Antimicrobial and cytotoxic arylazoenamines. Part III: Antiviral activity of selected classes of arylazoenamines

Bioorganic & Medicinal Chemistry 16 (2008) 8447–8465

Contents lists available at ScienceDirect

Bioorganic & Medicinal Chemistry

journal homepage: w ww.elsevier.com/locate/bmc

Antimicrobial and cytotoxic arylazoenamines. Part III: Antiviral activity

of selected classes of arylazoenamines ^q

c

Michele Tonelli ^a, Vito Boido ^a, Caterina Canu ^a, Anna Sparatore ^b, Fabio Sparatore ^a,, Maria Silvia Paneni ,

Maurizio Fermeglia ^c, Sabrina Pricl ^c, Paolo La Colla ^d, Laura Casula ^d, Cristina Ibba ^d, David Collu ^d,

Roberta Loddo ^d

*

^aDipartimento di Scienze Farmaceutiche, Università di Genova, Viale Benedetto XV 3, 16132 Genova, Italy

^bIstituto di Chimica Farmaceutica, Università di Milano, Viale Abruzzi 42, 20131 Milano, Italy

^cDipartimento di Ingegneria Chimica, dell’Ambiente e delle Materie prime, Università di Trieste, Via Valerio 10, 34127 Trieste, Italy

^dDipartimento di Scienze e Tecnologie Biomediche, Università di Cagliari, Cittadella Universitaria, S.S. 554, Km 4.500, 09042 Monserrato, Cagliari, Italy

a r t i c l e i n f o

a b s t r a c t

Article history:

Eighty-ﬁve arylazoenamines, characterized by different types of aryl and basic moieties, have been syn-

thesized and evaluated in cell-based assays for cytotoxicity and antiviral activity against a panel of ten

RNA and DNA viruses.

The most commonly affected viruses were, in decreasing order, CVB-2, RSV, BVDV, YFV, and Sb-1; the

remaining viruses were either not affected (HIV-1, VSV, and VV) or susceptible only to a very few com-

pounds (Reo-1 and HSV-1).

Received 19 May 2008

Accepted 13 August 2008

Available online 15 August 2008

Keywords:

Arylazoenamino compounds

Antiviral activity

RNA and DNA viruses

Pharmacophore models

Thirty-ﬁve compounds exhibited high activity, with EC₅₀in the range 0.8–10

lM, and other 28 com-

pounds had EC₅₀between 11 and 30 M, thus indicating that the arylazoenamine molecular pattern is

l

an interesting novel pharmacophore for antiviral agents against ssRNA viruses. Moreover, some com-

pounds (as 28, 32, 42, and 53) appear of high interest, being devoid of toxicity on the human MT-4 cells

(CC₅₀> 100 lM).

A ligand-based computational approach was employed to identify highly predictive pharmacophore

models for the most frequently affected viruses CVB-2, RSV, and BVDV. These models should allow the

design of second generation of more potent inhibitors of these human and veterinary pathogens.

1. Introduction

trimethylindoline,⁴as well as by reacting glyoxal monophenylhyd-

razone with secondary amine (piperidine⁵and pyrrolidine⁶). More-

Since 20 years, we have shown that by reacting

a

-(tert-amino)-

over, unstable arylazoenamines were supposed to be formed, at

ﬁrst, in the reaction of diazonium salts with unsaturated amines,

that, ﬁnally, gives rise to b-phenylhydrazones of substituted

glyoxals.⁷

ketones with excess of arylhydrazines, in the presence of acids, sta-

ble, red-orange arylazoenamines are formed, instead of (or in

addition to) the expected indoles.^1,2

It was assumed that the essential step for the formation of

this kind of compounds is the tautomerization of the initially

formed arylhydrazone into the corresponding arylhydrazoen-

amine, which is then oxidized by the excess of arylhydrazine,

with simultaneous formation of an arylamine and ammonia

(Scheme 1).

To the best of our knowledge, biological activities of arylazoen-

amines resulted unexplored until 1990,^8,9though they can be con-

sidered either as sub-structures of ortho (and para) aminoazo

compounds, or as vinylogues of aryltriazenes (either linear or cyc-

lic), many of which are endowed with either antimicrobial, anti-

parasitic or cytotoxic activities (Fig. 1).

Arylazoenamino compounds have been already obtained in the

past through different pathways, as by reacting arylhydrazines

with 2-(bromomethyl)-1-methylquinolinium bromide, or with

1,2,3,3-tetramethylindoleninium iodide,³by coupling diazonium

salts with the so-called methylene bases as 2-methylene-1,3,3-

Representative compounds of the ﬁrst type are the following:

2,6-diamino-3-phenylazo pyridine (pyridium), once used as uri-

nary antiseptic and analgesic; 5-(4-amino-1-naphthylazo)uracil,

highly effective against Schistosoma mansonii^10,11; 1-[3-(4-pheny-

lazo-5,6,7,8-tetrahydronaphth-1-yl)amino]-propylpiperidines¹²

and 1-[3-(4-phenylazo-5,6,7,8-tetrahydronaphth-1-yl)aminolup-

inane¹³very active against Mycobacterium tuberculosis; and Congo

Red, recently shown to be able to inhibit the deposition of prion

protein in amyloid plaques.^14,15

q

For notes I and II, see Refs. 8 and 9.

* Corresponding author. Tel.: +39 010 512721; fax: +39 010 3538358.

E-mail address: sparator@unige.it (F. Sparatore).

doi:10.1016/j.bmc.2008.08.028

8448

M. Tonelli et al. / Bioorg. Med. Chem. 16 (2008) 8447–8465

NHNH₂

R'

NH N=C CH

R"

R'

C=

R

R'"

N

NH NH

C

N

R""

R

R"

R

R'

_

N

R'

N

C

R'"

+

N=N

N

+^NH₃

NH₂

C

N

C

R"

+

R""

R

R"

R

Scheme 1.

H₂N

N=N

N

N H

NH₂R^-NH

N=N

O

R = H; CH₂CH₂N(C₂H₅)₂

NH

O

;

N=N

CH₂

NH

R

R = CH₂CH₂CH₂

N

H

N

NH₂

H₂N

N=N

SO₂ONa

R

SO₂ONa

Cl

N=N

N

R

CH₃

H₂N

N

N=N

N

Et

N

H₂N

O

C

O

NH₂

N

R

C

N

NH₂

N

CH₃

N

CH₃

N

H

N

N=N

N

O

C

O

CH₃

R = H; CONHCH ₂CH₂

CH₃

N

O

N

O

F

O

N=N

C

NH

F

N=N C

NH

F

N

C₂H₅

CH₂

N

H

N

H

N

CH₂)_n

Figure 1. Some biologically active azocompounds reminding arylazoenamines.

Triazenes as a class of compounds¹⁶hold an important position

as carcinogens and/or anticancer agents: 3,3-dimethyl-1-phenyl-

triazene and, even better, dacarbazine¹⁷are able to methylate

DNA, thus the former resulted in a powerful carcinogen, while

the latter is in clinical use for the treatment of malignant mela-

noma and Hodgkin’s disease. A structurally more complex triazene

M. Tonelli et al. / Bioorg. Med. Chem. 16 (2008) 8447–8465

8449

combines antitumor potential with inhibition of Pneumocystis cari-

2. Chemistry

nii.¹⁸Temozolamide,¹⁹closely related to dacarbazine, is currently

used to treat malignant glyomas. A recent development of this cyc-

lic acyltriazene is represented by the synthesis of benzotetrazepi-

nones,^20–24that are very promising agents against breast cancer.

It is worth noting that these compounds are weak alkylating agents

and may damage DNA by a novel mechanism.

The syntheses of arylazoenamines of general structures A–D

were carried out as indicated in Scheme 1 by heating the ethanolic

solution of relevant carbonyl derivative with the suitable arylhydr-

azine (or its hydrochloride), concd hydrochloric acid, and 85%

phosphoric acid in a molar ratio of 1:2:3:5. Yields were fairly good

except when the arylhydrazine is unsubstituted or is substituted

with electron-releasing groups. In the last cases a small amount

of the corresponding indole was formed and separated by column

chromatography.

Starting from 1-methyl-3-piperidone two isomeric compounds,

which can be separated by column chromatography, are simulta-

neously formed (Scheme 2). In all cases, the higher-melting isomer

shows the lower R_fin the T.L.C. As previously demonstrated,⁹the

low-melting isomer corresponds to the expected 3-arylazo-1-

methyl-1,4,5,6-tetrahydropyridine (C), while the high-melting

one corresponds to 2-arylazo-methylene-1-methylpyrrolidine

(D). The mechanism of this ring contraction has already been

discussed.⁹

The two isomeric compounds could also be obtained starting

from the corresponding 2-formyl-1-methylpyrrolidine arylhydraz-

one, but this procedure is more expensive and does not offer any

advantage.

The required octahydro-2H-quinolizin-1-one, octahydro-indoli-

zin-8-one, and 1-methyl-3-piperidone were prepared as described

in the literature.^34–361-Benzyl-3-piperidone is commercially

available.

Arylazoenamines could be also seen as simple azocompounds,

whose reactivity is somewhat enhanced by the conjugation with

the ene double bond, and as such they could be related to the even

more activated acyldiazenes. Simple aromatic azocompounds and

acyldiazenes are endowed with a variety of biological activities^25,26

(mainlyantifungaland cytotoxic), generallyrelatedtotheirabilityto

oxidize biological thiols, though other mechanisms should be sup-

posed in some cases. Antifungal²⁷and antiproliferative²⁸activities

have been found in a peculiar class of cyclic azocompounds (3,3-

disubstituted-3,4-dihydro-1,2,4-benzotriazines²⁹), which more-

over can display several other pharmacological activities depending

on the substituents that are present on the bicyclic system.^30,31

On this base, some years ago, thirty arylazoenamines were pre-

pared and tested^8,9for antimicrobial activity against eight Gram+

and Gramꢀ bacterial strains and against eight yeast like fungi.

While the antibacterial activity was very poor [MIC in the range

from 1 to 2 mg/ml (norﬂoxacine: MIC = 4–8

occasional values of 512 g/ml against Pseudomonas aeruginosa],

a moderate antifungal activity was displayed by most compounds,

with MIC from 32 to 512 g/ml against the whole set of tested fun-

gi (miconazole nitrate: MIC = 0.25–2 g/ml).

lg/ml) and only two

l

Moreover, a few of these compounds, which were additionally

tested against bovine viral diarrhoea virus (BVDV) and coxackie

virus type B2 (CVB-2), exhibited an interesting antiviral activity.

Therefore many new arylazoenamines have now been synthe-

sized, and together with part of the previously described com-

pounds were evaluated in cell-based assays for cytotoxicity and

antiviral activity against a large panel of RNA and DNA viruses.

On the whole, 85 arylazoenamines have been considered, which

may be allotted in seven groups (A–G) in relation to the nature of

the basic moiety (Fig. 2).

Compounds of structure E were obtained by coupling aryldiazo-

nium salts with 2-methylene-1,3,3-trimethylindoline as indicated

by König⁴(Scheme 3).

Finally, compounds of structures F and G were prepared by con-

densing glyoxal monoarylhydrazone with the suitable secondary

cyclic amine or piperazine (all commercially available) as indicated

by Severin et al.^5,6(Scheme 3).

3. Results and discussion

3.1. Biological activity—general considerations

Twenty-nine out of these compounds have been already de-

scribed: nine of structure A,^2,8ten of structure C,⁹six of structure

D,⁹two of structure E^3,32and two of structure F^5,6; one additional

compound of structure F (phenylazoethenemorpholine) is indi-

cated by Sommer³³as already known, but we failed to ﬁnd out

any data concerning its preparation and characterization.

Title compounds were evaluated for antiviral activity against

viruses representative of two of the three genera of Flaviviridae

family, that is, Flaviviruses (yellow fever virus, YFV) and Pestivirus-

es (bovine viral diarrhoea virus, BVDV), as Hepaciviruses can

N=N Ar

N

A: 1-17

B: 18-23

N=N Ar

CH₃

N

R

CH N=N Ar

N

CH

=N Ar

N

R

CH₃

D: 44-63

C: 24-43

E: 64, 65

X

CH=CH N=N Ar

N

Heteroaryl/Aryl

N

CH=CH N=N Ar

N

F: 66-73

G: 74-85

Figure 2. Structures of the investigated compounds.

8450

M. Tonelli et al. / Bioorg. Med. Chem. 16 (2008) 8447–8465

R

NH

N

CH₃

N

C

CH₃

R

or

R

N

CH

N

NH

N

CH

N

CH₃

D

CH₃

Scheme 2.

CH₃

CH N=N

+

-

CH₃

CH₂

R

N₂

+ Cl

N

R

CH₃

R

+

NH N=CH

CH=O

N=N CH=CH

N

HN

X

Scheme 3.

hardly be used in routine cell-based assays. Compounds were also

tested against representatives of other virus families. Among

ssRNA⁺were a retrovirus (human immunodeﬁciency virus type 1,

HIV-1) and two Picornaviruses (coxsackie virus type B2, CVB-2

and poliovirus type-1, Sabin strain, Sb-1); among ssRNA^ꢀwere a

Paramixoviridae (respiratory syncytial virus, RSV) and a Rhabdo-

viridae (vesicular stomatitis virus, VSV) representative. Among

double-stranded RNA (dsRNA) viruses was a Reoviridae represen-

tative (Reo-1). Two representatives of DNA virus families were also

included: herpes symplex type-1, HSV-1 (Herpesviridae), and vac-

cinia virus, VV (Poxviridae).

On the whole, 35 compounds have shown an EC₅₀6 10 lM

against at least one virus, 28 had EC₅₀between 11 and 30 lM,

and only 17 had a EC₅₀in the range 31–100 lM; thus, the aryl-

azoenamine molecular pattern is, indeed, an interesting novel

pharmacophore for antiviral agents against ssRNA viruses.

Cytotoxicity and antiviral activities of test and reference com-

pounds are reported in Tables 2–5.

3.2. Cytotoxicity on host cells

Test compounds showed different degrees of cytotoxicity

against the conﬂuent cell monolayers (in stationary growth) used

to support the multiplication of the different viruses.

The most susceptible to toxicity were the exponentially grow-

ing lymphoblastoid human cells (MT-4) used to grow HIV-1,

while the non-human host cell lines exhibited a progressively re-

duced sensitivity in the order MDBK > BHK > Vero-76. Indeed,

NM-108 (2⁰-b-methylguanosine), 6-azauridine, AZT (3⁰-azido-

thymidine), and ACG (acyclo-guanosine) were used as reference

inhibitors of ssRNA⁺, ssRNA^ꢀ, HIV-1, and DNA viruses, respectively.

Interestingly, 80 over 85 tested compounds exhibited antiviral

activity against one or more viruses: in particular, 27 compounds

exhibited a selective activity against a single virus, while 33, 15,

4, and 1 were, respectively, active against two, three, four, and ﬁve

viruses. None of the test compounds inhibited the multiplication of

HIV-1, VSV, and VV, and only three compounds exhibited a modest

activity against Reo-1. Similarly, only four compounds displayed

moderate activity against HSV-1, but it must be noted that, be-

cause of shortage of products, only 50 over 85 compounds were

tested against it. In both cases, the observed activity was secondary

to the activity on other viruses.

only 25.9% of compounds showed no toxicity (CC₅₀P 100

against MT-4 versus 47.1% on Vero-76 cells; on the contrary

45.9% of compounds exhibited a CC₅₀630 M for MT-4 compared

with 10.6–11.8% in the case of BHK and Vero-76 cells, respec-

tively. CC₅₀so low as 2 M were found for MT-4 cells, while

the lowest CC₅₀values for BHK and Vero-76 were, respectively,

10 and 11 M.

lM)

l

Cytotoxicity is clearly inﬂuenced by the nature of both the ami-

no heads and the arylazo moieties; however, the former seem to be

more important than the latter. Considering together the four cell

lines, it is observed that the azoethene derivatives of structure F

On the other hand, an increasing number of compounds exhib-

ited antiviral activity, sometimes of very high level, against, in the

order, Sb-1 (12), YFV (16), BVDV (28), RSV (34), and CVB-2 (62) (Ta-

ble 1). Twenty-eight of the 62 active compounds inhibited the mul-

tiplication of CVB-2 more strongly than the reference compound

and G are the most toxic, exhibiting for about 54% a CC₅₀6 30

while the tetrahydropyridines of structure C are the least toxic,

with only 2.5% of compounds with CC₅₀6 30 M. Viceversa,

ꢁ61% of compounds of the last type showed no toxicity

(CC₅₀P 100 M), while only 15% of compounds F and G have a

CC₅₀in this range. The remaining types of compounds (A, B, D,

lM,

NM-108 (EC₅₀= 20

compounds exhibited a EC₅₀value comparable with that of the

respective reference drug, NM-108 (EC₅₀= 1.7 M) and 6-azauri-

dine (EC₅₀= 1.2 M).

lM). Concerning BVDV and RSV, only a few

l

M. Tonelli et al. / Bioorg. Med. Chem. 16 (2008) 8447–8465

8451

Table 1

Number of active compounds on susceptible viruses and range of their EC₅₀

Virus

No. of active over

No. of active compounds (range of EC₅₀lM)

85 tested compounds^a

Sb-1^b

YFV^b

12

16

28

34

62

2(8)

6(12–27)

4(12–23)

11(11–21)

15(11–30)

1(45)

3(53–60)

4(7–10)

4(6–10)

6(6–9)

3(36–48)

2(40, 46)

1(40)

5(65–100)

6(60–100)

11(51–100)

BVDV^b

RSV^c

5(0.8–5.5)

6(0.9–5)

10(0.9–5)

CVB–2^b

10(6–10)

18(12–30)

13(32–50)

a

Compounds with EC₅₀> 100 lM or higher than CC₅₀are considered inactive.

Single-stranded, positive RNA virus.

Single-stranded, negative RNA virus.

b

c

and E) stay in the middle between the above extremes, with only

small differences among them.

The high cytotoxicity observed for a few compounds, particu-

larly against MT-4 cells (i.e., 7, 12, 55, 69, with CC₅₀= 2–4 lM),

while lowering their interest as antivirals, may warrant deeper

investigations to obtain novel antiproliferative agents.

Compounds endowed with the highest activity against a given

virus may also be selective for it (26: EC₅₀= 8 M for Sb-1; 32:

EC₅₀= 0.8 M for BVDV; 84: EC₅₀= 0.9 M for RSV), or may exhibit

simultaneously good activity on one or two other viruses, as com-

pound 75 (EC₅₀= 0.9 M for CVB-2) and 47 (EC₅₀= 7 M for YFV)

that inhibit also, respectively, RSV and BVDV and CVB-2 with

EC₅₀in the range 5–16 M.

l

3.3. Structure–activity relationships

Generally, compounds acting on three or more viruses display

only a moderate or even a modest degree of activity. Interesting

exceptions are represented by compounds 28, 47, 76, 83, and, par-

Also the antiviral activity, as already seen for cytotoxicity, is

inﬂuenced by the nature of the amino head and the arylazo moi-

ety; while the former seems to mainly address the activity versus

the different types of viruses, the latter principally regulate the

intensity of the antiviral action. However, in some cases the deﬁni-

tion of the real role of each structural feature may result very dif-

ﬁcult and even puzzling. Indeed, the shift of a substituent from a

position to another of the aromatic moiety may produce a qualita-

tive more than a quantitative variation of activity. Moreover, not

all the possible combinations of the basic heads and aromatic moi-

eties could be explored.

Among the ﬁve compounds completely devoid of antiviral

activity, the lack of activity can be related either to the nature of

the basic head (indoline: 64, 65), or to the aromatic moiety (3,5-

bis-triﬂuoromethylphenyl: 36, 55; pentaﬂuorophenyl: 41), since

the corresponding aromatic moieties of the former and the basic

heads of the latter are commonly present in very active

compounds.

Despite these shortcomings, some general rules can be drawn.

The arylazohexahydroquinolizines A and the arylazoethenam-

ines F deﬁnitely appear as the most suitable structures for activity

against CVB-2; indeed, only one compound (16) of the ﬁrst type

and no one of the second failed to inhibit its multiplication.

Also most of the structurally related arylazohexahydroindoli-

zines B and arylazoethenepiperazines G are active against CVB-2,

though acting also against Sb-1 and RSV, respectively. About a half

of compounds active against RSV is found in groups F and G; the

remaining being evenly distributed among the others.

ticularly, 81, that exhibited a EC₅₀= 7 lM against CVB-2, YFV and

RSV. Most of the nineteen large spectrum compounds belong

(13/19) to groups C and D; the remaining are found in groups G

(3/19), A (2/19), and B (1/19).

3-(4-Chlorophenylazo)-1-methyltetrahydropyridine (28) is ac-

tive on Sb-1, CVB-2, RSV and Reo virus, with EC₅₀= 8, 5, 20 and

32 lM, respectively. The shift of chlorine from para to ortho posi-

tion (26) produces the loss of the activity on CVB-2, RSV and Reo

viruses, leaving unchanged that on Sb-1; however when the chlo-

rine is placed on meta position (27) the activity on CVB-2 and RSV

is maintained (though reduced), but that on Sb-1 and Reo is lost,

while appears a good activity on BVDV. The activity on Sb-1 is lost

also when the 4-chlorophenylazo substituent of compound 28 is

attached on the methylenepyrrolidine moiety (46) that is isomeric

to the tetrahydropyridine ring of 28. Curiously, compound 46

exhibits the same spectrum of activity (CVB-2, RSV, and BVDV) of

the meta chloro derivative 27. Differently from what is observed

for the isomeric 4-chlorophenylazo derivatives 28 and 46, the iso-

meric 3-nitrophenylazo compounds 32 and 50 maintain a high and

selective activity against BVDV.

The varied activities shown by the four isomeric compounds 26,

27, 28, and 46 demonstrate the intriguing complexity of the struc-

ture–activity relationships for this kind of compounds. Therefore,

in an attempt to reach a better understanding of the structural

requirements for antiviral activity on the 85 arylazoenamines A–

G, a ligand-based computational approach was employed in order

to derive a highly predictive pharmacophore model for the most

frequently affected viruses CVB-2, RSV, and BVDV. Results of this

study are discussed in the next section.

Interestingly, no one of compounds of F and G groups was active

against BVDV.

The arylazotetrahydropyridines C and the arylazomethylene-

pyrrolidines D share in almost equal measure the activity against

CVB-2 and BVDV or on both types of viruses. It is worth noting that

the exchange of the methyl group with a benzyl abolishes the

activity against BVDV, leaving or enhancing that against CVB-2.

Sole activity on YFV is never found in any tested compound, but

in association with activity on CVB-2 and/or BVDV is most com-

monly found among arylazoenamines F and arylazoethenepipera-

Despite the high antiviral activity of many compounds, the cor-

responding selectivity indexes (SI) are most commonly below 10,

due to the high toxicity on the host cells often observed. Neverthe-

less, 17 compounds exhibited a selectivity index in the range from

>10 to >125, for at least one of the viruses CVB-2, BVDV, RSV, and

Sb-1. The reference compound NM-108 exhibits a SI >5 and >58.8

for CVB-2 and BVDV viruses, respectively.

Thus, taking into account the EC₅₀and SI values, the best com-

pounds can be arranged in the following order of decreasing po-

tency and selectivity; for CVB-2: 75 > 66 P 53 P 28; for BVDV:

32 > 22 > 42 = 50 > 16; for RSV: 84 > 23, and for Sb-1: 26 P 28. In

these sequences even the least potent compounds compare well

with the reference compounds. It must be observed, however, that

in view of a therapeutical application the cytotoxicity on human

zines G. Activity against YFV is shown by only

compounds of group A, but by none of group B.

1 over 17

Selective activity on Sb-1 is shown by a single compound of

group C, while activity on Sb-1 associated with activity on CVB-2

and BVDV is exhibited by several compounds of the same and

other groups, especially by hexahydroindolizines B.

Table 2

Cytotoxicity against MT-4, MDBK, BHK and Vero-76 cell lines and BVDV, YFV, CVB-2, RSV, and Sb-1 inhibitory activity of arylazoenamines of structures A and B

N=N Ar

N

A: 1-17

B: 18-23

M) YFV EC₅₀

a

b

c

d

e

f

g

h

i

Compound

Aryl=

MT-4 CC₅₀

(

lM)

MDBK CC₅₀

(lM)

BVDV EC₅₀

(lM)

BHK CC₅₀

(

l

(l

M)

Vero-76 CC₅₀

(

lM)

CVB-2 EC₅₀

(l

M)

RSV EC₅₀

(

lM)

Sb-1 EC₅₀(lM)

1

2

3

4

5

6

7

8

4-F-Ph^l

2-Cl-Ph

3-Cl-Ph

4-Cl-Ph

3-Br-Ph

4-Br-Ph

3-CF₃-Ph

3-NO₂-Ph

4-NO₂-Ph

3,4-DiCl-Ph

3,5-DiCF₃-Ph

3-CF₃-4-F-Ph

3-CF₃-4-Cl-Ph

3-CF₃-4-Br-Ph^m

3-NO₂-4-Cl-Ph

1-Naphthyl

7-Chloro-1-

quinolyl

94

78

16

42

9

56

>100

20

60

10

59

8

>100

30

31

>100

35

>100

18

>100

40

>56

>100

>20

>60

>10

>59

>8

>100

93

46

69

49

62

40

87

>100

54

19

60

41

>100

10

100

>93

>46

>69

>49

P62

>40

>87

P100

>54

>19

>60

>41

80

>100

60

80

>100

92

27

32

22

19

18

10

21

71

26

38

41

45

64

42

51

>90

51

>100

60

18

>100

>60

>80

>100

>92

25

>100

>92

>100

>90

>100

75

27

4

>100

90

>100

20

37

91

35

34

3

19

34

33

12

15

>100

86

9

100

>100

100

>100

100

90

>100

25

>100

>90

>85

10

11

12

13

14

15

16

17

>30

>31

>100

>35

61

>18

6

5

>10

>20

>48

20

48

>90

>85

85

18

19

20

21

22

23

4-Cl-Ph

>100

26

30

18

52

>100

2

30

68

67

21

>30

>68

>67

>21

3.5

>100

1.7

>100

34

95

61

51

>100

90

>34

>95

>61

>51

>100

1

75

10

54

30

9

44

>100

20

>75

>100

>90

7

20

60

45

12

>90

>100

>20

3-NO₂-Ph

3,4-DiCl-Ph

3-CF₃-4-Cl-Phⁿ

1-Naphthyl

1-Phthalazyl

>100

100

90

>100

20

91

>100

NM-108 (2⁰-b-methylguanosine)

>100

1.2

6-Azauridine

>100

26

>20

a

Compound concentration (

l

M) required to reduce the viability of mock-infected MT-4 (CD4⁺human T cells containing an integrated HTLV-1 genome) cells by 50%, as determined by the MTT method.

M) required to reduce the viability of mock-infected MDBK (bovine normal kidney) cells by 50%, as determined by the MTT method.

M) required to achieve 50% protection of MDBK cells from BVDV (bovine viral diarrhoea virus) induced cytopathogenicity, as determined by the MTT method.

M) required to reduce the viability of mock-infected BHK (hamster normal kidney ﬁbroblast) monolayers by 50%, as determined by the MTT method.

M) required to achieve 50% protection of BHK cells (kidney ﬁbroblast) from YFV (yellow fever virus) induced cytopathogenicity, as determined by the MTT method.

M) required to reduce the viability of mock-infected VERO-76 (monkey normal kidney) monolayers by 50%.

M) required to reduce the plaque number of CVB 2 (coxsackie virus B2) by 50% in VERO-76 monolayers.

M) required to reduce the plaque number of RSV (respiratory syncitial virus) by 50% in VERO-76 monolayers.

M) required to reduce the plaque number of Sb-1 (poliovirus type-1, Sabin strain) by 50% in VERO-76 monolayers.

b

c

d

e

f

g

h

i

l

Active also on Reo-1 virus (EC₅₀= 34

Active also on HSV-1 virus (EC₅₀= 25

Active also on HSV-1 virus (EC₅₀= 33

lM).

m

n

l

M).

l

M).

Table 3

Cytotoxicity against MT-4, MDBK, BHK, and Vero-76 cell lines and BVDV, YFV, CVB-2, RSV, and Sb-1 inhibitory activity of arylazoenamines of structure C

N=N Ar

N

R

C: 24-43

a

b

c

d

e

f

g

h

i

Compound

Aryl=

MT-4 CC₅₀

(l

M)

MDBK CC₅₀

(l

M)

BVDV EC₅₀

(lM)

BHK CC₅₀

(l

M)

YFV EC₅₀

(lM)

Vero-76 CC₅₀

(l

M)

CVB-2 EC₅₀

(lM)

RSV EC₅₀

(

lM)

Sb-1 EC₅₀(lM)

R = CH₃

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

Phenyl

4-F-Ph

2-Cl-Ph

>100

P100

>100

83

>100

43

>100

62

50

71

>100

48

>100

87

>100

>48

100

12

>58

19

60

90

0.8

11

>100

7

>64

>100

>28

40

>61

>100

5

>100

90

>100

85

>100

95

>100

38

82

80

44

>100

>90

>100

>85

>100

71

100

>100

>95

>100

>38

>82

>80

>44

>100

P100

77

>100

75

>100

53

>100

34

5

27

>100

>71

14

>100

35

34

75

30

>100

40

20

>100

>77

8

>75

8

>100

25

>100

>70

3-Cl-Ph

4-Cl-Ph^l

58

90

3-Br-Ph

>100

62

>100

73

64

>100

28

>100

61

>100

100

>100

71

>100

70

90

80

>100

P100

18

4-Br-Ph^m

3-CF₃-Ph

>100

>71

>100

>70

>90

>80

>100

100

>100

3-NO₂-Ph

4-NO₂-Ph

4-CH₃-Ph

3,4-DiCl-Ph

3,5-DiCF₃-Ph

3-CF₃-4-F-Ph

3-CF₃-4-Cl-Ph

3-CF₃-4-Br-Ph

3-NO₂-4-Cl-Ph

PentaF-Ph

1-Naphthyl

>90

>80

>100

40

>100

76

>100

R = CH₂-C₆H₅

43

NM-108 (2⁰-b-

3-CF₃-4-Cl-Ph

43

>100

28

>100

P28

1.7

32

90

22

1

48

>100

4

20

>48

>100

>48

>100

methylguanosine)

6-azauridine

2

>100

26

20

>20

1.2

>20

For the meaning of ^a–i: see Table 2.

^lActive also on Reo-1 virus (EC₅₀= 32

l

M). ^mActive also on Reo-1 virus (EC₅₀= 76

lM).

Table 4

Cytotoxicity against MT-4, MDBK, BHK, and Vero-76 cell lines and BVDV, YFV, CVB-2, RSV, and Sb-1 inhibitory activity of arylazoenamines of structures D and E

CH₃

H

N

R

CH N=N Ar

N

CH₃

N=N Ar

D: 44-63

E: 64, 65

a

b

c

d

e

f

g

h

i

Compound

Aryl=

MT-4 CC₅₀

(l

M)

MDBK CC₅₀

(l

M)

BVDV EC₅₀

(lM)

BHK CC₅₀

(l

M)

YFV EC₅₀

(

lM)

Vero-76 CC₅₀

(lM)

CVB-2 EC₅₀

(lM)

RSV EC₅₀

(lM)

Sb-1 EC₅₀(lM)

R=CH₃

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

Phenyl

4-F-Ph

4-Cl-Ph

3-Br-Ph

4-Br-Ph

3-CF₃-Ph

3-NO₂-Ph

4-NO₂-Ph^l

4-CH₃-Ph

4-CH₃O-Ph

3,4-diCl-Ph^m

3,5-diCF₃-Ph

3-CF₃-4-F-Phⁿ

3-CF₃-4-Br-Ph

PentaF-Ph

>100

100

>100

55

23

49

>100

59

85

52

>100

90

>100

67

90

36

10

18

21

9

11

21

5.5

19

20

18

16

>36

100

46

>100

80

55

41

>100

50

>100

77

23

48

>100

36

P55

7

82

75

92

80

100

>100

100

93

19

>75

24

16

36

>100

79

60

5

>100

>93

71

100

>100

20

12

20

>80

20

>100

75

>100

>93

>100

75

>82

>75

>92

>80

>100

>50

>100

65

>100

>77

>23

>48

>41

>100

53

>93

>100

60

25

>100

39

2

11

>100

41

>100

P100

>100

24

8

>100

R=CH₂-C₆H₅

59

60

61

62

63

64

65

Phenyl

4-Cl-Ph

4-CH₃-Ph

3,4-diCl-Ph^o

3-CF₃-4-Cl-Ph

Phenyl^p

22

17

10

19

20

30

55

>100

74

18

22

79

14

20

72

>100

>74

>18

>22

>79

>14

>20

>72

1.7

31

20

25

50

22

47

21

90

>31

>20

>25

>50

>22

>47

>21

1

87

70

50

55

80

73

>100

16

23

8

21

30

>87

>70

>50

>55

>80

>73

>100

>70

>50

>55

>80

>73

>100

30

>73

>100

20

4-Cl-Ph

NM-108 (2⁰-b-

methylguanosine)

6-Azauridine

2

>100

26

20

>20

1.2

>20

For the meaning of^a–i: see Table 2; ^lmixture (2:1) with C isomer as seen from NMR spectrum; ^mactive also on HSV-1 virus (EC₅₀= 40

l

M); ⁿactive also on HSV-1 virus (EC₅₀= 60

l

M); ^omixture (3:1) with C isomer as seen from NMR

spectrum; ^pas perchlorate.

Table 5

Cytotoxicity against MT-4, MDBK, BHK, and Vero-76 cell lines and BVDV, YFV, CVB-2, RSV, and Sb-1 inhibitory activity of arylazoenamines of structures F and G

X

CH=CH N=N Ar

N

Heteroaryl/Aryl

N

CH=CH N=N

N

F: 66-73

G: 74-85

a

b

c

d

e

f

g

h

i

Compound

X=

Ar=

MT-4 CC₅₀

(lM)

MDBK CC₅₀

(lM)

BVDV EC₅₀

(l

M)

BHK CC₅₀

(l

M)

YFV EC₅₀

(lM)

Vero-76 CC₅₀

M)

CVB-2 EC₅₀

(l

M)

RSV EC₅₀

(

lM)

Sb-1 EC₅₀(lM)

(

l

66

67

68

69

70

71

72

73

74

75

76

77

78

79

80

81

82

83

84

85

/

Phenyl

4-Cl-Ph

4-EtO-CO-Ph

Ph

4-EtO-CO-Ph

Ph

8

63

30

100

7

>100

28

14

13

P100

10

5

10

13

7

7.5

9

20

8

19

>63

>30

>100

>7

>100

>28

>14

>13

>100

>10

>5

>10

>13

>7

>7.5

>9

82

>82

46

>100

23

>100

>75

>100

42

>62

20

>69

10

7.5

9

70

22

>100

15

>100

30

22

55

50

34

35

14

11

45

11

45

4

50

7

35

7

6

12

6

0.9

5

6

30

>22

>100

>15

>100

15

>30

4

7

5

11

6

9

5

7

4

13

0.9

6

>100

1.2

>70

>22

>100

>15

27

61

26

2

33

7

>100

67

>100

75

>100

60

62

50

69

47

29

43

48

49

46

90

>100

CH₂

O

>30

>22

>55

>50

>34

>35

>14

>11

>45

>11

>45

>11

>38

>100

>20

S

Ph

19

17

19

17

15

32

17

16

19

21

30

41

21

29

>100

2

EtO-CON Ph

2-Pyridinyl

2-Pyrimidinyl

Phenyl

2-F-Ph

4-F-Ph

2-Cl-Ph

3-Cl-Ph

4-Cl-Ph

5

>11

7

>11

14

>11

5

3-CF₃-Ph

4-NO₂-Ph

4-CH₃CO-Ph

4-CH₃O-Ph

>20

>8

>19

1.7

12

>49

>46

1

11

38

>100

20

NM-108 (2⁰-b-methylguanosine)

>100

20

>20

6-Azauridine

>100

26

For the meaning of^a–i: see Table 2.

8456

M. Tonelli et al. / Bioorg. Med. Chem. 16 (2008) 8447–8465

cells is more signiﬁcant than that on host cells, and therefore,

among the foregoing sequences, compounds 53, 28, 32, and 42 ap-

pear of high interest being devoid of toxicity on the human MT-4

of 84 disposes over the RAr feature of Hypo1_RSV, whereas one

nitrogen atom of the aza group again maps the remaining HBD

feature.

cells (CC₅₀> 100

l

M).

As expected, for the majority of the less active compounds we

found that they do not map adequately or eventually miss some

of the relevant hypothesis features; for some of these molecules,

their lack of afﬁnity could be primarily ascribed to their inability

to achieve an energetically favorable conformation required by

the pharmacophore and shared by the active compounds.

Taken together, all these evidences concur to indicate that each

Hypo1 model provides reasonable pharmacophoric characteristics

of the three virus inhibitors for the components of their corre-

sponding activities.

3.4. Molecular modeling

3.4.1. Pharmacophore models generation

It is a widely accepted concept that three-dimensional (3D)

pharmacophore modeling is a well-behaved approach to quantita-

tively explore the common chemical characteristics among a con-

siderable number of different structures. Indeed, a computed

pharmacophore model can, in principle, be as good as the informa-

tion data input. Nevertheless, to achieve such a quality, at least

three must obey rules should be respected in a three-dimensional

quantitative structure–activity relationship (3D-QSAR) generation

using CATALYST: (i) the training set must be constituted by a wide

population (at least 16 items) of structurally diverse representative

covering at least four orders of magnitude of activity, (ii) the most

active compound should inevitably be included in the training set,

and (iii) all biological data must be obtained by homogeneous

procedures.

In our case, three training sets (TrS)—consisting of 18, 43, and

21 compounds for BVDV, CVB-2, and RSV, respectively—were pre-

pared by considering structural diversities and by the widest cov-

erage of in vitro activity range possible (see Table 6). The molecules

in each TrS were selected according to the following criteria: (i)

each TrS for each virus should contain structures from each class

of active compounds and (ii) each TrS should cover the molecular

bioactivities (EC₅₀) as widely as possible. Should there be only

one compound with maximum or minimum order of bioactivity

in a class, this compound had to be assigned to the training set.

Compounds 51 and 62, although active against BVDV and/or

CVB-2, were not included in the corresponding training sets as they

were experimentally tested as a mixture of isomers.

A set of 10 hypotheses was automatically generated for each of

the three BVDV_TrS, CVB-2_TrS, and RSV_TrS training sets using

the HypoGen module in CATALYST. Hydrophobic aromatic (HAr)

sites, hydrogen bond acceptor (HBA), hydrogen bond donor

(HBD), positive ionizable (PI), and ring aromatic (RAr) moieties

were selected as the functional features to describe the antiviral

activity of the compounds. The best predictive hypothesis (Hypo1)

for each training set obtained from HypoGen has three features for

BVDV, and two features for CVB-2 and RSV. In detail, Hypo1_BVDV

is characterized by two hydrogen bond donors and one hydropho-

bic aromatic, while both Hypo1_CVB-2 and Hypo1_RSV have one

hydrogen bond donor and one aromatic ring (see Fig. 3). These

three hypotheses were also all characterized by the highest cost

differences, the lowest root-mean-square (RMSD) deviations, and

the highest correlation coefﬁcients (see below).

3.4.2. Cost analysis

When generating an hypothesis, CATALYST performs a ﬁxed cost

calculation, which represents the simplest model that ﬁts all data

perfectly, and a null cost calculation, which presumes that there

is no relationship in the data set, and that the experimental activ-

ities are normally distributed about their mean. Accordingly, a gen-

erated pharmacophore hypothesis with a score that is substantially

lower than that of the corresponding null hypothesis is likely to be

statistically signiﬁcant. Two further criteria should be satisﬁed in

assessing the validity of a given hypothesis: (a) the cost of the

hypothesis should be close to the ﬁxed cost and (b) the conﬁgura-

tional cost, that is the magnitude of the hypothesis space for a given

TrS of compounds, should be less than 17. If this last cost exceeds

17, there are more degrees of freedom in the TrS that the CATALYST

algorithm can properly handle and, consequently, the correspond-

ing pharmacophore is likely to be poorly meaningful.

The quality of the three statistically most signiﬁcant hypotheses

out of the ten generated by HypoGen for each TrSs (Hypo1_BVDV,

Hypo1_CVB-2, and Hypo1_RSV, respectively) was evaluated by

considering the relevant cost functions (see Table 7).

Detailedly, for BVDV the cost range between Hypo1_BVDV and

the ﬁxed cost is 8.59, while that between the null hypothesis and

Hypo1_BVDV is 77.69 (see Table 7, ﬁrst row). According to these

scores, criterion (a) is satisﬁed as the total cost of Hypo1_BVDV

is much closer to the ﬁxed cost than to the null cost. Furthermore,

a high correlation coefﬁcient (q) of 0.98, a RMSD value of 0.50, and

a conﬁgurational cost of 12.78 (criterion (b)) allow us to conclude

that Hypo1_BVDV is a reliable pharmacophore model with high

predictivity for our compounds. Treating the 43 test compounds

active against CVB-2 with HypoGen yielded the corresponding best

hypothesis Hypo1_CVB-2 (see Table 7, second row), characterized

by a total cost of 107.92. The total ﬁxed cost for this run was

89.62 and the cost of the null hypothesis was 173.57. Thus, the

large difference between the null and total hypothesis costs

(65.65) and the closeness of the hypothesis cost to the ﬁxed total

cost (18.30), coupled with a low conﬁgurational cost (13.56), a

In harmony with the predicted activities, the three most active

compounds in the three TrSs (i.e., compounds 32 in BVDV_TrS, 75

in CVB-2_TrS, and 84 in RSV_TrS in Table 6) can be nicely mapped

onto the corresponding Hypo1 models by the best ﬁt values, as can

be appreciated from Figure 4. As we can see from this ﬁgure, com-

pound 32 ﬁts very well all features of the pharmacophore model

Hypo1_BVDV. In fact, the ﬁrst HBA is located on the nitro group,

the phenyl moiety overlaps with the HAr feature, and the second

HBA is mapped by the aza group. Noticeably, although the other

two hypotheses possess the same chemical features, the different

relative orientations and distances of these groups in the two cases

impose different mapping modes. Thus, in the case of compound

75, it is the aromatic ring bound to the aza group that overlaps

neatly over the RAr site, while one of the two nitrogen of the

pyrimidine substituent serves as an HBA in Hypo1_CVB-2. On the

other hand, the aromatic moiety of the acetylphenyl substituent

good RMSD (0.94), and a high q value (0.97), all speak in favor of

the statistical signiﬁcance of the top-ranked hypothesis. Finally,

also for Hypo1_RSV the ample difference between the null and to-

tal hypothesis cost (79.85), the high correlation coefﬁcient (0.95),

and the excellent RMSD value (0.38) (see Table 7, last row) indi-

cates a reliable predictivity of the corresponding 3D-QSAR pharma-

cophore and conﬁrm that it is not a chance correlation.

3.4.3. Validation of the pharmacophore models

To cross-validate the observed correlations, we prepared three

test sets (TsS) of further 9, 17, and 12 compounds for BVDV,

CVB-2, and RSV, respectively. Interestingly, good correlation coefﬁ-

cients (0.97, 0.95, and 0.96) were observed when a regression anal-

ysis was performed by mapping each test set onto the features of

the corresponding best pharmacophore hypothesis. The predicted

and the experimental EC₅₀values for the three TsSs along with

M. Tonelli et al. / Bioorg. Med. Chem. 16 (2008) 8447–8465

8457

Table 6

Estimated and experimentally determined activity values of the training set compounds for BVDV, CVB-2, and RSV

BVDV

EC_50,exp.

CVB-2

EC_50,exp.

M)

RSV

EC_50,exp.

Training set

compounds

EC_50,est.

M)

Error^a

Training set

compounds

EC_50,est.

Error^a

Training set

compounds

EC_50,est.

M)

Error^a

(lM)

(

l

(l

(lM)

(l

M)

(l

32

22

42

17

50

47

33

27

45

53

52

49

39

30

14

9

0.8

3.5

5

5.5

9

0.9

7.9

3

7.8

4.6

1.1

2.3

75

43

66

28

53

76

79

85

72

77

69

71

61

21

18

6

73

35

83

59

47

4

60

46

29

1

0.9

4

5

6

7

8

1.6

5.8

3.1

7

2.2

5.8

6

5.3

5.9

6

1.8

1.5

84

73

82

75

79

77

23

76

45

83

71

30

44

48

4

0.9

4

5

6

7

11

12

13

15

18

20

25

30

40

75

1.1

1.2

1.8

1.0

7.2

4.1

4.5

5

10

9.6

15

10

18.9

19

34

32

34

29

34

33

68

65

93

ꢀ1.7

ꢀ1.3

1.6

1.4

ꢀ1.1

ꢀ0.8

1.7

1.2

ꢀ2.3

1.2

1.0

15

13

19

12

19

26

16

41

55

58

93

94

96

ꢀ1.2

1.4

11

12

18

20

21

40

60

61

86

90

1.6

1.1

ꢀ1.1

ꢀ1.5

ꢀ1.0

1.3

1.1

1.0

ꢀ1.3

1.3

1.1

1.7

1.6

1.3

5

10

ꢀ1.4

1.4

1.5

ꢀ1.3

1.0

12

15

9

16

10

13

ꢀ1.1

1.1

1.0

ꢀ1.0

9

1.7

10

12

14

16

19

23

24

27

32

34

35

36

42

44

45

50

51

53

54

60

64

71

76

100

ꢀ1.1

1.4

1.6

59

66

27

15

58

41

ꢀ1.0

31

56

ꢀ1.2

ꢀ1.1

ꢀ2.0

ꢀ1.5

1.3

1.1

100

ꢀ1.2

ꢀ1.1

6.9

11

20

23

13

28

26

28

35

31

30

38

42

44

58

51

56

48

55

67

70

81

66

100

1.2

ꢀ1.8

1.2

ꢀ1.0

1.0

1.1

ꢀ1.1

ꢀ0.9

ꢀ1.2

ꢀ1.1

ꢀ1.0

1.2

2

38

37

48

14

22

12

68

17

15

25

19

52

13

8

1.0

1.1

ꢀ1.1

1.0

1.1

42

57

ꢀ1.2

108

1.1

a

Value in the error column represent the ratio of the estimated activity to measured activity, or its negative inverse if the ratio is less than one.

the respective errors are shown in Table 8. The actual activity val-

ues are within the limits of uncertainty (i.e., 2.0), thus conﬁrming

the validity of the original 3D pharmacophores.

4. Conclusions

Several sets of arylazoenamines (A–G, for a total of 85 com-

pounds), characterized by different types of aryl and basic moie-

ties, have been synthesized and assayed for antiviral activity

against a panel of ten RNA and DNA viruses. Eighty over 85 com-

pounds exhibited selective or large spectrum antiviral activity; in

particular, 12 were active on Sb-1, 16 on YFV, 28 on BVDV, 34 on

RSV, and 62 on CVB-2. Two viruses (Reo-1 and HSV-1) were sus-

ceptible only to very few compounds; the remaining (HIV-1, VSV,

and VV) were not affected by any compound.

Thirty-ﬁve compounds have shown high activity, with EC₅₀in

the range 0.8–10

pounds had EC₅₀between 11 and 30

arylazoenamine molecular pattern is an interesting novel pharma-

A ﬁnal evaluation of the statistical relevance of the 3D pharma-

cophore models, the Fisher method was applied using the Cat-

Scramble program available in the CATALYST suite of programs.

According to the validation procedure, the experimental activity

of the compounds in each training test was scrambled randomly,

and the resulting new training set was used for a new HypoGen

run. The parameters adopted in running these calculations were

the same employed in the initial HypoGen calculations and, since

a 98% conﬁdence level was selected, 49 random hypotheses were

generated. The resulting data clearly indicate that all values gener-

ated after randomization produced hypotheses with no predictive

values similar or close to the corresponding Hypo1s. Indeed, none

of the outcome hypotheses has lower cost score, better correlation

or smaller root-mean-square deviation than the initial one. Table 9

lists the ﬁrst 10 lowest total score values of the resulting 49

hypotheses for BVDV, CVB-2, and RSV, respectively. In conclusion,

there is a 98% chance for the best hypothesis to represent a true

correlation in the training set activity data for each antiviral class

of compounds.

l

M, against at least one virus and other 28 com-

lM, thus indicating that the

cophore for antiviral agents against ssRNA viruses.

Some interesting correlations between the structure and the

type of affected virus were shown. The arylazohexahydroquinoli-

zines A and the aryazoethenamines F appear as the most suitable

structures for activity on CVB-2. The arylazotetrahydropyridines

C and the arylazomethylenepyrrolidines D share in almost equal

measure the activity on CVB-2 and BVDV, while the RSV was

8458

M. Tonelli et al. / Bioorg. Med. Chem. 16 (2008) 8447–8465

Figure 3. Top scoring HypoGen pharmacophore for BVDV (top, left), CVB-2 (top, right), and RSV (bottom, center). Hypothesis features are color-coded as follows: cyan,

hydrophobic aromatic (HAr); green, hydrogen bond acceptor (HBA), dark gold, aromatic ring (RAr). The distances (Å) between all features in the corresponding hypotheses are

also shown.

Figure 4. Mapping of the three most active compounds of the training sets onto the corresponding top-ranked 3D pharmacophore models (top, left), compound 32 onto

Hypo1_BVDV; (top, right), compound 75 onto Hypo1_CVB-2; (bottom, center) compound 84 onto Hypo1_RSV. Hypothesis features are color-coded as follows: cyan,

hydrophobic aromatic (HAr); green, hydrogen bond acceptor (HBA), light blue/gray, aromatic ring (RAr).

Table 7

Results for the best pharmacophore hypotheses (Hypo1) generated by CATALYST HypoGen using the training sets for BVDV, CVB-2, and RSV

Hypotheses

Total ﬁxed cost

Total Hypo1 cost

Null cost

Conﬁgurational cost

RMSD^a

q

Features

Hypo1_BVDV

Hypo1_CVB-2

Hypo1_RSV

93.89

89.62

103.01

102.48

107.92

121.15

180.17

173.57

201.00

12.78

13.56

14.23

0.50

0.42

0.38

0.98

0.97

0.95

HBA, HBA, HAr

HBA, RA

All cost units are in bits.

a

Deﬁned in terms of the deviation of the log(estimated activity), from the log(experimental activity), normalized by the log(uncertainty).

M. Tonelli et al. / Bioorg. Med. Chem. 16 (2008) 8447–8465

8459

Table 8

Estimated and experimentally determined activity values of the test set compounds for BVDV, CVB-2, and RSV

BVDV

EC_50,exp.

M)

CVB-2

EC_50,exp.

M)

RSV

EC_50,exp.

M)

Test set

compounds

EC_50,est

M)

Error^a

Test set

compounds

EC_50,est

M)

Error^a

Test set

compounds

EC_50,est.

(lM)

Error^a

(

l

(

l

(

l

(

l

(

l

16

35

44

48

54

29

46

57

26

6

7

8

9

1.3

1.9

67

78

74

81

5

44

7

3

4

5

6

7

3.5

ꢀ1.1

1.3

ꢀ1.1

1.0

80

85

81

74

78

3

29

28

46

25

2

5

6

7

9

18

20

30

60

75

4.9

7

8.5

6

12

16

20

19

24

51

64

ꢀ1.0

1.2

6.4

5.5

6.6

10

11

16

19

21

46

100

19

13

14

13

25

40

94

1.2

ꢀ1.2

ꢀ1.1

ꢀ1.5

1.2

18

19

21

22

26

30

34

35

38

40

41

71

18

21

19

28

25

28

29

38

28

33

44

38

62

1.3

1.1

ꢀ1.1

1.1

ꢀ1.1

ꢀ1.2

1.3

ꢀ1.1

9

ꢀ1.0

ꢀ1.1

ꢀ1.0

1.1

ꢀ1.3

ꢀ1.2

1.1

1.2

20

63

27

70

10

39

11

56

ꢀ1.3

ꢀ1.2

24

ꢀ1.1

a

Values in the error column represent the ratio of the estimated activity to measured activity, or its negative inverse if the ratio is less than one.

Table 9

Output parameters of the 10 lowest cost hypotheses resulting from the statistical evaluation according to the Cat-Scramble validation procedure for BVDV, CVB-2, and RSV

BVDV CVB-2 RSV

Hypothesis

q

RMSD

Total cost

Hypothesis

1

q

RMSD

Total cost

Hypothesis

1

q

RMSD

Total cost

1

2

3

4

5

6

7

8

0.77

0.76

0.74

0.73

0.71

0.69

0.66

0.64

0.98

0.75

0.78

0.79

0.80

0.81

0.84

0.88

0.92

1.01

1.11

0.50

131.66

133.14

133.75

135.22

137.29

140.01

141.56

142.81

144.35

145.22

102.48

0.84

0.81

0.79

0.78

0.75

0.73

0.72

0.97

0.91

0.93

1.02

1.01

1.05

1.13

1.21

1.24

1.25

1.26

0.42

158.04

159.51

181.09

183.63

184.64

190.07

192.55

193.78

194.16

198.23

107.92

0.74

0.72

0.68

0.66

0.63

0.61

0.57

0.55

0.52

0.47

0.95

0.68

0.75

0.84

0.90

0.96

0.99

1.05

1.08

1.14

1.21

0.38

155.35

157.87

161.69

168.21

169.97

170.30

174.08

174.53

178.04

181.69

121.15

2

3

4

5

6

7

8

10

2

3

4

5

6

7

8

10

9

10

Hypo1

mainly affected by arylazoethenepiperazines G. Many compounds

exhibited a selectivity index (CC₅₀for host cells/EC₅₀for a given

virus) in the range from >10 to >125.

5. Experimental

5.1. General

Taking into account the EC₅₀and SI values, the best compounds

can be arranged in the following order of decreasing potency and

Chemicals, solvents, and reagents used for syntheses were pur-

chased from Sigma–Aldrich, Fluka, or Lancaster, and were used

without any further puriﬁcation, unless otherwise speciﬁed. Not

commercially available arylhydrazines, as 4-F-3-CF₃,³⁷4-Cl-3-

CF₃,³⁸4-Br-3-CF₃,³⁹4-ethoxycarbonyl-phenylhydrazine,⁴⁰and 1-

naphthylhydrazine,⁴⁰were prepared according to Enders.⁴⁰

Column chromatography (CC): neutral alumina (Al₂O₃), activity

1 (Merck). Mps: Büchi apparatus, uncorrected. ¹H NMR spectra:

Varian Gemini-200 spectrometer; CDCl₃; d in ppm rel to Me₄Si as

internal standard. J in Hz; Ind = hexahydroindolizine, Pip = piperi-

dine, Py = tetrahydropyridine, Pyrr = pyrrolidine, Pz = piperazine,

Q = hexahydroquinolizine. Elemental analyses were performed on

a Carlo Erba EA-1110 CHNS-O instrument in the Microanalysis Lab-

oratory of the Department of Pharmaceutical Sciences of Genoa

University.

selectivity;

for

CVB-2:

75 > 66 P 53 P 28;

for

BVDV:

32 > 22 > 42 = 50 > 16; for RSV: 84 > 23, and for Sb-1: 26 P 28. In

these sequences, even the least potent compounds still compare

well with the reference drugs NM-108 and 6-azauridine.

Compounds 53, 28, 32, and 42 appear of particular interest

being also devoid of toxicity on the human MT-4 cells

(CC₅₀> 100 lM).

A ligand-based computational approach was employed to iden-

tify molecular structure requirements in the available series of

compounds to act as effective viral inhibitors. Three highly predic-

tive pharmacophore models were generated based on 18, 43, and

21 training set compounds for BVDV, CVB-2, and RSV, respectively,

which consist of two hydrogen bond donors and one hydrophobic

aromatic in the case of BVDV (Hypo1_BVDV), and one hydrogen

bond donor and one ring aromatic for the two other viral species

(Hypo1_CVB-2 and Hypo1_RSV, respectively). The application of

our 3D pharmacophore models to three test sets of different com-

pounds showed that each model is able to accurately predict the

activity of the compounds toward the respective virus. Since the

target proteins for the antiviral activities of our series of com-

pounds is as yet unknown, our three-dimensional QSAR pharmaco-

phore models should constitute useful tools in the design of second

generations of different, more potent inhibitors of these human

and veterinary pathogens.

5.2. Aryl/heteroarylazoenamines of structures A, B, C, and D (1–

63): general method

To a solution of a-aminoketone (3 mmol) {hexahydro-2H-quin-

olizin-1-one,³⁴hexahydro-8(5H)-indolizinone,³⁵1-Methyl-3-pip-

eridone³⁶} in 15 mL of abs. ethanol were added, in the following

order, aryl/heteroarylhydrazine hydrochloride (6 mmol) or the cor-

responding free base, 0.75 mL of conc. hydrochloric acid, and

0.9 mL of 85% phosphoric acid. The mixture was reﬂuxed with stir-

8460

M. Tonelli et al. / Bioorg. Med. Chem. 16 (2008) 8447–8465

ring for 3 h. After cooling, ethanol was evaporated at reduced pres-

sure, water was added, and the acid solution was basiﬁed with 2 M

NaOH and extracted with Et₂O. After drying, the solvent was re-

moved obtaining a red-orange oil from which the volatile sub-

stances were removed by heating until 80–100 °C under vacuum

(0.2 torr). The residue was puriﬁed by CC (CH₂Cl₂).

The low (C) and high (D) melting point isomers, obtained in the

reaction with 1-methyl-3-piperidone, were separated by eluting

ﬁrstly with CH₂Cl₂and then CH₂Cl₂+ 2%Et₂NH.

3.15–3.30 (m, 6H, C(2, 4, 6) of Q); 7.32–7.78 (m, 5 arom. H);

8.74–8.84 (m, 1 arom. H). Anal. calcd for C₁₉H₂₁N₃: C 78.32, H

7.26, N, 14.42; found: C 78.67, H 7.03, N 14.02.

5.2.8. 1-[4-(7-Chloro)quinolylazo]-3,4,6,7,8,9-hexahydro-2H-

quinolizine (17)

Yield: 58%. CC (CH₂Cl₂). Mp 203–206 °C (Et₂O). ¹H NMR (CDCl₃):

1.74–2.14 (m, 6H, C(3, 7, 8) of Q); 2.89 (t, J = 6.4, 2H, C(9) of Q);

3.29–3.50 (m, 6H, C(2, 4, 6) of Q); 7.35–8.84 (m, 5 arom. H). Anal.

calcd for C₁₈H₁₉ClN₄: C 66.15, H 5.86, N 17.14; found: C 66.04, H

6.04, N 17.09.

In the case of phthalazylazoenamine 23, the reaction mixture

was neutralized and then shaken with CH₂Cl₂which extracted

the compound as hydrochloride.

Compounds 1–4, 6, 9, 10, 13, 15, 24, 26–28, 32-35, 38, 40, 44,

5.2.9. 8-(4-Chlorophenylazo)-1,2,3,5,6,7-hexahydroindolizine

(18)

46, and 51–54 have been already described.^2,8,9

Yield: 25%. Mp 165–167 °C (Et₂O). ¹H NMR (CDCl₃): 1.95–2.20

(m, 4H, C(2, 6) of Ind); 2.65 (t, J = 6.1, 2H, C(7) of Ind); 3.22 (t,

J = 7.8, 2H, C(1) of Ind); 3.33 (t, J = 6.0, 2H, C(5) of Ind); 3.55 (t,

J = 6.6, 2H, C(3) of Ind); 7.25 (d, J = 2.2, 2 arom. H); 7.50 (d,

J = 2.2, 2 arom. H). Anal. calcd for C₁₄H₁₆ClN₃: C 64.24, H 6.16, N

16,05: found: C 64.45, H 6.20, N 16.19.

5.2.1. 1-(3-Bromophenylazo)-3,4,6,7,8,9-hexahydro-2H-

quinolizine (5)

Yield:42%. Mp 115–117 °C (Et₂O). ¹H NMR(CDCl₃): 1.61–1.88(m,

6H, C(3, 7, 8) of Q); 2.59 (t, J = 6.5, 2H, C(9) of Q); 3.08–3.36 (m, 6H,

C(2, 4, 6) of Q); 7.03–7.63 (m, 4 arom. H). Anal. calcd for C₁₅H₁₈BrN₃:

C 56.26, H 5.67, N 13.12; found: C 56.45, H 5.69, N 13.09.

5.2.10. 8-(3-Nitrophenylazo)-1,2,3,5,6,7-hexahydroindolizine

(19)

5.2.2. 1-(3-Triﬂuoromethylphenylazo)-3,4,6,7,8,9-hexahydro-

2H-quinolizine (7)

Yield: 43%. Mp 157–159 °C (Et₂O). ¹H NMR (CDCl₃): 1.98–2.22

(m, 4H, C(2, 6) of Ind); 2.68 (t, J = 6.2, 2H, C(7) of Ind); 3.29 (t,

J = 7.9, 2H, C(1) of Ind); 3.38 (t, J = 6.0, 2H, C(5) of Ind); 3.64 (t,

J = 6.6, 2H, C(3) of Ind); 7.37–8.40 (m, 4 arom. H). Anal. calcd for

Yield: 25%. Mp 102–103 °C (Et₂O). ¹H NMR (CDCl₃): 1.60–1.97

(m, 6H, C(3, 7, 8) of Q); 2.63 (t, J = 6.4, 2H, C(9) of Q); 3.07–3.33

(m, 6H, C(2, 4, 6) of Q); 7.22–7.77 (m, 4 arom. H). Anal. calcd for

C₁₆H₁₈F₃N₃: C 62.13, H 5.87, N 13.58; found: C 62.00, H 6.04, N

C₁₄H₁₆N₄O₂: C 61.75, H 5.92. N 20.58; found: C 61.78, H 6.04, N

13.23.

20.45.

5.2.3. 1-(3-Nitrophenylazo)-3,4,6,7,8,9-hexahydro-2H-

quinolizine (8)

5.2.11. 8-(3,4-Dichlorophenylazo)-1,2,3,5,6,7-

hexahydroindolizine (20)

Yield: 46%. CC (Et₂O). Mp 132–135°C (Et₂O). ¹H NMR (CDCl₃):

1.63–2.00 (m, 6H, C(3, 7, 8) of Q); 2.63 (t, J = 6.4, 2H, C(9) of Q);

3.12–3.37 (m, 6H, C(2, 4, 6) of Q); 7.30–8.32 (m, 4 arom. H). Anal.

calcd for C₁₅H₁₈N₄O₂: C 62.92, H 6.34; N 19.57; found: C 62.61, H

6.39, N 19.58.

Yield: 30%. CC (Et₂O). Mp 176–177 °C (Et₂O). ¹H NMR (CDCl₃):

1.87–2.11 (m, 4H, C(2, 6) of Ind); 2.59 (t, J = 6.4, 2H, C(7) of Ind);

3.17 (t, J = 7.8, 2H, C(1) of Ind); 3.29 (t, J = 5.9, 2H, C(5) of Ind);

3.54 (t, J = 6.6, 2H, C(3) of Ind); 7.18–7.34 (m, 2 arom. H); 7.58 (d,

J = 2.0, 1 arom. H). Anal. calcd for C₁₄H₁₅Cl₂N₃: C 56.77, H 5.10, N

14.19; found: C 56.43, H 5.08, N 14.00.

5.2.4. 1-(3,5-Bis-triﬂuoromethylphenylazo)-3,4,6,7,8,9-

hexahydro-2H-quinolizine (11)

Yield: 72%. Mp 147–149 °C (Et₂O). ¹H NMR (CDCl₃): 1.64–1.99

(m, 6H, C(3, 7, 8) of Q); 2.63 (t, J = 6.5, 2H, C(9) of Q); 3.12–3.36

(m, 6H, C(2, 4, 6) of Q); 7.40–7.90 (m, 3 arom. H). Anal. calcd for

5.2.12. 8-(4-Chloro-3-triﬂuoromethylphenylazo)-1,2,3,5,6,7-

hexahydroindolizine (21)

Yield: 81%. Mp 161–162.5 °C (Et₂O). ¹H NMR (CDCl₃): 1.88–2.12

(m, 4H, C(2, 6) of Ind); 2.59 (t, J = 6.4, 2H, C(7) of Ind); 3.17 (t,

J = 7.9, 2H, C(1) of Ind); 3.29 (t, J = 5.8, 2H, C(5) of Ind); 3.54 (t,

J = 7.1, 2H, C(3) of Ind); 7.32 (d, J = 8.6, 1 arom. H); 7.56 (dd,

J = 8.6, 2.0, 1 arom. H); 7.82 (d, J = 2.4, 1 arom. H). Anal. calcd for

C₁₇H₁₇F₆N₃: C 54,11, H 4.54, N 11.11; found: C 54.00, H 4.72, N

11.02.

5.2.5. 1-(4-Fluoro-3-triﬂuoromethylphenylazo)-3,4,6,7,8,9-

hexahydro-2H-quinolizine (12)

C₁₅H₁₅ClF₃N₃: C 54.64, H 4.59, N 12.74; found: C 54.47, H 4.57, N

Yield: 88%. Mp 133–134 °C (Et₂O). ¹H NMR (CDCl₃): 1.60–1.98

(m, 6H, C(3, 7, 8) of Q); 2.58 (t, J = 6.5, 2H, C(9) of Q); 3.07–3.34

(m, 6H, C(2, 4, 6) of Q); 6.98–7.80 (m, 3 arom. H). Anal. calcd for

12.69.

5.2.13. 8-(1-Naphthylazo)-1,2,3,5,6,7-hexahydroindolizine (22)

Yield: 32%. CC (CH₂Cl₂). Mp 148–151 °C (CH₂Cl₂). ¹H NMR

(CDCl₃): 1.92–2.10 (m, 4H, C(2, 6) of Ind); 2.77 (t, J = 6.4, 2H, C(7)

of Ind); 3.17–3.30 (m, 4H, C(1, 5) of Ind); 3.46 (t, J = 6.9, 2H, C(3)

of Ind); 7.33–7.78 (m, 6 arom. H); 8.78–8.85 (m, 1 arom. H). Anal.

calcd for C₁₈H₁₉N₃: C 77.95, H 6.90, N 15.15; found: C 78.01, H 6.94,

15.13.

C

₁₆H₁₇F₄N₃: C 58.71, H 5.23, N 12.84; found: C 58.51, H 5.43, N

12.59.

5.2.6. 1-(4-Bromo-3-triﬂuoromethylphenylazo)-3,4,6,7,8,9-

hexahydro-2H-quinolizine (14)

Yield: 37%. CC (CH₂Cl₂). Mp 155–157 °C (Et₂O). ¹H NMR (CDCl₃):

1.62–1.98 (m, 6H, C(3, 7, 8) of Q); 2.60 (t, J = 6.5, 2H, C(9) of Q);

3.07–3.35 (m, 6H, C(2, 4, 6) of Q); 7.40–7.84 (m, 3 arom. H). Anal.

calcd for C₁₆H₁₇BrF₃N₃: C 49.50, H 4.41, N 10.82; found: C 49.53, H

4.32, N 10.74.

5.2.14. 8-(1-Phthalazinylazo)-1,2,3,5,6,7-hexahydroindolizine

hydrochloride (23)

Yield: 49%. Mp 174–176 °C (Et₂O). ¹H NMR (CDCl₃): 1.90–2.10

(m, 4H, C(2, 6) of Ind); 2.72 (t, J = 6.2, 1H, C(7) of Ind); 2.93 (t,

J = 6.2, 1H, C(7) of Ind); 3.08 (t, J = 7.0, 2H, C(1) of Ind); 3.73–3.86

(m, 2H, C(5) of Ind); 4.10–4.28 (m, 2H, C(3) of Ind); 7.88–8.72

(m, 4 arom. H); 9.10 (d, J = 8.0, 1 arom. H). Anal. calcd for

5.2.7. 1-(1-Naphthylazo)-3,4,6,7,8,9-hexahydro-2H-quinolizine

(16)

Yield: 58%. CC (CH₂Cl₂). Mp 153–155 °C (Et₂O). ¹H NMR (CDCl₃):

1.63–2.00 (m, 6H, C(3, 7, 8) of Q); 2.79 (t, J = 6.4, 2H, C(9) of Q);

M. Tonelli et al. / Bioorg. Med. Chem. 16 (2008) 8447–8465

8461

C₁₆H₁₇N₅+ HCl+H₂O: C 57.57, H 6.04, N 20.97; found: C 57.39, H

6.25, 20.25.

(t, J = 6.0, 2H, C(6) of Py); 7.27 (s, 1H, C(2) of Py). Anal. calcd for

₁₂H₁₀F₅N₃: C 49.49, H 3.46, N 14.43; found: C 49.51, H 3.32, N 14.00.

C

5.2.23. 1-Methyl-3-(1-naphthylazo)-1,4,5,6-tetrahydropyridine

(42)

5.2.15. 3-(4-Fluorophenylazo)-1-methyl-1,4,5,6-

tetrahydropyridine (25)

Yield: 42%. Mp 116–119 °C (Et₂O). ¹H NMR (CDCl₃): 2.08 (q, 2H,

C(5) of Py); 2.82 (t, J = 6.7, 2H, C(4) of Py); 3.14 (s, CH₃N); 3.29 (t,

J = 5.7, 2H, C(6) of Py); 7.35 (s, 1H, C(2) of Py); 7.44–7.90 (m, 6

arom. H); 8.79–8.87 (m, 1 arom. H). Anal. calcd for C₁₆H₁₇N₃: C

76.46, H 6.82, N 16.72; found: C 76.46, H 6.74, N 16.39.

Yield: 25%. Mp 82–84 °C (pentane). ¹H NMR (CDCl₃): 1.98 (q,

2H, C(5) of Py); 2.60 (t, J = 6.4, 2H, C(4) of Py); 3.06 (s, CH₃N);

3.22 (t, J = 5.8, 2H, C(6) of Py); 6.97–7.08 (m, 2 arom. H); 7.20 (s,

1H, C(2) of Py); 7.49–7.59 (m, 2 arom. H). Anal. calcd for

C₁₂H₁₄FN₃: C 65.73, H 6.44, N 19.16; found. C 65.52, H 6.15, N

19.28.

5.2.24. 1-Benzyl-3-(4-chloro-3-triﬂuoromethylphenylazo)-

1,4,5,6-tetrahydropyridine (43)

5.2.16. 3-(3-Bromophenylazo)-1-methyl-1,4,5,6-

tetrahydropyridine (29)

Yield: 27%. Mp 118–119 °C (Et₂O). ¹H NMR (CDCl₃): 1.96 (q, 2H,

C(5) of Py); 2.64 (t, J = 5.8, 2H, C(4) of Py); 3.23 (t, J = 5.7, 2H, C(6) of

Py); 4.48 (s, 2H, CH₂C₆H₅); 7.22–8.02 (m, 8 arom. H and 1H, C(2) of

Py). Anal. calcd for C₁₉H₁₇ClF₃N₃: C, 60.08, H 4.51, N 11.06; found:

C 59.74, H 4.47, N 10.95.

Yield: 20%. Mp 96–99 °C (Et₂O). ¹H NMR (CDCl₃): 1.99 (q, 2H,

C(5) of Py); 2.61 (t, J = 6.4, 2H, C(4) of Py); 3.12 (s, CH₃N); 3.26 (t,

J = 5.8, 2H, C(6) of Py); 7.20 (s, 1H, C(2) of Py); 7.24–7.77 (m, 4

arom. H). Anal. calcd for C₁₂H₁₄BrN₃: C 51.44, H 5.04, N 15.00;

found: C 51.59, H 5.07, N 14.82.

5.2.25. 2-(4-Fluorophenylazo)methylene-1-methylpyrrolidine

(45)

Yield: 23%. Mp 107–108 °C (pentane). ¹H NMR (CDCl₃): 2.07 (q,

2H, C(4) of Pyrr); 2.98 (s, CH₃N); 3.27 (t, J = 7.7, 2H, C(3) of Pyrr);

3.57 (t, J = 7.0, 2H, C(5) of Pyrr); 7.98–8.12 (m, 1H, @CH– and 2

arom. H); 7.50–7.61 (m, 2 arom. H). Anal. calcd for C₁₂H₁₄FN₃: C

65.73, H 6.44, N 19.16; found: C 65.60, H 6.12, N 19.31.

5.2.17. 3-(4-Bromophenylazo)-1-methyl-1,4,5,6-

tetrahydropyridine (30)

Yield: 22%. Mp 112–114 °C (Et₂O). ¹H NMR (CDCl3): 1.92 (q, 2H,

C(5) of Py); 2.54 (t, J = 6.4, 2H, C(4) of Py); 3.04 (s, CH₃N); 3.18 (t,

J = 5.8, 2H, C(6) of Py); 7.17 (s, 1H, C(2) of Py); 7.34–7.42 (m, 4

arom. H). Anal. calcd for C₁₂H₁₄BrN₃: C 51.44, H 5.04, N 15.00;

found: C 51. 40, H 4.52, N 14.78.

5.2.26. 2-(3-Bromophenylazo)methylene-1-methylpyrrolidine

(47)

Yield: 22%. Mp 103–107 °C (Et₂O). ¹H NMR (CDCl₃): 2.08 (q, 2H,

C(4) of Pyrr); 3.01 (s, CH₃N); 3.30 (t, J = 7.8, 2H, C(3) of Pyrr); 3.60

(t, J = 7.0, 2H, C(5) of Pyrr); 7.14 (s, 1H, @CH–); 7.20–7.76 (m, 4

arom. H). Anal. calcd for C₁₂H₁₄BrN₃: C 51.44, H 5.04, N 15.00;

found: C 51.41, H 4.98, N 14.87.

5.2.18. 1-Methyl-3-(3-triﬂuoromethylphenylazo)-1,4,5,6-

tetrahydropyridine (31)

Yield: 20%. Mp 76–77 °C (pentane). ¹H NMR (CDCl₃): 1.93 (q,

2H, C(5) of Py); 2.57 (t, J = 6.5, 2H, C(4) of Py); 3.06 (s, CH₃N);

3.20 (t, J = 5.8, 2H, C(6) of Py); 7.23 (s, 1H, C(2) of Py); 7.27–7.81

(m, 4 arom. H). Anal. calcd for C₁₃H₁₄F₃N₃: C 57.99, H 5.24, N

15.61; found: 57.65, H 5.38, N 15.75.

5.2.27. 2-(4-Bromophenylazo)methylene-1-methylpyrrolidine

(48)

Yield: 33%. Mp 150–152.5 °C (Et₂O). ¹H NMR (CDCl₃): 2.02 (q,

2H, C(4) of Pyrr); 2.95 (s, CH₃N); 3.20 (t, J = 7.8, 2H, C(3) of Pyrr);

3.53 (t, J = 7.0, 2H, C(5) of Pyrr); 7.09 (s, 1H, @CH–); 7.34–7.42

(m, 4 arom. H). Anal. calcd for C₁₂H₁₄BrN₃: C 51.44, H 5.04, N

15.00; found: C 51.07, H 5.17, N 14.80.

5.2.19. 3-(3,5-Bis-triﬂuoromethylphenylazo)-1-methyl-1,4,5,6-

tetrahydropyridine (36)

Yield: 20%. Mp 86–87 °C (pentane). ¹H NMR (CDCl₃): 2.02 (q,

2H, C(5) of Py); 2.64 (t, J = 6.7, 2H, C(4) of Py); 3.18 (s, CH₃N);

3.31 (t, J = 5.8, 2H, C(6) of Py); 7.38 (s, 1H, C(2) of Py); 7.55–8.03

(m, 3 arom. H). Anal. calcd for C₁₄H₁₃F₆N₃: C 49.86, H 3.89, N

12.46; found: C 49.91, H 3.60, N 12.30.

5.2.28. 1-Methyl-2-(3-triﬂuoromethylphenylazo)methylenepy-

rrolidine (49)

Yield: 15%. Mp 107–108 °C (Et₂O). ¹H NMR (CDCl₃): 2.10 (q, 2H,

C(4) of Pyrr); 3.03 (s, CH₃N); 3.32 (t, J = 7.9, 2H, C(3) of Pyrr); 3.62

(t, J = 7.0, 2H, C(5) of Pyrr); 7.18 (s, 1H, @CH–); 7.33–7.86 (m, 4

arom. H). Anal. calcd for C₁₃H₁₄F₃N₃: C 57.99, H 5.24, N 15.61;

found: C 58.11, H 5.55, N 15.33.

5.2.20. 3-(4-Fluoro-3-triﬂuoromethylphenylazo)-1-methyl-

1,4,5,6-tetrahydropyridine (37)

Yield: 30%. Mp 81–84 °C (pentane). ¹H NMR (CDCl₃): 2.00 (q,

2H, C(5) of Py); 2.61 (t, J = 6.7, 2H, C(4) of Py); 3.13 (s, CH₃N);

3.26 (t, J = 5.9, 2H, C(6) of Py); 7.10–7.87 (m, 1H, C(2) of Py and 3

arom. H). Anal. calcd for C₁₃H₁₃F₄N₃: C 54.36, H 4.56, N 14.63;

found: C 54.03, H 4.22, N 13.67.

5.2.29. 1-Methyl-2-(3-nitrophenylazo)methylenepyrrolidine

(50)

Yield: 30%. Mp 117–119 °C (Et₂O). ¹H NMR (CDCl₃): 2.06 (q, 2H,

C(4) of Pyrr); 3.00 (s, CH₃N); 3.27 (t, J = 7.8, 2H, C(3) of Pyrr); 3.60

(t, J = 7.0, 2H, C(5) of Pyrr); 7.18–8.32 (m, 1H, @CH– and 4 arom. H).

Anal. calcd For C₁₂H₁₄N₄O₂: C 58.53, H 5.73, N 22.75; found: C

58.44, H 5.89, N 22.73.

5.2.21. 3-(4-Bromo-3-triﬂuoromethylphenylazo)-1-methyl-

1,4,5,6-tetrahydropyridine (39)

Yield: 34%. Mp 94–95 °C (Et₂O). ¹H NMR (CDCl₃): 2.01 (q, 2H,

C(5) of Py); 2.63 (t, J = 6.4, 2H, C(4) of Py); 3.16 (s, CH₃N); 3.30 (t,

J = 5.7, 2H, C(6) of Py); 7.33 (s, 1H, C(2) of Py); 7.59–7.93 (m, 3

arom. H). Anal. calcd for C₁₃H₁₃BrF₃N₃: C 44.85, H 3.76, N, 12.07,

found: C 44.95, H 4.06 N 11.76.

5.2.30. 1-Methyl-2-(4-nitrophenylazo)methylenepyrrolidine

and 1-Methyl-3-(4-nitrophenylazo)-1,4,5,6-tetrahydropyridine

(mixture 2:1) (51)

Yield: 60%. Mp 180 °C (EtOH). ¹H NMR (CDCl₃): 1.98 (q,

0.33 ꢂ 2H, C(5) of Py); 2.14 (q, 0.67 ꢂ 2H, C(4) of Pyrr); 2.63 (t,

J = 6.4, 0.33 ꢂ 2H, C(4) of Py); 3.10 (s, 0.67 ꢂ 3H, NCH₃of Pyrr);

3.18 (s, 0.33 ꢂ 3H, NCH₃of Py); 3.22–3.37 (m, 0.67 ꢂ 2H, C(3) of

5.2.22. 1-Methyl-3-pentaﬂuorophenylazo-1,4,5,6-

tetrahydropyridine (41)

Yield: 24%. Mp 100–101 °C (pentane). ¹H NMR (CDCl₃): 1.95 (q,

2H, C(5) of Py); 2.55 (t, J = 6.5, 2H, C(4) of Py); 3.11 (s, CH₃N); 3.25

8462

M. Tonelli et al. / Bioorg. Med. Chem. 16 (2008) 8447–8465

Pyrr) and 0.33 ꢂ 2H, C(6) of Py); 3.70 (t, J = 7.0, 0.67 ꢂ 2H, C(5) of

Pyrr); 7.36–7.40 (m, 1H, @CH– of Pyrr and Py); 7.58 (d, J = 9.0, 4

arom. H); 8.16 (d, J = 9.2, 4 arom. H). Anal. calcd for C₁₂H₁₄N₄O₂:

C 58.52, H 5.73, N 22.75; found: C 58.75, H 5.72, N 22.87.

(t, J = 7.7, 0.75 ꢂ 2H, C(3) of Pyrr); 3.59 (t, J = 7.1, 0.75 ꢂ 2H, C(5)

of Pyrr); 4.45 (s, 0.25 ꢂ 2H, CH₂C₆H₅of Py); 4.54 (s, 0.75 ꢂ 2H,

CH₂C₆H₅of Pyrr); 7.20–7.70 (m, 1H, @CH– of Pyrr and Py and 8

arom. H). Anal. calcd for C₁₈H₁₇Cl₂N₃: C 62.43, H 4.95, N 12.13;

found: C 62.68, H 4.96, N 12.16.

5.2.31. 2-(3,5-Bis-triﬂuoromethylphenylazo)methylene-1-

methylpyrrolidine (55)

5.2.39. 1-Benzyl-2-(4-chloro-3-triﬂuoromethylphenylazo)-

methylenepyrrolidine (63)

Yield: 15%. Mp 128–130 °C (Et₂O). ¹H NMR (CDCl₃): 2.07(q, 2H,

C(4) of Pyrr); 3.01 (s, CH₃N); 3.28 (t, J = 7.8, 2H, C(3) of Pyrr); 3.61

(t, J = 7.0, 2H, C(5) of Pyrr); 7.19 (s, 1H, @CH–); 7.45–7.70 (m, 3

arom. H). Anal. calcd for C₁₄H₁₃F₆N₃: C 49.86, H 3.89, N 12.46;

found: C 50.42, H 3.56, N 12.35.

Yield: 37%. Mp 114–116 °C (Et₂O). ¹H NMR (CDCl₃): 2.12 (q, 2H,

C(4) of Pyrr); 3.38 (t, J = 7.8, 2H, C(3) of Pyrr); 3.62 (t, J = 7.1, 2H,

C(5) of Pyrr); 4.47 (s, 2H, CH₂C₆H₅); 7.15–7.98 (m, 1H, @CH– and

8 arom. H). Anal. calcd for C₁₉H₁₇ClF₃N₃: C 60.08, H 4.51, N

11.06; found: C 59.95, H 4.49, N 10.95.

5.2.32. 2-(4-Fluoro-3-triﬂuoromethylphenylazo)methylene-1-

methylpyrrolidine (56)

5.3. 2-(Arylazomethylene)-1,3,3-trimethylindoline (64–65)

Yield: 18%. Mp 120–123 °C (Et₂O). ¹H NMR (CDCl₃): 2.03 (q, 2H,

C(4) of Pyrr); 2.96 (s, CH₃N); 3.23 (t, J = 7.8, 2H, C(3) of Pyrr); 3.56

(t, J = 7.0, 2H, C(5) of Pyrr); 7.00–7.75 (m, 1H, @CH– and 3 arom. H).

Anal. calcd for C₁₃H₁₃F₄N₃: C 54.36, H 4.56, N 14.63; found: C

54.34, H 4.56, N 14.60.

The title compounds were prepared through the method used by

König and Muller⁴with minor modiﬁcation. A solution of aniline or

4-chloroaniline (15 mmol) in 6 mL of conc. hydrochloridric acid and

15 mL of water was cooled in an ice bath and a solution of sodium ni-

trite (15 mmol) in 3.5 mL of cold water was added dropwise. The

resulting solution of diazonium salt was then poured slowly into a

stirred emulsion of 1,3,3-trimethyl-2-methyleneindoline⁴¹in

15 mL of water and subsequently added of sodium acetate (10 g).

After 1 h at 0 °C with stirring, the aqueous phase was extracted with

CH₂Cl₂, theorganicphasewasdriedwithsodiumsulfateandconcen-

trated to dryness, obtaining, in the case of 64, an oil which crystal-

lized when treated with the equivalent amount of 1N perchloric

acid in abs. ethanol. Compound 65 was a crystalline solid.³²

5.2.33. 2-(4-Bromo-3-triﬂuoromethylphenylazo)methylene-1-

methylpyrrolidine (57)

Yield: 43%. Mp 126–128 °C (Et₂O). ¹H NMR (CDCl₃): 2.12 (q, 2H,

C(4) of Pyrr); 3.05 (s, CH₃N); 3.31 (t, J = 7.9, 2H, C(3) of Pyrr); 3.65

(t, J = 7.1, 2H, C(5) of Pyrr); 7.19 (s, 1H, @CH–); 7.50–7.90 (m, 3

arom. H). Anal. calcd for C₁₃H₁₃BrF₃N₃: C 44.85, H 3.76, N 12.07;

found: C 44.76, H 3.89, N 11.89.

5.2.34. 1-Methyl-2-(pentaﬂuorophenylazo)methylenepyrroli-

dine (58)

5.3.1. 2-(Phenylazomethylene)-1,3,3-trimethylindoline

perchlorate (64)

Yield: 64%. Mp 239–240 °C. Anal. calcd for C₁₈H₁₉N₃+HClO₄: C

57.22, H 5.34, N 11.12; found: C 57.30, H 5.29, N 11.12. Already de-

scribed as hydroiodide,³hydrochloride,⁴and free base,^3,42which,

however, in our hands did not crystallize well.

Yield: 15%. Mp 154–155 °C (pentane). ¹H NMR (CDCl₃): 2.10 (q,

2H, C(4) of Pyrr); 3.07 (s, CH₃N); 3.24 (t, J = 7.8, 2H, C(3) of Pyrr);

3.67 (t, J = 7.1, 2H, C(5) of Pyrr); 7.23 (s, 1H, @CH–). Anal. calcd

for C₁₂H₁₀F₅N₃: C 49.49, H 3.46, N 14.43; found: C 49.58, H 3.22,

N 14.18.

5.2.35. 1-Benzyl-2-phenylazomethylenepyrrolidine (59)

Yield: 43%. Mp 121–123 °C (Et₂O). ¹H NMR (CDCl₃): 2.08 (q, 2H,

C(4) of Pyrr); 3.38 (t, J = 7.7, 2H, C(3) of Pyrr); 3.56 (t, J = 7.0, 2H,

C(5) of Pyrr); 4.53 (s, 2H, CH₂C₆H₅); 7.03–7.73 (m, 1H, @CH– and

10 arom. H). Anal. calcd for C₁₈H₁₉N₃: C 77.95, H 6.90, N 15.15;

found: C 77,74, H 6.89, N 15.15.

5.3.2. 2-(4-Chlorophenylazomethylene)-1,3,3-trimethyl-

indoline (65)

Yield: 50%. Mp 132–134 °C (MeOH) lit.³²141.5 °C. ¹H NMR

(CDCl₃): 1.76 (s, 6H, 2 CH₃of Indoline); 3.35 (s, CH₃N); 6.84–7.33

(m, 5 arom. H); 7.38 (d, J = 9.2, 2 arom. H); 7.61 (d, J = 9.2, 2 arom.

H). Anal. calcd for C₁₈H₁₈ClN₃: C 69.33, H 5.82, N 13.48; found: C

69.13, H 5.85, N 13.47.

5.2.36. 1-Benzyl-2-(4-chlorophenylazo)methylenepyrrolidine

(60)

Yield: 32%. Mp 127–129 °C (EtOH). ¹H NMR (CDCl₃): 2.09 (q, 2H,

C(4) of Pyrr); 3.37 (t, J = 7.8, 2H, C(3) of Pyrr); 3.57 (t, J = 7.1, 2H,

C(5) of Pyrr); 4.53 (s, 2H, CH₂C₆H₅); 6.98–7.76 (m, 1H, @CH– and

9 arom. H). Anal. calcd for C₁₈H₁₈ClN₃: C 69.34, H 5.82, N 13.48;

found: C 69.02, H 5.78, N 13.40.

5.4. 2-Amino-1-arylazoethenes of structure F and G (66–85):

general method

a) A solution of the relevant glyoxal monoarylhydrazone

(6.75 mmol) and secondary amine (6.5 mmol) in 60 mL of

anhydrous toluene was stirred at room temperature for

3 h. After removing the solvent under vacuum, the solid

was crystallized with the suitable solvent.

5.2.37. 1-Benzyl-2-(4-methylphenylazo)methylenepyrrolidine

(61)

Yield: 50%. Mp 122–124 °C (Et₂O). ¹H NMR (CDCl₃): 2.07 (q, 2H,

C(4) of Pyrr); 2.34 (s, CH₃); 3.37 (t, J = 7.6, 2H, C(3) of Pyrr); 3.53 (t,

J = 7.0, 2H, C(5) of Pyrr); 4.51 (s, 2H, CH₂C₆H₅); 7.07–7.63 (m, 1H,

@CH– and 9 arom. H). Anal. calcd for C₁₉H₂₁N₃: C 78.32, H 7.26,

N 14.42; found: C 78.02, H 7.24, N 14.38.

b) The required glyoxal monophenylhydrazone was prepared

as described by Severin et al.⁵; in the same way, glyoxal 4-

chlorophenylhydrazone⁴³

and

glyoxal

4-ethoxycar-

bonylphenylhydrazone were prepared. The latter hydrazone

is new. Yield: 84%. Mp 163–165 °C (Et₂O). Anal. calcd for

C

₁₁H₁₂N₂O₃: C 59.99, H 5.49, N 12.72; found: C 60.19, H

5.55, N 12.71. ¹H NMR (CDCl₃): 1.26 (t, J = 7.0, CH₃of COOEt);

4.25 (q, J = 7.0, CH₂of COOEt); 7.24 (d, J = 8.7, 2 arom. H);

7.36 (d, J = 9.3, 1H, @CH–); 7.88 (d, J = 8.8, 2 arom. H); 9.48

(d, J = 10.0, 1H, CHO); 10.97 (s, NH, collapses with D₂O).

5.2.38. 1-Benzyl-2-(3,4-dichlorophenylazo)methylenepyrroli-

dine and 1-benzyl-3-(3,4-dichlorophenylazo)-1,4,5,6-tetrahy-

dropyridine (mixture 3:1) (62)

Yield: 57%. Mp 113–115 °C. ¹H NMR (CDCl₃): 1.94 (q, 0.25 ꢂ 2H,

C(5) of Py); 2.10 (q, 0.75 ꢂ 2H, C(4) of Pyrr); 2.61 (t, J = 6.5,

0.25 ꢂ 2H, C(4) of Py); 3.21 (t, J = 6.0, 0.25 ꢂ 2H, C(6) of Py); 3.38

Compounds 66 and 69 were already described.^5,6

M. Tonelli et al. / Bioorg. Med. Chem. 16 (2008) 8447–8465

8463

5.4.1. 1-(4-Chlorophenylazo)-2-(1-pyrrolidino)ethene (67)

5.4.10. 2-[4-(2-Fluorophenyl)piperazin-1-yl]-1-phenylazoe-

Yield: 41%. Mp 138–140 °C (Et₂O/pentane). ¹H NMR (CDCl₃):

1.91–2.08 (m, 4H, C(3, 4) of Pyrr); 3.25–3.57 (m, 4H, C(2, 5) of Pyrr);

7.08 (d, J = 10.8, 1 vinylic H); 7.30 (d, J = 8.7, 2 arom. H); 7.50 (d,

J = 8.6, 2 arom. H); 7.66 (d, J = 10.8, 1 vinylic H). Anal. calcd for

thene (77)

Yield: 68%. Mp 129–131 °C (CH₂Cl₂/Et₂O). ¹H NMR (CDCl₃): 3.16

(t, J = 5.0, 4H, C(3, 5) of Pz); 3.54 (t, J = 5.0, 4H, C(2, 6) of Pz); 6.85–

7.66 (m, 9 arom. H and 2 vinylic H). Anal. calcd for C₁₈H₁₉FN₄: C

69.66, H 6.17, N 18.05; found: C 69.24, H 5.98, N 17.85.

C₁₂H₁₄ClN₃: C 61.15, H 5.99, N 17.83; found: C 60.81, H 5.98, N 17.32.

5.4.2. 1-(4-Carbethoxyphenylazo)-2-(1-pyrrolidino)ethene (68)

Yield:80%. Mp 163–164 °C (Et₂O). ¹H NMR(CDCl₃): 1.37(t, J = 7.1,

CH₃of COOEt); 1.84–2,10 (m, 4H, C(3, 4) of Pyrr); 3.35–3.54 (m, 4H,

C(2, 5) of Pyrr); 4.34 (q, J = 7.0, CH₂of COOEt); 7.17 (d, J = 10.8, 1

vinylic H); 7.58 (d, J = 8.6, 2 arom. H); 7.71 (d, J = 10.6, 1 vinylic H);

8.02 (d, J = 8.6, 2 arom. H). Anal. calcd for C₁₅H₁₉N₃O₂: C 65.91, H

7.01, N 15.37; found: C 66.08, H 7.06, N 15.44.

5.4.11. 2-[4-(4-Fluorophenyl)piperazin-1-yl]-1-phenylazoeth-

ene (78)

Yield: 68%. Mp 168–170 °C (toluene). ¹H NMR (CDCl₃): 3.17 (t,

J = 5.0, 4H, C(3, 5) of Pz); 3.51 (t, J = 5.2, 4H, C(2, 6) of Pz); 6.85–

7.66 (m,

₁₈H₁₉FN₄+ 0.5 H₂O: C 67.69, H 6.31, N 17.54; found: C 67.82, H

6.14, N 17.82.

9 arom. H and 2 vinylic H). Anal. calcd for

C

5.4.3. 1-(4-Carbethoxyphenylazo)-2-(1-piperidino)ethene (70)

Yield: 63%. Mp 122–124 °C (Et₂O). ¹H NMR (CDCl₃): 1.37 (t,

J = 7.1, CH₃of COOEt); 1.60–1.74 (m, 6H, C(3, 4, 5) of Pip); 3.30–

3.44 (m, 4H, C(2, 6) of Pip); 4.34 (q, J = 7.0, CH₂of COOEt); 7.30–

7.50 (m, 2 vinylic H); 7.58 (d, J = 8.8, 2 arom. H); 8.03 (d, J = 8.8,

2 arom. H). Anal. calcd for C₁₆H₂₁N₃O₂: C 66.88, H 7.37, N 14.62;

found: C 66.60, H 7.27, N 14.59.

5.4.12. 2-[4-(2-Chlorophenyl)piperazin-1-yl]-1-phenylazoeth-

ene (79)

Yield: 49%. Mp 83–85 °C (CH₂Cl₂/Et₂O). ¹H NMR (CDCl3): 3.13

(t, J = 4.8, 4H, C(3, 5) of Pz); 3.55 (t, J = 4.9, 4H, C(2, 6) of Pz);

6.97–7.65 (m, 9 arom. H and 2 vinylic H). Anal. calcd for

C₁₈H₁₉ClN₄: C 66.15, H 5.86, N 17.14; found: 66.00, H 5.96, N 17.30.

5.4.13. 2-[4-(3-Chlorophenyl)piperazin-1-yl]-1-phenylazoe-

thene (80)

5.4.4. 2-(4-Morpholino)-1-phenylazoethene (71)

Yield: 54%. Mp 106 °C (Et₂O). ¹H NMR (CDCl₃): 3.37 (t, J = 4.9,

4H, morpholine); 3.78 (t, J = 4.9, 4H, morpholine); 7.15 (d,

J = 11.0, 1 vinylic H); 7.35–7.70 (m, 5 arom. H and 1 vinylic H).

Anal. calcd for C₁₂H₁₅N₃O: C 66.34, H 6.96, N 19.34; found: C

66.03, H 6.85, N 19.25.

Yield: 75%. Mp 134–135 °C (CH₂Cl₂/Et₂O). ¹H NMR (CDCl₃): 3.27

(t, J = 5.1, 4H, C(3, 5) of Pz); 3.50 (t, J = 5.2, 4H, C(2, 6) of Pz); 6.76–

7.63 (m, 9 arom. H and 2 vinylic H). Anal. calcd for C₁₈H₁₉ClN₄: C

66.15, H 5.86, N 17.14; found: 66.19, H 5.92, N 17.28.

5.4.14. 2-[4-(4-Chlorophenyl)piperazin-1-yl]-1-phenylazoe-

thene (81)

5.4.5. 1-Phenylazo-2-(4-thiomorpholino)ethene (72)

Yield: 48%. Mp 88–89 °C (Et₂O). ¹H NMR (CDCl₃): 2.76 (t, J = 5.0,

4H, thiomorpholine); 3.70 (t, J = 5.1, 4H, thiomorpholine); 7.10–

7.70 (m, 5 arom. H and 2 vinylic H). Anal. calcd for C₁₂H₁₅N₃S: C

61.77, H 6.48, N 18.01, S 13.74; found: C 61.47, H 6.46, N 18.35,

S 13.43.

Yield: 87%. Mp 197–198 °C (toluene). ¹H NMR (CDCl₃): 3.22 (t,

J = 5.1, 4H, C(3, 5) of Pz); 3.51 (t, J = 5.2, 4H, C(2, 6) of Pz); 6.80–

7.65 (m, 9 arom. H and 2 vinylic H). Anal. calcd for C₁₈H₁₉ClN₄: C

66.15, H 5.86, N 17.14; found: 65.84, H 6.00, N 17.39.

5.4.15. 1-Phenylazo-2-[4-(3-triﬂuoromethylphenyl)piperazin-

1-yl]ethene (82)

5.4.6. 2-[(4-Ethoxycarbonyl)piperazin-1-yl]-1-phenylazoethene

(73)

Yield: 69%. Mp 146–148 °C (CH₂Cl₂/Et₂O). ¹H NMR (CDCl₃): 3.32

(t, J = 5.1, 4H, C(3, 5) of Pz); 3.53 (t, J = 5.2, 4H, C(2, 6) of Pz); 7.04–

7.64 (m, 9 arom. H and 2 vinylic H). Anal. calcd for C₁₉H₁₉F₃N₄: C

63.32, H 5.31, N 15.55; found: C 63.06, H 5.39, N 15.91.

Yield: 53%. Mp 106–107 °C (CH₂Cl₂/Et₂O). ¹H NMR (CDCl₃): 1.29

(t, J = 7.1, CH₃of COOEt); 3.35 (t, J = 5.2, 4H, C(3, 5) of Pz); 3.60 (t,

J = 5.1, 4H, C(2, 6) of Pz); 4.18 (q, J = 7.1, CH₂of COOEt); 7.11–7.66

(m, 5 arom. H and 2 vinylic H). Anal. calcd for C₁₅H₂₀N₄O₂: C 62.48,

H 6.99, N 19.43; found: C 62.47, H 7.00, N 19.53.

5.4.16. 2-[4-(4-Nitrophenyl)piperazin-1-yl]-1-phenylazoethene

(83)

5.4.7. 1-Phenylazo-2-[4-(pyridin-2-yl)piperazin-1-yl]ethene

(74)

Yield: 90%. Mp 215–217 °C (CH₂Cl₂/Et₂O). ¹H NMR (CDCl₃):

3.50–3.61 (m, 8H, C(2, 3, 5, 6) of Pz); 6.86 (d, J = 9.4, 2 arom. H,

p-NO₂-phenyl); 7.18–7.65 (m, 5 arom. H and 2 vinylic H); 8.17

(d, J = 9.4, 2 arom. H, p-NO₂-phenyl). Anal. calcd for C₁₈H₁₉N₅O₂:

C 64.08, H 5.68, N 20.76; found: C 64.01, H 5.71, N 20.55.

Yield: 75%. Mp 114–116 °C (CH₂Cl₂/Et₂O). ¹H NMR (CDCl₃): 3.50

(t, J = 5.1, 4H, C(3, 5) of Pz); 3.71 (t, J = 5.2, 4H, C(2, 6) of Pz); 6.65–

7.65 (m, 5 arom. H, 2 vinylic H and 3H, C(3, 4, 5) of Pyridine); 8.20–

8.26 (m, 1H, C(6) of Pyridine). Anal. calcd for C₁₇H₁₉N₅: C 69.60, H

6.53, N 23.87; found: 69.62, H 6.57, N 24.01.

5.4.17. 2-[4-(4-Acetylphenyl)piperazin-1-yl]-1-

phenylazoethene (84)

5.4.8. 1-Phenylazo-2-[4-(pyrimidin-2-yl)piperazin-1-yl]ethene

(75)

Yield: 79%. Mp 188–190 °C (toluene). ¹H NMR (CDCl₃): 2.55 (s,

CH₃CO); 3.45–3.60 (m, 8H, C(2, 3, 5, 6) of Pz); 6.86 (d, J = 9.4, 2

arom. H, p-NO₂-phenyl); 7.18–7.65 (m, 5 arom. H and 2 vinylic

H); 8.17 (d, J = 9.4, 2 arom. H, p-NO₂-phenyl). Anal. calcd for

Yield: 74%. Mp 140–142 °C (CH₂Cl₂/Et₂O). ¹H NMR (CDCl₃): 3.46

(t, J = 5.2, 4H, C(3, 5) of Pz); 3.98 (t, J = 5.3, 4H, C(2, 6) of Pz); 6.57 (t,

J = 5.0, 1H, C(5) of Pyrimidine); 7.20–7.65 (m, 5 arom. H and 2 vinylic

H); 8.35 (d, J = 4.9, 2H, C(4, 6) of Pyrimidine). Anal. calcd for

C₂₀H₂₂N₄O + 0.5H₂O: C 69.64, H 6.75, N 16.31; found: C 70.34, H

6.58, N 16.28.

C₁₆H₁₈N₆: C 65.29, H 6.16, N 28.55; found: C 65.01, H 6.14, N 28.79.

5.4.18. 2-[4-(4-Methoxyphenyl)piperazin-1-yl]-1-phenylazoe-

thene (85)

5.4.9. 1-Phenylazo-2-(4-phenylpiperazin-1-yl)ethene (76)

Yield: 77%. Mp 159–161 °C (CH₂Cl₂/Et₂O). ¹H NMR (CDCl₃): 3.29

(t, J = 5.4, 4H, C(3, 5) of Pz); 3.54 (t, J = 5.5, 4H, C(2, 6) of Pz); 6.90–

7.67 (m, 10 arom. H and 2 vinylic H). Anal. calcd for C₁₈H₂₀N₄: C

73.94, H 6.89, N 19.16; found: C 73.64, H 6.97, N 19.40.

Yield: 60%. Mp 173–175 °C (toluene). ¹H NMR (CDCl₃): 3.16 (t,

J = 5.0, 4H, C(3, 5) of Pz); 3.53 (t, J = 5.1, 4H, C(2, 6) of Pz); 3.79 (s,

CH₃O); 6.97–7.65 (m, 9 arom. H and 2 vinylic H). Anal. calcd for

C₁₉H₂₂N₄O: C 70.78, H 6.88, N 17.38; found: C 70.51, H 6.81, N 17.39.

8464

M. Tonelli et al. / Bioorg. Med. Chem. 16 (2008) 8447–8465

5.5. Cell-based assays

To this end, Vero-76 cells were seeded in 24-well plates at a den-

sity of 2 ꢂ 10⁵cells/well and were allowed to form conﬂuent mon-

olayers by incubating overnight in growth medium at 37 °C in a

humidiﬁed CO₂(5%) atmosphere. Then, monolayers were infected

5.5.1. Compounds

Compounds were dissolved in DMSO at 100 mM and then di-

luted in culture medium.

with 250

lL of proper virus dilutions to give 50–100 PFU/well. Fol-

lowing removal of unadsorbed virus, 500

lL of Dulbecco’s modiﬁed

5.5.2. Cells and viruses

Eagle’s medium supplemented with 1% inactivated FCS and 0.75%

methyl-cellulose, without or with serial dilutions of test com-

pounds, were added. Cultures were incubated at 37 °C for 2 (Sb-1

and VSV), 3 (CVB-2, VV and HSV-1) or 5 days (RSV) and then ﬁxed

with PBS containing 50% ethanol and 0.8% crystal violet, washed

and air-dried. Plaques were then counted. EC₅₀(50% effective con-

centration) was calculated by linear regression technique.

Cell lines were purchased from American Type Culture Collec-

tion (ATCC). The absence of mycoplasma contamination was

checked periodically by the Hoechst staining method. Cell lines

supporting the multiplication of RNA and DNA viruses were the

following: CD4⁺human T cells containing an integrated HTLV-1

genome (MT-4); Madin Darby bovine kidney (MDBK); baby ham-

ster kidney (BHK-21) and monkey kidney (Vero-76) cells.

5.6. Molecular modeling

5.5.3. Cytotoxicity assays

For cytotoxicity tests, run in parallel with antiviral assays,

MDBK and BHK cells were resuspended in 96 multiwell plates at

an initial density of 6 ꢂ 10⁵and 1 ꢂ 10⁶cells/mL, respectively, in

maintenance medium, without or with serial dilutions of test com-

pounds. Cell viability was determined after 48–96 h at 37 °C in a

humidiﬁed CO₂(5%) atmosphere by the 3-(4,5-dimethylthiazol-

2-yl)-2,5-diphenyl-tetrazolium bromide (MTT) method.⁴⁴Vero-76

cells were resuspended in 24 multiwell plates at an initial density

of 4 ꢂ 10⁵cells/mL. The cell number of Vero-76 monolayers was

determined by staining with the crystal violet dye.

For cytotoxicity evaluations, exponentially growing cells de-

rived from human hematological tumors [CD4⁺human T-cells con-

taining an integrated HTLV-1 genome (MT-4)] were seeded at an

initial density of 1 ꢂ 10⁵cells/mL in 96-well plates in RPMI-1640

medium supplemented with 10% fetal calf serum (FCS), 100 U/mL

5.6.1. Pharmacophore modeling

The model structures of all compounds were built using the

CATALYST 2D–3D sketcher,⁴⁵and the conformational search of each

compound was carried out using the Poling algorithm^46–48and

generalized CHARMM force ﬁeld⁴⁹as implemented in the program.

Molecular ﬂexibility was taking into account by considering each

compound as a collection of conformers representing a different

area of conformational space accessible to the molecule within a

given energy range. The ‘best quality’ generation option was

adopted to select representative conformers over a 0–20 kcal/mol

interval above the computed global energy minimum in the con-

formational space. Since the number of conformers generated for

each compound has been limited to a maximum of 250, this search

procedure should identify the best three-dimensional arrangement

of chemical functions explaining the activity variations among the

compound training set.

penicillin G and 100 lg/ml streptomycin. Cell cultures were then

incubated at 37 °C in a humidiﬁed, 5% CO₂atmosphere in the ab-

sence or presence of serial dilutions of test compounds. Cell viabil-

ity was determined after 96 h at 37 °C by the MTT method.

Based on the conformations for each compound, CATALYST 4.9

software package was used to generate plausible three-dimen-

sional pharmacophore models. During hypotheses generation, the

software attempts to minimize a cost function of two main terms:

the ﬁrst penalizes the deviation between the estimated activities of

the training set molecules and their experimental values, while the

second penalizes the complexity of the hypothesis. The uncertain

factor for each compound represents the ratio range of uncertainty

in the activity value based on the expected statistical straggling of

biological data collection. Uncertainty inﬂuences the ﬁrst step—

also called the constructive phase—of the hypothesis generating

process.⁵⁰In this work, an uncertainty of 2.0 was preferred over

the default factor of 3.0, as the experimental activities of our com-

pounds barely span the required four orders of magnitude.

An analysis of the functional groups characterizing our com-

pounds suggested that hydrophobic aromatic (HAr) sites, hydrogen

bond acceptors (HBA), hydrogen bond donors (HBD), positive ion-

izable points (PI), and ring aromatic (RAr) sites could effectively

map their critical chemical features and, hence, describe the anti-

viral activity of our compounds. Accordingly, these ﬁve features

were selected to constitute the essential information in the auto-

mated hypothesis generation process.

Two validation procedures were followed to determine the sta-

tistical relevance and the validity of the proposed 3D pharmaco-

phore models: the test set prediction model, and the Cat-

Scramble method. In this work, the former procedure consisted in

the collection of further, different compounds into three sets

(TsS), and in performing a regression analysis by mapping each

TsS onto the best pharmacophore hypothesis for each virus. The

obtainment of high correlation coefﬁcients generated using the

test set compounds revealed the good correlation between the ac-

tual and estimated activities and, hence, the predictive validity of

the corresponding 3D hypothesis. The Cat-Scramble validation pro-

cedure is based on Fisher’s randomization test.⁵¹The goal of this

5.5.4. Antiviral assay

Activity of compounds against human immunodeﬁciency virus

type-1 (HIV-1) was based on inhibition of virus-induced cytopath-

ogenicity in MT-4 cells acutely infected with a multiplicity of infec-

tion (m.o.i.) of 0.01. Brieﬂy, 50

were added to each well of ﬂat-bottom microtitre trays containing

l

L of RPMI containing 1 ꢂ 10⁴MT-4

50

lL of RPMI, without or with serial dilutions of test compounds.

Then, 20

lL of a HIV-1 suspension containing 100 CCID₅₀were

added. After a 4-day incubation, cell viability was determined by

the MTT method.

Activity of compounds against yellow fever virus (YFV) and Reo

virus type-1 (Reo-1) was based on inhibition of virus-induced cyto-

pathogenicity in acutely infected BHK-21 cells. Activities against

bovine viral diarrhoea virus (BVDV), in infected MDBK cells, were

also based on inhibition of virus-induced cytopathogenicity.

BHK and MDBK cells were seeded in 96-well plates at a density

of 5 ꢂ 10⁴and 3 ꢂ 10⁴cells/well, respectively, and were allowed to

form conﬂuent monolayers by incubating overnight in growth

medium at 37 °C in a humidiﬁed CO₂(5%) atmosphere. Cell mono-

layers were then infected with 50

lL of a proper virus dilution (in

serum-free medium) to give a m.o.i = 0.01. One hour later, 50

lL of

MEM Earle’s medium, supplemented with inactivated fetal calf ser-

um (FCS), 1% ﬁnal concentration, without or with serial dilutions of

test compounds, were added. After 3–4 days of incubation at 37 °C,

cell viability was determined by the MTT method.

Activity of compounds against coxsackie virus type B2 (CVB-2),

Polio Virus type-1 Sabin strain (Sb-1), vesicular stomatitis virus

(VSV), vaccinia virus (VV), herpes virus 1 (HSV-1) and respiratory

syncytial virus (RSV), A-2 strain, in infected Vero-76 cells, was

determined by plaque reduction assays in Vero-76 cell monolayers.

M. Tonelli et al. / Bioorg. Med. Chem. 16 (2008) 8447–8465

8465

Florio, T. Int. J. Immunopathol. Pharmacol. 2006, 19, 339. and references

reported therein.

16. Kimball, D. B.; Haley, M. M. Angew. Chem. Int. Ed. 2002, 41, 3338.

17. Gesher, J. A.; Hickman, J. A.; Simmonds, R. J.; Stevens, M. F. G.; Vaughan, K.

Biochem. Pharmacol. 1981, 30, 89.

18. Stevens, M. F. G.; Phillip, K. S.; Rathbone, D. L.; O’Shea, D. M.; Queener, S. F.;

Schwalbe, C. H.; Lambert, P. A. J. Med. Chem. 1997, 40, 1886.

19. Prous, J.; Graul, A.; Castañer, J. Drugs Future 1994, 19, 746.

20. Jean-Claude, B. J.; Just, G. J. Chem. Soc. Perkin Trans. I 1991, 2525.

21. Jean-Claude, B. J.; Just, G. Heterocycles 1998, 48, 1347.

22. Jean-Claude, B. J.; Mustafa, A.; Damian, Z.; De Marte, J.; Vasilescu, D. E.; Yen, R.;

Chan, T. H.; Leyland-Jones, B. Biochem. Pharmacol. 1999, 288, 753.

23. Jean-Claude, B. J.; Mustafa, A.; Watson, A. J.; Damian, Z.; Vasilescu, D.; Chan, T.

H.; Leyland-Jones, B. J. Pharmacol. Exp. Ther. 1999, 288, 484.

24. Dumont-Hornebeck, B.; Strube, Y. N.; Vasilescu, D.; Jean-Claude, B. J. Bioorg.

Med. Chem. Lett. 2000, 10, 2325.

25. Miyadera, T. In The Chemistry of Hydrazo, Azo and Azoxy Groups; Patai, S., Ed.; J.

Wiley: London, New York, 2002; pp 495–539.

26. Polanc, S. J. Heterocyclic Chem. 2005, 42, 401.

27. Reich, M. F.; Fabio, P. F.; Lee, V. J.; Kuck, N. A.; Testa, R. T. J. Med. Chem. 1989, 32,

2474.

28. Barbieri, F.; Sparatore, A.; Alama, A.; Novelli, F.; Bruzzo, C.; Sparatore, F. Oncol.

Res. 2003, 13, 455.

type of validation is to check whether there is a strong correlation

between the chemical structures and the biological activity. This is

done by randomizing the activity data associated with the training

set compounds, generating pharmacophore hypothesis using the

same features and parameters employed to develop the original

pharmacophore model. The statistical signiﬁcance is calculated

according to the following formula:

significance ¼ 100 ꢂ ½1 ꢀ ð1 þ x=yÞꢃ

where x is the total number of hypotheses having a total cost lower

than the original (best) hypothesis, and y is the total number of

HypoGen runs (initial + random runs). Thus, for instance, 49 random

spreadsheets (i.e., 49 HypoGen runs) have to be generated to obtain

a 98% conﬁdence level. Should any randomized data set result in the

generation of a 3D pharmacophore with similar or even better cost

values, root-mean-square deviations, and correlation coefﬁcients,

then it is likely that the original hypothesis does reﬂect a chance

correlation.

29. Sparatore, F.; Cerri, R. J. Heterocyclic Chem. 1979, 16, 1001.

30. Sparatore, F.; Boido, A. Farmaco 1991, 46(Suppl.), 233.

31. Novelli, F.; Sparatore, F. Farmaco 2002, 57, 871.

Acknowledgments

32. Gault, F.; Wahl, H. Compt. Rend. 1955, 240, 983.

33. Sommer, S. Angew. Chem. Int. Ed. 1977, 16, 58.

Financial support from Italian MIUR (FIRB RBNE01J3SK01) is

gratefully acknowledged. The authors thank O. Gagliardo for per-

forming elemental analyses.

34. Reckhow, W. A.; Tarbell, D. S. J. Am. Chem. Soc. 1952, 74, 4960.

35. Kunieda, T.; Koga, K.; Yamada, S.-I. Chem. Pharm. Bull. 1967, 15, 337.

36. Hirsch, J. A.; Jarmas, A. A. J. Org. Chem. 1978, 43, 4106.

37. Haylor, B. D. G.; Smith, P. H. G. Eur. Pat. Appl. EP 634,413, 1995; Chem. Abstr.

1995, 122, 214088.

References and notes

38. Nagarajan, K.; Talwalker, P. K.; Kulkarni, C. L.; Venkateswarlu, A.; Prabhu, S. S.;

Nayak, G. V. Ind. J. Chem. Sect. B 1984, 23B, 1243.

39. Challand, S. R.; Copp, F. C.; Denyer, C. V.; Eakins, K. E.; Walker, J. M. G.;

Whittaker, N.; Caldwell, A. G. Eur. Pat. Appl. EP 22,578, 1981; Chem. Abstr. 1981,

95, 62193.

40. Enders, E. In Houben-Weyl; Müller, E., Ed.; Methoden der organischen Chemie;

G. Thieme Verlag: Stuttgart, 1967; Vol. 10/II, pp 201–214.

41. Plancher, G. Gazz. Chim. It. 1898, 28 II, 418.

1. Sparatore, F.; Boido, V. Farmaco 1988, 43, 1099.

2. Boido Canu, C.; Boido, V.; Sparatore, F.; Sparatore, A.; Susanna, V.; Russo, S.;

Cenicola, M. L.; Marmo, E. Farmaco, Ed. Sci. 1988, 43, 819.

3. Rosenhauer, E. Chem. Ber. 1924, 57, 1192.

4. König, W.; Müller, J. Chem. Ber. 1924, 57, 144.

5. Severin, T.; Adam, R.; Lerche, H. Chem. Ber. 1975, 108, 1756.

6. Lerche, H.; Severin, T. Chem. Ber. 1978, 111, 1195.

7. Crary, J. W.; Quayle, O. R.; Lester, C. T. J. Am. Chem. Soc. 1956, 78, 5584.

8. Sparatore, A.; Canu Boido, C.; Boido, V.; Sparatore, F.; Debbia, E.; Pesce Schito, A.

Farmaco 1990, 45, 867.

42. Colonna, M.; Bruni, P. Gazz. Chim. It. 1965, 95, 857.

43. Clark, J. D.; Heise, J. D.; Shah, A. S.; Peterson, J. C.; Chou, S. K.; Levine, J.; Karakas,

A. M.; Ma, Y.; Ng, K.-Y.; Patelis, L.; Springer, J. R.; Stano, D. R.; Wettach, R. H.;

Dutra, G. A. Org. Proc. Res. Dev. 2004, 8, 176.

44. Pauwels, R.; Balzarini, J.; Baba, M.; Snoeck, R.; Schols, D.; Herdewijn, P.;

Desmyster, J.; De Clercq, E. J. Virol. Methods 1988, 20, 309.

45. CATALYST 4.9; Accelrys Inc., San Diego, CA, USA, 2004.

46. Smellie, A.; Teig, S. L.; Towbin, P. J. Comput. Chem. 1994, 16, 171.

47. Smelie, A.; Kahn, S. D.; Teig, S. L. J. Chem. Inf. Comput. Sci. 1995, 35, 285.

48. Smelie, A.; Kahn, S. D.; Teig, S. L. J. Chem. Inf. Comput. Sci. 1995, 35, 295.

49. Brooks, B. R.; Bruccoleri, R. E.; Olafson, B. D.; States, D. J.; Swaminathan, S.;

Karplus, M. J. J. Comput. Chem. 1983, 4, 187.

9. Canu Boido, C.; Boido, V.; Sparatore, F.; Sparatore, A.; Bombieri, G.; Benetollo,

F.; Debbia, E.; Pesce Schito, A. Farmaco 1993, 48, 749.

10. Elslager, E. F.; Worth, D. F. J. Med. Chem. 1963, 6, 444.

11. Elslager, E. F.; Capps, D. B.; Kurz, D. H.; Short, F. W.; Werbel, L. M.; Worth, D. F. J.

Med. Chem. 1966, 9, 378.

12. Werbel, L. M.; Elslager, E. F.; Fisher, M. W.; Gavrilis, Z. B.; Phillips, A. A. J. Med.

Chem. 1968, 11, 411.

13. Vazzana, I.; Sparatore, F.; Fadda, G.; Manca, C. Farmaco 1993, 48, 737.

14. Milhavet, O.; Mange, A.; Casanova, D.; Lehmann, S. J. Neurochem. 2000, 74, 222.

15. Corsaro, A.; Paludi, D.; Villa, V.; D’Arrigo, C.; Chiovitti, K.; Thellung, S.;

Russo, C.; Di Cola, D.; Ballerini, P.; Patrone, E.; Schettini, G.; Aceto, A.;

50. Güner, O. F. In Pharmacophore Perception Development and Use in Drug Design;

International University Line: La Jolla, CA, USA, 2000; p 173.

51. Fisher, R. The Design of Experiments; Hafner Publishing: New York, NY, USA, 1966.

Article Doi

Antimicrobial and cytotoxic arylazoenamines. Part III: Antiviral activity of selected classes of arylazoenamines

DOI: 10.1016/j.bmc.2008.08.028

Source and publish data:

Authors:

Article abstract of DOI:10.1016/j.bmc.2008.08.028

Full text of DOI:10.1016/j.bmc.2008.08.028

Products guided by the article

R&D Labs maybe for 61178-18-5

Relevant to this article

Hot Product