Synthesis and biological evaluation of first N-alkyl syn dimeric 4-aryl-1,4-dihydropyridines as competitive HIV-1 protease inhibitors

Article abstract of DOI:10.1016/S0223-5234(01)01228-4

A first series of novel N-alkyl substituted syn dimeric 4-aryl-1,4-dihydropyridines 12-17 have been synthesised and evaluated as HIV-1 protease inhibitors in in vitro assays. While the N-methyl derivatives 12 and 13 were almost inactive, with IC₅₀-values of about 225 μM, the N-benzyl compounds with varied ester groups all exhibited stronger activities, with IC₅₀-values of 11-12 μM for the presently best compounds 16 and 17 with ethyl ester functions. The type of HIV-1 protease inhibition of the novel inhibitors was characterised as competitive. With the increase of observed activity from N-methyl derivatives to N-benzyl compounds the binding mode may correspond to that of cyclic ureas with hydrophobic interactions of the four aromatic residues to the S1/S1′ and S2/S2′ regions of HIV-1 protease.

Full text of DOI:10.1016/S0223-5234(01)01228-4

Eur. J. Med. Chem. 36 (2001) 367–374
´
© 2001 Editions scientifiques et me´dicales Elsevier SAS. All rights reserved
PII: S0223-5234(01)01228-4/FLA
Short Communication
Synthesis and biological evaluation of first N-alkyl syn dimeric
4-aryl-1,4-dihydropyridines as competitive HIV-1 protease inhibitors
Andreas Hilgerotha*, Andreas Billichb, Hauke Liliec
aDepartment of Pharmacy, Martin-Luther University Halle–Wittenberg, Wolfgang-Langenbeck Straße 4,
D-06120 Halle, Germany
bDepartment of Antiretroviral Therapy, Novartis Forschungsinstitut GmbH, Brunnerstraße 59, A-1235 Vienna, Austria
cDepartment of Biochemistry and Biotechnology, Martin-Luther University Halle–Wittenberg, Kurt-Mothes Straße 3,
D-06120 Halle, Germany
Received 5 January 2001; accepted 16 March 2001
Abstract – A first series of novel N-alkyl substituted syn dimeric 4-aryl-1,4-dihydropyridines 12–17 have been synthesised and
evaluated as HIV-1 protease inhibitors in in vitro assays. While the N-methyl derivatives 12 and 13 were almost inactive, with
IC50-values of about 225 mM, the N-benzyl compounds with varied ester groups all exhibited stronger activities, with IC50-values of
11–12 mM for the presently best compounds 16 and 17 with ethyl ester functions. The type of HIV-1 protease inhibition of the novel
inhibitors was characterised as competitive. With the increase of observed activity from N-methyl derivatives to N-benzyl
compounds the binding mode may correspond to that of cyclic ureas with hydrophobic interactions of the four aromatic residues
´
to the S1/S1% and S2/S2% regions of HIV-1 protease. © 2001 Editions scientifiques et me´dicales Elsevier SAS
Syn dimeric 4-aryl-1,4-dihydropyridines / HIV-1 protease inhibitor / Competitive inhibition
1. Introduction
bioavailability. However, hopeful inhibitors of a first
series of cyclic ureas, e.g. DMP 323 (figure 1), failed
in clinical trials, with unsatisfactory bioavailabilities
[5]. From a series of 4-hydroxy-2-pyrones as non-pep-
tidic PIs with better oral bioavailability, one represen-
tative PNU-140690 (figure 1) is presently undergoing
clinical trials [6]. Nevertheless, the development of
novel non-peptidic lead structures is a great challenge
to the drug design of today. Cage dimeric 4-aryl-1,4-
dihydropyridines have recently been introduced as a
novel class of non-peptidic PIs [7, 8].
N-Substituted cage dimeric 4-aryl-1,4-dihydropy-
ridines I possess functional groups similar to those of
cyclic ureas (figure 1). Certain conformities in the
molecular properties of the cage dimeric target struc-
ture I and cyclic ureas were suggested by molecular
modeling [9]. One main functional difference between
cyclic ureas and these cage dimers is the urea carbonyl
group that directly binds to Asp25A and 25B, replac-
ing a water molecule and the hydroxymethyl groups
of the cage dimers, which were also suggested for
Since the discovery of HIV-1 protease (PR) as a
novel target enzyme for the development of HIV-1
protease inhibitors (PI), certain peptidic PIs have
been established in HIV-therapy combined with nu-
cleoside analogues or reverse transcriptase inhibitors
[1–3]. Resistance development against peptidic PIs,
which should be lowered by those combined thera-
pies, is one main problem today because the resis-
tances described earlier against single PIs mutated to
certain cross-resistances against all the marketed pep-
tidic PIs [4]. The additional problems of the present
peptidic PIs, poor oral bioavailability and severe side
effects, just in long-term therapies like cardiovascular
morbidity and mortality, strengthened the develop-
ment of non-peptidic PIs that promised better
* Corresponding author
E-mail address: hilgeroth@pharmazie.uni-halle.de
(A. Hilgeroth).

368
A. Hilgeroth et al. / European Journal of Medicinal Chemistry 36 (2001) 367–374
direct binding to those amino acids by molecular
modeling.
2. Chemistry
Syn dimer 4 has been given by the solid-state
photodimerisation reaction of monomeric 4-phenyl-
1,4-dihydropyridine 1 [10] using unfiltered light of
Ultra Vitalux lamps® with u]270 nm [11] (figure 2).
Monomeric derivatives 2 and 3, which were them-
selves produced by cyclocondensation reactions of
benzaldehyde, methyl and ethyl propiolate, respec-
tively, and benzylamine in acetic acid, yielded syn
dimers 5 and 6 by solution dimerisation reactions
besides anti dimers 7 and 8 using filtered light at
u>313 nm [12] (figure 2). The syn configuration was
proved by the cyclisation reaction of the dimers to
cage compounds on irradiation with unfiltered light
We now present a first series of syn dimeric 4-aryl-
1,4-dihydropyridines with target structure II supplied
with ester carbonyl functions instead of the hydroxy-
methylene functions. Compared with the cage dimers
the ring-opened character of the syn dimers provided
higher flexibility for the binding of their aromatic
residues to the hydrophobic regions S1/S1% and S2/S2%
of PR. The variously substituted target structure II
was simply yielded by few reaction steps, thus follow-
ing a facile synthetic route.
The novel non-peptidic PIs have been evaluated
regarding their PI-activity and the type of inhibition
of PR in an in vitro system.
Figure 1. Structures of cyclic urea DMP 323, 4-hydroxy-2-pyrone PNU-140690, cage dimeric 4-aryl-1,4-dihydropyridine I and syn
dimeric target structure II (with chosen N-benzyl substitution) discussed in the text.

A. Hilgeroth et al. / European Journal of Medicinal Chemistry 36 (2001) 367–374
369
Figure 2. Synthesis of intermediate compounds 4–8.
of the carbamide ester chromophore between u=
304–330 nm, the IR absorption bands between w=
1669–1679 cm−1 as well as by the unchanged
chemical shifts of the vinylogous protons at lꢀ7.4
ppm, compared with the syn dimers 4–6 and 9–11.
(See Section 6 for selected compounds 4 and 12.)
under excitation of the carbamide ester chromophore
as has been demonstrated for reported syn dimers
9–11, while anti dimers remain stable on further
irradiation (A. Hilgeroth, unpublished results).
Selective ester group reductions in all the syn
dimers 4–6 and 9–11 to the corresponding alcoholic
target structures 12–17 succeeded under optimised
conditions using lithium aluminium hydride (LiAlH4)
at −8°C in dry THF without any affecting of the
carbamide ester function (figure 3). The vinylogous
carbamide ester structure in 12–17 was proved spec-
troscopically by the almost unchanged UV absorption
3. Pharmacology
The in vitro inhibitory activities of the synthesised
syn dimers 12–17 against PR have been evaluated.
Saquinavir has been used as the reference compound

370
A. Hilgeroth et al. / European Journal of Medicinal Chemistry 36 (2001) 367–374
Figure 3. Synthesis of target structures 12–17.
with a determined percentage inhibition of 99% at a
concentration of 1 mM and a Ki value of 2.1 nM. The
IC50-values of the syn dimers are presented (table I).
Due to the poor solubilities of all the derivatives 12–17
except 16, their IC50-values could not be determined
directly. They were estimated based on triple inhibitory
determinations at 1 (17), 6.25 (14,15) or 50 mM for the
other compounds assuming parallel dose–response
curves according to Eq. (1). The determination of the
presented Ki values (table I) followed Ref. [13].
suggest a favourable binding of the aromatic N-benzyl
substituents to the hydrophobic binding regions S1/
S1%or S2/S2% of PR. Such a binding will not be possible
in the case of the small N-methyl derivatives. More-
over, for those reported cyclic ureas with hydrophobic
binding of all four aromatic residues to PR, N-benzyl
substituents are also favoured substituents for binding
to those hydrophobic regions [5]. The change of ester
substitution from methyl to ethyl derivatives in 16 and
17 slightly increases the inhibitory activity, with an
IC50-value of 11 mM and Ki of 4.8 mM for the presently
best compound 17. So syn dimers 12–15 are moderately
potent inhibitors of PR.
v
IC50
=
(1)
vmax [I]+IC50
The novel PIs were proved to be competitive in-
hibitors of PR. From the Lineweaver–Burk plot of
4. Results and discussion
Table I. HIV-1 protease inhibition of syn dimers 12–17.
Within the series of investigated methyl ester deriva-
tives 12–15 both the N-methyl derivatives 12 and 13
were almost inactive, with IC50-values>200 mM. The
introduction of N-benzyl substituents in 14 and 15
mainly increases the inhibitory activities, with IC50 <30
mM and inhibitory constants of about 10 mM. The
differences in the activity data from N-methyl to
N-benzyl substitution of the methyl ester derivatives
Compound
IC50 (mM)
Ki (mM)
12
13
14
15
16
17
254922.9
222920.5
2191.5
2996.3
12.190.9
1190.7
11099.5
9698.8
8.790.6
12.192.6
5.890.4
4.890.3

A. Hilgeroth et al. / European Journal of Medicinal Chemistry 36 (2001) 367–374
371
cation. UV spectra were recorded on a Diode Array
1
spectrophotometer 8452A. H-NMR (400 or 500 MHz)
spectra were measured using tetramethylsilane as inter-
nal standard. TLC was performed on E. Merck 5554
silica gel plates. Mass spectra were measured with an
AMD 402 mass spectrometer. Elemental analyses indi-
cated by the symbols of the elements were within 0.4%
of the theoretical values and were performed using a
Leco CHNS-932 apparatus (table II). The synthesis of
compounds 9–11 was recently reported in Ref. [11]
(9,11) and Ref. [12] (10).
Figure 4. Lineweaver–Burk-Plot for inhibitor 16. The sub-
strate concentration was varied in the range 15–500 mM. The
inhibitor concentration was fixed at 15 mM. From linear fits of
the data the parameters Vmax and Km/app were calculated with
Vmax (mM min−1)=37 4 and Km (mM)=214 38 for the unin-
Table II. Elemental analyses of compounds.
C
H
N
hibited reaction and Vmax (mM min−1)=45
4 and Kapp
2
(mM)=777 98 for the inhibited reaction. The Ki value was
determined with 5.7 mM, thus well corresponding to the one
given in table I.
Calc.
Found
3
72.71
72.63
5.82
5.56
3.85
3.73
Calc.
Found
4
Calc.
Found
5
Calc.
Found
6
73.64
73.27
6.44
6.06
3.58
3.52
velocities of substrate cleavage at various substrate
concentrations for uninhibited and inhibited reactions
with selected N-benzyl derivative 16 the maximum
velocity of substrate cleavage nearly remains un-
changed while the calculated Michaelis–Menten con-
stant increases from 214 to 777 mM (figure 4).
66.90
66.87
5.92
5.82
4.93
4.82
72.71
72.76
5.82
5.89
3.85
3.99
Calc.
Found
7
Calc.
Found
8
Calc.
Found
12
Calc.
Found
13
Calc.
Found
14
Calc.
Found
15
Calc.
Found
16
73.64
73.34
6.44
6.41
3.58
3.43
5. Conclusions
72.71
72.61
5.82
5.67
3.85
3.67
The results of this study show that the N-benzyl
substituent as an aromatic one is absolutely necessary
for certain binding activities of the novel PIs. The
very slight difference in the biological activities of the
methyl and ethyl ester derivatives may indicate that
there will be no favourable interactions of the varied
alcoholic component of the ester groups with the
active site of PR. The non-significant protein binding
[14] and the poor metabolism of the dimeric, non-pep-
tidic PIs compared with the ureas [15] are advanta-
geous with respect to bioavailability. So the
development of this promising class of PIs is strongly
encouraged.
73.64
73.38
6.44
6.55
3.58
3.32
67.16
66.95
6.71
6.48
5.22
4.88
65.40
65.55
6.69
7.00
4.77
4.59
74.21
74.29
6.37
6.65
4.12
4.10
70.57
70.71
6.46
6.28
3.74
3.51
Calc.
Found
17
Calc.
Found
75.62
75.60
6.63
6.64
4.01
4.00
6. Experimental protocols
6.1. Chemistry
72.80
72.80
6.64
6.76
3.69
3.67
Commercial reagents were used without further purifi-

372
A. Hilgeroth et al. / European Journal of Medicinal Chemistry 36 (2001) 367–374
2
6.1.1. Dimethyl 1-benzyl-1,4-dihydro-4-phenylpyridine-
3,5-dicarboxylate (2)
4.90 (d, 2H, CHAN, J=15.0 Hz), 6.86–7.43 (m, 20H,
aromatic H), 7.83 (s, 2H, H-2, -6). EIMS; m/z: 726 [M+].
Anal. C44H42N2O8 (C, H, N).
Methyl propiolate (1.68 g, 20 mmol), benzaldehyde
(1.06 g, 10 mmol) and benzylamine (1.07 g, 10 mmol) were
heated in 1 mL of glacial acetic acid on a steam bath for
15 min. The reaction mixture was then poured into
ice-water (100 mL), from which 2.5 g (69%) of 2 crystallised
6.1.5. Tetraethyl 1,5-dibenzyl-1,5,8,8bi-tetrahydro-
4,8-diphenylcyclobuta[1,2b:3,4b%]dipyridine-
3,4ai,7,8ai(4H,4biH)tetracarboxylate (6)
Prepared in a way similar to that described for 5 from
3 (0.40 g, 1.02 mmol). Yield: 0.14 g (36%). White needles
1
on stirring (m.p.: 158–161°C). H-NMR (CDCl3): l=
3.59 (s, 6H, CH3OOC), 4.56 (s, 2H, CH2N), 4.89 (s, 1H,
H-4), 7.10–7.42 (m, 12H, aromatic H, H-2, -6). EIMS;
m/z: 363 [M+]. Anal. C22H21NO4 (C, H, N).
1
(m.p.: 138–145°C). H-NMR (CDCl3): l=0.87 (t, 6H,
CH3CH2OOCꢀC4a, C8a, 3J=7.2 Hz), 1.15 (t, 6H,
3
CH3CH2OOCꢀC3, C7, J=7.0 Hz), 3.53 (AMX3, 2H,
CH3CHMOOCꢀC4a, C8a, 2J=10.7 Hz, 3J=7.2 Hz), 3.63
(AMX3, 2H, CH3CHAOOCꢀC4a, C8a, 2J=10.7 Hz,
3J=7.2 Hz), 3.99 (AMX3, 2H, CH3CHMOOCꢀC3, C7,
6.1.2. Diethyl 1-benzyl-1,4-dihydro-4-phenylpyridine-
3,5-dicarboxylate (3)
Prepared in a way similar to that described for 2 from
ethyl propiolate (1.96 g, 20 mmol), benzaldehyde (1.06 g,
10 mmol) and benzylamine (1.07 g, 10 mmol). Yield: 2.8
2J=10.7 Hz, J=7.0 Hz), 4.05 (s, 2H, H-4, -8), 4.06
3
2
3
(AMX3, 2H, CH3CHAOOCꢀC3, C7, J=10.7 Hz, J=
1
g (72%). Yellow powder (m.p.: 145–147°C). H-NMR
7.0 Hz), 4.46 (s, 2H, H-4b, -8b), 4.59 (d, 2H, CHBN,
(CDCl3): l=1.17 (t, 6H, CH3CH2 ,3J=7.0 Hz), 4.08 (q,
2J=14.8 Hz), 4.71 (d, 2H, CHAN, J=14.8 Hz), 6.88–
2
3
4H, CH3CH2, J=7.0 Hz), 4.46 (s, 2H, CH2N), 4.91 (s,
7.40 (m, 20H, aromatic H), 7.61 (s, 2H, H-2, -6). EIMS;
m/z: 782 [M+]. Anal. C48H50N2O8 (C, H, N).
1H, H-4), 7.09–7.73 (m, 12H, aromatic H, H-2, -6).
ESIMS; m/z: 392 [M+H+]. Anal. C24H25NO4 (C, H, N).
6.1.6. Tetramethyl 1,5-dibenzyl-1,5,8,8bh-tetrahydro-
4,8-diphenylcyclobuta[1,2b:3,4b%]dipyridine-
3,4ah,7,8ai(4H,4biH)tetracarboxylate (7)
6.1.3. Tetramethyl 1,5,8,8bi-tetrahydro-1,5-
dimethyl-4,8-diphenylcyclobuta[1,2b:3,4b%]dipyridine-
3,4ai,7,8ai(4H,4biH)tetracarboxylate (4)
Crystalline 1 (1.0 g, 1.74 mmol) with a layer thickness
of 1 mm was irradiated as recently described [11]. After
three days the reaction product was dissolved in boiling
toluene (100 mL), from which 0.71 g (71%) of 4 crystallised
on cooling (m.p.: 226–230°C). UV (CHCl3): umax (log m)
The volume of the mother liquid from the preparation
of 5 was further reduced leading to crystallisation of anti
dimer 7 as a white powder with a total yield of 0.13 g (32%)
1
(m.p.: 251–256°C). H-NMR (CDCl3): l=3.16 (s, 6H,
CH3OOCꢀC4a, C8a), 3.52 (s, 6H, CH3OOCꢀC3, C7), 4.13
2
(s, 2H, H-4, -8), 4.38 (d, 2H, CHBN, J=15.0 Hz), 4.46
1
(s, 2H, H-4, -8), 4.60 (d, 2H, CHAN, 2J=15.0 Hz),
6.58–7.42 (m, 20H, aromatic H), 7.46 (s, 2H, H-2, -6).
EIMS; m/z: 726 [M+]. Anal. C44H42N2O8 (C, H, N).
244 (3.9), 297 (4.35). IR (KBr): w1732, 1668. H-NMR
(CDCl3): l=3.23 (s, 6H, CH3N), 3.33 (s, 6H,
CH3OOCꢀC4a, C8a), 3.55 (s, 6H, CH3OOCꢀC3, C7), 3.54
(s, 2H, H-4, -8), 4.38 (s, 2H, H-4b, -8b), 7.00–7.15 (m,
10H, aromatic H), 7.41 (s, 2H, H-2, -6). ESIMS; m/z: 597
[M+Na+]. Anal. C32H34N2O8 (C, H, N).
6.1.7. Tetraethyl 1,5-dibenzyl-1,5,8,8bh-tetrahydro-
4,8-diphenylcyclobuta[1,2b:3,4b%]dipyridine-
3,4ah,7,8ai(4H,4biH)tetracarboxylate (8)
6.1.4. Tetramethyl 1,5-dibenzyl-1,5,8,8bi-
tetrahydro-4,8-diphenylcyclobuta[1,2b:3,4b%]dipyridine-
3,4ai,7,8ai(4H,4biH)tetracarboxylate (5)
After separation of 6 the anti-dimer 8 crystallises from
the solution with a collected yield 0.13 g (32%) as a white
1
powder (m.p.: 231–236°C). H-NMR (CDCl3): l=1.11
2 (0.40 g, 1.10 mmol) was dissolved in 40 mL methanol–
THFunderstirring. Thesolutionwasirradiatedinaquartz
flask placed in a copper (II) sulphate bath (1.25 M) with
an Ultra-Vitalux® lamp from a distance of 60 cm for about
4 weeks. After reducing the solution volume syn dimer 5
crystallised, finally yielding 0.14 g (35%) as white powder
(m.p.: 237–240°C). 1H-NMR ((CD3)2SO): l=3.02 (s, 6H,
CH3OOC), 3.43 (s, 6H, CH3OOC), 3.83 (s, 2H, H-4, -8),
4.19 (s, 2H, H-4b, -8b), 4.49 (d, 2H, CHBN, 2J=15.0 Hz),
(t, 6H, CH3CH2OOCꢀC4a, C8a, 3J=7.2 Hz), 1.16 (t, 6H,
CH3CH2OOCꢀC3, C7, 3J=7.2 Hz), 3.58 (q, 4H,
CH3CH2OOCꢀC4a, C8a, 3J=7.2 Hz), 3.94 (AMX3, 2H,
2
3
CH3CHMOOCꢀC3, C7, J=10.7 Hz, J=7.2 Hz), 4.06
2
3
(AMX3, 2H, CH3CHAOOCꢀC3, C7, J=10.7 Hz, J=
7.2 Hz), 4.17 (s, 2H, H-4, -8), 4.39 (d, 2H, CHBN, 2J=14.7
Hz), 4.52 (s, 2H, H-4b, -8b), 4.67 (d, 2H, CHAN, 2J=14.7
Hz), 6.94–7.42 (m, 20H, aromatic H), 7.46 (s, 2H, H-2,
-6). EIMS; m/z: 782 [M+]. Anal. C48H50N2O8 (C, H, N).

A. Hilgeroth et al. / European Journal of Medicinal Chemistry 36 (2001) 367–374
373
3
6.1.8. General procedure for compounds 12–17
2J=14.1 Hz), 4.61 (t, 2H, exchangeable, OH, J=3.1
Hz), 7.09–7.43 (m, 20H, aromatic H), 7.49 (s, 2H, H-2,
-6). EIMS; m/z: 670 [M+]. Anal. C42H42N2O6·0.5H2O (C,
H, N).
6.1.8.1. Dimethyl 1,5,8,8bi-tetrahydro-
4ai,8ai-dihydroxymethyl-1,5-dimethyl-
4,8-diphenylcyclobuta[1,2b:3,4b%]dipyridine-3,7
(4H, 4biH)dicarboxylate (12)
6.1.8.4. Dimethyl 1,5-dibenzyl-1,5,8,8bi-tetrahydro-
4ai,8ai-dihydroxymethyl-4,8-bis(4-methoxyphenyl)-
cyclobuta[1,2b:3,4b%]dipyridine-3,7(4H,4biH)-
dicarboxylate (15)
4 (0.04g, 0.07 mmol) was dissolved in dry THF (25 mL)
and the solution was cooled down to −8°C. Then a
solution of lithium aluminium hydride (1.4 mL, 1.4 mmol)
in THF (1 M) was added. After 6 h the reaction mixture
was hydrolysed with 2 mL of a solution of potassium
hydroxide (20%) at 0°C. The water layer was then
extracted with 100 mL chloroform three times. The
combined extracts were dried over sodium sulphate. On
reducing the extraction volume crude 12 crystallised and
was recrystallised from methanol–water yielding 0.026 g
(71%) of pure 12 as a white powder (m.p.: 255–260°C).
UV (CHCl3): umax (lg m) 242 (3.4), 330 (3.4). IR (KBr):
w1670. 1H-NMR ((CD3)2SO): l=2.94 (‘d’, 2H, CHBOH,
2J=10.5 Hz), 3.11 (s, 6H, CH3N), 3.15 (‘d’, 2H, CHAOH,
2J=10.5 Hz), 3.31 (s, 2H, H-4b, -8b), 3.34 (s, 6H,
CH3OOC), 3.47 (s, 2H, H-4, -8), 4.57 (s, br, 2H, exchange-
able, OH), 7.05–7.16 (m, 10H, aromatic H), 7.32 (s, 2H,
H-2, -6). ESIMS; m/z: 519 [M+H+]. Anal.
C30H34N2O6·H2O (C, H, N).
Yield: 0.016 g (43%) (m.p.: 235–245°C). 1H-NMR
3
(CDCl3): l=0.86 (d, 2H, exchangeable, OH, J=1.0
Hz), 3.18 (dd, after D2O addition, d, 2H, CHBOH,
3
2J=12.0, J=1 Hz), 3.29 (s, 2H, H-4b, -8b), 3.34 (‘d’,
2H, CHAOH, 2J=12.0 Hz), 3.53 (s, 6H, CH3OOC), 3.60
(s, 2H, H-4, -8), 3.72 (s, 6H, CH3O), 4.45 (d, 2H, CHBN,
2
2J=15.0 Hz), 4.55 (d, 2H, CHAN, J=15.0 Hz), 6.61–
6.66 (m, br, 8H, aromatic H of 4-CH3OPh), 7.26–7.45 (m,
10H, aromatic H of Bzl), 7.54 (s, 2H, H-2, -6). EIMS; m/z:
730 [M+]. Anal. C44H46N2O8·H2O (C, H, N).
6.1.8.5. Diethyl 1,5-dibenzyl-1,5,8,8bi-tetrahydro-
4ai,8ai-dihydroxymethyl-4,8-diphenyl-cyclo-
buta[1,2b:3,4b%]dipyridine-3,7(4H,4biH)dicarboxylate
(16)
Yield: 0.029 g (81%) (m.p.: 261–263°C). 1H-NMR
((CD3)2SO): l=1.10 (t, 6H, CH3CH2, 3J=7.0 Hz), 2.91
(dd, after D2O addition d, 2H, CHBOH, 2J=10.9,
3J=3.9 Hz), 3.19 (dd, after D2O addition d, 2H, CHAOH,
6.1.8.2. Dimethyl 1,5,8,8bi-tetrahydro-4ai,8ai-dihydro-
xymethyl-1,5-dimethyl-4,8-bis(4-methoxyphenyl)-
cyclobuta[1,2b:3,4b%]dipyridine-3,7(4H,4biH)-
dicarboxylate (13)
3
2J=10.9, J=3.9 Hz), 3.48 (s, 2H, H-4b, -8b), 3.62 (s,
2
2H, H-4, -8), 3.82–3.92 (m, 4H, CH3CHAM, J=10.9,
Yield: 0.026 g (72%) (m.p.: 210–225°C). 1H-NMR
3J=7 Hz), 4.45 (d, 2H, NCHB, 2J=14.3 Hz), 4.55 (d, 2H,
3
2
(CDCl3): l=1.31 (dd, 2H, exchangeable, OH, J=8.0
Hz, 5.0 Hz), 3.20 (s, 6H, CH3N), 3.21 (s, 2H, H-4b, -8b),
NCHA, J=14.3 Hz), 4.59 (t, 2H, exchangeable, OH,
3J=3.9 Hz), 7.00–7.45 (m, 22H, aromatic H, H-2, -6).
ESIMS; m/z: 721 [M+Na+]. Anal. C44H46N2O6 (C, H, N).
2
3.28 (dd, after D2O addition d, 2H, CHBOH, J=11.5,
3J=8.0 Hz), 3.46 (s, 2H, H-4, -8), 3.47 (dd, after D2O
addition d, 2H, CHAOH, 2J=11.5, 3J=5.0 Hz), 3.51 (s,
6H, CH3OOC), 3.74 (s, 6H, CH3O), 6.75–6.90 (m, 8H,
aromatic H), 7.39 (s, 2H, H-2, -6). FDMS; m/z: 578 [M+].
Anal. C32H38N2O8·0.5H2O (C, H, N).
6.1.8.6. Diethyl 1,5-dibenzyl-1,5,8,8bi-tetrahydro-
4ai,8ai-dihydroxymethyl-4,8-bis(4-methoxyphenyl)-
cyclobuta[1,2b:3,4b%]dipyridine-3,7(4H,4biH)-
dicarboxylate (17)
Yield: 0.027 g (75%) (m.p.: 257–260°C). 1H-NMR
((CD3)2SO): l=1.11 (t, 6H, CH3CH2, 3J=7.0 Hz), 2.92
(dd, after D2O addition d, 2H, CHBOH, 2J=10.0,
3J=3.1 Hz), 3.12 (dd, after D2O addition d, 2H, CHAOH,
6.1.8.3. Dimethyl 1,5-dibenzyl-1,5,8,8bi-tetrahydro-
4ai,8ai-dihydroxymethyl-4,8-diphenylcyclo-
buta[1,2b:3,4b%]dipyridine-3,7(4H,4biH)dicarboxylate
(14)
3
2J=10.0, J=3.1 Hz), 3.43 (s, 2H, H-4b, -8b), 3.56 (s,
Yield: 0.015 g (41%) (m.p.: 222–225°C). 1H-NMR
((CD3)2SO): l=3.06 (dd, after D2O addition d, 2H,
CHBOH, 2J=12.0, 3J=3.1 Hz), 3.25 (dd, after D2O
addition d, 2H, CHAOH, 2J=12.0, 3J=3.1 Hz), 3.38 (s,
2H, H-4b, -8b), 3.42 (s, 6H, CH3OOC), 3.54 (s, 2H, H-4,
-8), 4.44 (d, 2H, CHBN, 2J=14.1 Hz), 4.57 (d, 2H, CHAN,
2H, H-4, -8), 3.66 (s, 6H, CH3O), 3.81–3.94 (m, 4H,
2
3
CH3CHAM, J=10.7, J=7.0 Hz), 4.43 (d, 2H, NCHB,
2J=14.5 Hz), 4.61 (d, 2H, NCHA, 2J=14.5 Hz), 4.61 (t,
2H, exchangeable, OH, 3J=3.1 Hz), 6.54–7.49 (m, 20H,
aromatic H). FDMS; m/z: 758 [M+]. Anal. C46H50N2O8
(C, H, N).

374
A. Hilgeroth et al. / European Journal of Medicinal Chemistry 36 (2001) 367–374
8
[3] Stellbrink H.-J., Dt. Arztebl. 94A (1997) 2497–2503.
6.2. In vitro assay of PR inhibition
[4] Vella St, Palmisano L., Antiviral Res. 45 (2000) 1–7.
[5] De Lucca G.V., Lam Y.S., Drugs Future 23 (1998) 987–994.
[6] Aristoff P.A., Drugs Future 23 (1998) 995–999.
The enzymatic activity of PR was measured by fol-
lowing the cleavage of the substrate H-Lys-Ala-Arg-Val-
Leu-p-nitrophenylalanine-Glu-Ala-Nle-NH2 [16]. PR
was incubated at 37°C in 0.1 M MES, 0.37 M NaCl and
4 nM EDTA, pH 6.25, with 280 mM substrate in the
presence or absence of inhibitors in a photometer cu-
vette. From the decrease of absorbance at 298 nm, the
initial reaction rates were calculated [13]. Compounds
were added from stock solutions in DMSO.
[7] Hilgeroth A., Billich A., Arch. Pharm. Pharm. Med. Chem. 332
(1999) 3–5.
[8] Hilgeroth A., Billich A., Arch. Pharm. Pharm. Med. Chem. 332
(1999) 380–384.
[9] Zo¨llner R., Hilgeroth A., Arch. Pharm. Pharm. Med. Chem.
331 (1998) 31 (Suppl. 1).
[10] Chennat U., Eisner U., J. Chem. Soc., Perkin Trans. I 10 (1975)
926–927.
[11] A. Hilgeroth, U. Baumeister, F.W. Heinemann, Eur. J. Org.
Chem. (1998) 1213–1218.
[12] A. Hilgeroth, U. Baumeister, F.W. Heinemann, Eur. J. Org.
Acknowledgements
Chem. (2000) 245–249.
[13] Scholz D., Billich A., Charpiot B., Ettmayer P., Lehr P.,
Rosenwirth B., Schreiner E., Gstach H.J., Med. Chem. 37
(1994) 3079–3089.
A.H. is grateful for the support of his work by the
German Pharmaceutical Society (DPhG).
[14] Hilgeroth A., Langner A., Pharmazie 55 (2000) 542–543.
[15] Hilgeroth A., Langner A., Arch. Pharm. Pharm. Med. Chem.
333 (2000) 195–197.
References
[16] Richards A.D., Phylip L.H., Farmeries W.G., Scaborough
P.E., Alvarez A., Dunn B.M., Hirel Ph.-H., Konvalinka J.,
Strop P., Pavlickova L., Kosta V., J. Biol. Chem. 265 (1990)
7733–7736.
[1] Hilgeroth A., Pharm. Uns. Zeit 27 (1998) 22–25.
[2] Hilgeroth A., Pharm. Uns. Zeit 27 (1998) 111–116.
.