrodinger Suite 2011; Schrodinger Inc., LLC, New York, 2011.

(40) Hanessian, S.; Therrien, E.; van Otterlo, W. A.; Bayrakdarian,

M.; Nilsson, I.; Fjellstrom, O.; Xue, Y. Phenolic P2/P3 core motif as

̈

thrombin inhibitors: Design, synthesis, and X-ray co-crystal structure.

Bioorg. Med. Chem. Lett. 2006, 16, 1032−1036.

(41) Tucker, T. J.; Lumma, W. C.; Lewis, S. D.; Gardell, S. J.; Lucas,

B. J.; Baskin, E. P.; Woltmann, R.; Lynch, J. J.; Lyle, E. A.; Appleby, S.

D.; Chen, I.-W.; Dancheck, K. B.; Vacca, J. P. Potent noncovalent

thrombin inhibitors that utilize the unique amino acid ^D-dicyclohex-

P

dx.doi.org/10.1021/jm401169a | J. Med. Chem. XXXX, XXX, XXX−XXX

Article Doi

Synthesis and biological evaluation of direct thrombin inhibitors bearing 4-(piperidin-1-yl)pyridine at the P1 position with potent anticoagulant activity

DOI: 10.1021/jm401169a

Source and publish data:

Authors:

Article abstract of DOI:10.1021/jm401169a

Full text of DOI:10.1021/jm401169a

Products guided by the article

R&D Labs maybe for 865611-13-8

Relevant to this article

Hot Product