Rhodanine derivatives as novel inhibitors of PDE4

Article abstract of DOI:10.1016/j.bmcl.2008.01.117

The discovery, synthesis and in vitro activity of a novel series of rhodanine based phosphodiesterase-4 (PDE4) inhibitors is described. Structure-activity relationship studies directed toward improving potency led to the development of submicromolar inhibitors 2n and 3i (IC₅₀ = 0.89 & 0.74 μM). The replacement of rhodanine with structurally related heterocycles was also investigated and led to the synthesis of pseudothiohydantoin 7 (IC₅₀ = 0.31 μM).

Full text of DOI:10.1016/j.bmcl.2008.01.117

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 18 (2008) 2032–2037
Rhodanine derivatives as novel inhibitors of PDE4
Mark W. Irvine,a,* Graham L. Patrick,a Justin Kewney,b
Stuart F. Hastingsb and Simon J. MacKenzieb
aDepartment of Chemistry, University of Paisley, High Street, Paisley, Renfrewshire PA1 2BE, UK
bScottish Biomedical, Todd Campus, West of Scotland Science Park, Glasgow G20 0XA, UK
Received 9 November 2007; revised 29 January 2008; accepted 29 January 2008
Available online 2 February 2008
Abstract—The discovery, synthesis and in vitro activity of a novel series of rhodanine based phosphodiesterase-4 (PDE4) inhibitors
is described. Structure-activity relationship studies directed toward improving potency led to the development of submicromolar
inhibitors 2n and 3i (IC50 = 0.89 & 0.74 lM). The replacement of rhodanine with structurally related heterocycles was also inves-
tigated and led to the synthesis of pseudothiohydantoin 7 (IC50 = 0.31 lM).
Ó 2008 Elsevier Ltd. All rights reserved.
The phosphodiesterase (PDE) enzymes are responsible
for hydrolyzing secondary messengers cAMP and
cGMP to their 50-monophosphate counterparts. Phos-
phodiesterase type 4 (PDE4) is a cAMP-specific PDE
expressed in inflammatory and airway smooth muscle
cells. By elevating intracellular levels of cAMP, PDE4
inhibition can downregulate the inflammatory response
and induce smooth muscle relaxation.1 With anti-
inflammatory and bronchodilatory activity, PDE4
inhibitors have attracted considerable interest as thera-
peutic agents for inflammatory diseases such as asthma
In a search for novel inhibitors, the Scottish Biomedical
and COPD (chronic obstructive pulmonary disease).2
lead generation library was scanned against PDE4 via
high throughout screening. Prominent amongst the
However, the development of 1st generation inhibitors
resultant hits was a series of derivatives containing a
(e.g. rolipram) was precluded, due to the propensity of
5-benzylidenerhodanine7 substructure (Fig. 1). Ben-
these compounds to induce dose-limiting side-effects of
zylidenerhodanines have been reported as small
molecule inhibitors of numerous targets such as cycloox-
nausea and vomiting.2d,3 Several 2nd generation inhibi-
tors appear to be better tolerated, with the most ad-
ygenase and 5-lipoxygenase,8a b-lactamase,8b cathepsin
vanced to date Òbeing cilomilast (ArifloÒ)4 and
D,8c HCV NS3 protease,8d aldose reductase,8e protein
roflumilast (Daxas ).5 Both of these compounds have
mannosyl transferase8f and JNK-stimulating phospha-
completed phase III clinical trials and are presently
tase.8g Herein, we report the synthesis and structure-
awaiting regulatory approval for the treatment of asth-
ma and/or COPD.6
activity relationships (SAR) for a series of ben-
zylidenerhodanines which inhibit PDE4 in the
submicromolar range.
Keywords: Rhodanine; Phosphodiesterase 4; PDE4 inhibitor; Asthma;
COPD.
*
Corresponding author. Tel.: +44 141 848 3209; fax: +44 141 848
3204; e-mail: mark.irvine@ntlworld.com
Figure 1. 5-Benzylidenerhodanine substructure.
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.01.117

M. W. Irvine et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2032–2037
2033
Scheme 1. Reagents and conditions: (a) RCHO, NaOAc, AcOH,
reflux, 3 h or RCHO, piperidine, EtOH, reflux, 1.5 h.
Scheme 4. Reagents and conditions: (a) activated silica gel 60, toluene,
80–85 °C, 24 h, in dark; (b) NH4OH, NH4Cl, EtOH, reflux.
However, due to the weak acidity of its a-hydrogens, 8
could not be directly condensed with the requisite benz-
aldehyde. Instead, this compound was initially reacted
with acetic anhydride to generate 9. The N-acetyl group
is electron-withdrawing and facilitates the condensation
reaction by enhancing the acidity of the a-hydrogens.
This activating group was easily removed from the prod-
ucts in situ by using an excess of base during the
condensation.10
Scheme 2. Reagents and conditions: (a) RCHO, NaOAc, AcOH,
reflux, 4 h; (b) Bz–Br, Cs2CO3, NaI, DMF, 60 °C, 19 h.
Modification of the exocyclic alkene linker was of obvi-
ous interest for SAR studies (Scheme 4). Reduction of
the alkene moiety in 2n (yielding 12) was achieved in
good yield using Hantzsch ester 11.11 Despite employing
conditions previously reported as suitable for condens-
ing rhodanine with ketones,12 we were unable to synthe-
size compound 14 from rhodanine 1 (R1@H) and 3,4-
dimethoxyacetophenone 13. The exact reason for the
failure of this reaction is unknown although steric and
electronic factors may be involved.
The benzylidenerhodanines in Tables 1 and 2 were syn-
thesized in excellent yield via a Knoevenagel type con-
densation between the requisite rhodanine 1 and
benzaldehyde (Scheme 1). Those benzaldehydes not
available commercially were synthesized via well docu-
mented procedures.
Compounds 6a, 6b and 7 were synthesized from 2,4-thia-
zolidinedione 4 and pseudothiohydantoin9 5 in an almost
identical fashion to their rhodanine counterparts (Scheme
2). The N3 position of 6b was alkylated under standard
conditions to afford 6c in good yield (Scheme 2).
All compounds prepared were evaluated for their
in vitro activity against PDE4B and PDE3A using the
enzymatic assay described previously.13 The respective
IC50 values are shown in Tables 1–3.
3-Benzylidene-2-pyrrolidinone derivatives 10a and 10b
were synthesized from 2-pyrrolidinone 8 (Scheme 3).
To establish basic SAR information, we initially synthe-
sized simple rhodanine derivatives with either a mono-
or di-substituted benzylidene ring. The only active
PDE4 inhibitor to be generated via this strategy was
the 30,40-dimethoxy derivative 2a (IC50 = 17.90 lM). In
addition to PDE4, 2a displayed activity against PDE3
(IC50 = 10.50 lM), an enzyme whose inhibition has been
linked with cardiotoxicity.1 In an effort to increase both
potency and selectivity toward PDE4, the main struc-
tural areas of 2a (i.e. rhodanine ring, catechol subunit
and alkene linker) were subjected to chemical
modification.
Scheme 3. Reagents and conditions: (a) MeONa or EtONa, RCHO,
toluene; (b) Ac2O, reflux, 1.5 h; (c) RCHO, NaH, THF, 0–10 °C.

2034
M. W. Irvine et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2032–2037
proved optimal at the 40-position with the introduction
of larger alkoxy groups resulting in marked decreases
in inhibitor potency.
Table 1. Modification of the catechol subunit
To investigate the binding role of the catechol oxygens
we synthesized the 30,40-methylenedioxy derivative 2s.
The alkoxy oxygens in 2a have the ability to act as
hydrogen bond acceptors (HBAs), meaning the spatial
orientation of the lone pairs on these atoms and the sub-
sequent dipoles they create will be critical to the success
of any such interaction with the active site. Like roli-
pram,20 the dipoles created by the catechol lone pairs
in 2a appear to be orientated in the same general direc-
tion (Fig. 2).
Compound R1
R2
R3
PDE4B
a
PDE3A
a
IC50 (lM) IC50 (lM)
2a15
2b15
2c
2d15
2e16
2f17
2g15
2h15
2i18
2j
3-OMe 4-OMe
H
H
H
H
H
H
H
H
H
17.90
NI
NI
10.50
NT
NT
NT
NT
NT
NT
NT
NT
9.46
39.81
80
H
3-OMe
H
H
4-OMe
4-OH
4-OMe
NI
NI
3-OH
3-OH
NI
NI
3-OMe 4-OH
2-OMe 3-OMe
54.65
However, the cyclic framework in 2s will undoubtedly
force the dipoles in this compound to be directed away
3-OMe 4-OMe 5-OMe 64.82
3-OMe 4-OEt
3-OEt
3-OEt
H
H
H
H
H
6.91
5.30
6.51
3.08
0.89
30.42
1.94
NI
21
from each other (Fig. 2). This modification abolishes
inhibitory activity and suggests that the alkoxy oxygens
are indeed participating in hydrogen bonding.
2k
4-OMe
4-OEt
2l15
2m
3-OAllyl 4-OMe
3-OcPent 4-OMe
NI
2n
2o
10.83
28.50
NI
3-OMe 4-OcPent H
The preceding modifications suggest that when the ben-
zylidenerhodanines interact with PDE4, the alkoxy
groups bind within two hydrophobic pockets while the
catechol oxygens participate in hydrogen bonding with
a suitable donor residue (Fig. 3). This hypothesized
binding mode is in agreement with published crystal
structures of PDE4 in complex with rolipram and other
catechol ether inhibitors.22
2p
3-OcHex 4-OMe
3-OBz 4-OMe
H
H
H
H
2q19
2r
NT
24.99
NT
NI
3-OcPent 4-OEt
3,4-OCH2O–
1.85
NI
2s15
Rolipram
2.00
NI, no inhibition.
NT, not tested.
a Mean of at least two experiments.
After the catechol subunit, we turned our attention to
the rhodanine ring. Two modifications were made to
this area with the first being the introduction of N-al-
kyl substitution (Table 2). As shown, the N-methyl,
N-ethyl, N-allyl and N-carboxymethyl derivatives of
2a were all devoid of activity (3a–3d). The introduc-
tion of N-alkyl substitution was also detrimental to
2n, although it is interesting to note that the resultant
compounds (3e–3g) were inactive against PDE3. The
poor activity of the N-alkylated derivatives suggests
that the NH proton may be participating in hydrogen
bonding. Whilst detrimental to the preceding com-
pounds, N-alkyl substitution proved hugely beneficial
to 2f and 2g. For example, the introduction of an
N-ethyl group converted 2f into an active, albeit rela-
tively weak, PDE4 inhibitor (3h). The increase was
even more dramatic for 2g with an N-ethyl moiety
transforming this previously inactive compound into
a submicromolar inhibitor (3i). The activity displayed
by these compounds was surprising and difficult to
explain considering the inactivity of their parent struc-
tures and that N-alkyl substitution had proved detri-
mental to 2a and 2n. However, it is possible that
the N-ethyl groups may allow the rhodanine moieties
in 3h and 3i to mimic the binding mode of the cate-
chol subunit in 2n. In this theoretical binding mode,
the N-ethyl group would occupy the same space as
the 30-cyclopentyloxy moiety whilst the rhodanine car-
bonyl would mimic the hydrogen-bonding role of the
catechol oxygens (Fig. 4). An interesting point to note
is that methylation of the hydroxyl groups in 3h and
3i (generating 3b in each case) abolishes all activity.
This observation provides strong evidence that the
Table 2. Effect of N-alkyl substitution
Compound R1
R2
R3
PDE4B
a
PDE3A
a
IC50 (lM) IC50 (lM)
3a23
3b
OMe
OMe Me
OMe Et
OMe Allyl
NI
NI
NI
NT
NT
NT
NT
NI
OMe
OMe
OMe
3c
3d
OMe CH2CO2H NI
3e
OcPent OMe Me
OcPent OMe Allyl
2.87
4.13
3f
3g
NI
NI
OcPent OMe CH2CO2H 3.07
OMe Et
OH Et
3h
3i24
OH
OMe
11.20
0.74
NI
NI
NI, no inhibition.
NT, not tested.
a Mean of at least two experiments.
The first area that we focused on was the catechol sub-
unit. Modifications to this region (Table 1) established a
SAR similar to that previously elucidated for the cate-
chol subunit in rolipram.14 Briefly, both methoxy groups
were essential for activity with a 30,40-dialkoxy substitu-
tion pattern proving optimal. At the 30-position, increas-
ing the size of the alkoxy substituent significantly
enhanced activity with the highest potency residing in
the cyclopentyloxy derivative 2n. In contrast, a methoxy

M. W. Irvine et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2032–2037
Table 3. Replacement of rhodanine heterocycle
2035
Compound
PDE4B
a
PDE3A
a
IC50 (lM) IC50 (lM)
6a25
14.10
1.09
NI
NI
6b
Figure 3. Proposed binding of 2n with PDE4 active site.
6c
NI
NT
NI
NI
6c, which was synthesized to further explore the space
surrounding the N3 position, was inactive. Since the
thioketone in 2a and 2n could be replaced without ad-
versely affecting activity, we decided to investigate the
complete removal of this moiety. Interestingly, 2-pyr-
rolidinone derivatives 10a and 10b were 1.6- and 4.5-
fold weaker than their rhodanine counterparts. This
was somewhat surprising as the activities of 6a, 6b
and 7 suggested that the thioketone was not essential
for inhibitory activity. However, the drop in potency
may simply stem from 10a and 10b adopting a slightly
altered binding conformation due to a steric clash be-
tween the H4 and H20 hydrogens.
7
0.31
28.30
10a10
10b
4.01
NI
The alkene linker between the catechol and heterocyclic
rings was the final area to undergo modification
(Scheme 4). Reduction of the alkene moiety in 2n yielded
benzylrhodanine 12, which was some 2.5-fold weaker
than its unsaturated parent (IC50 = 2.20lM). This drop
in activity was not particularly surprising as the greater
flexibility of the benzyl linker in 12 will allow this com-
pound to adopt more binding conformations than its ri-
gid benzylidene counterpart. However, it is also possible
that the drop in activity is due to a loss in conjugation
between the catechol and heterocyclic rings. For exam-
ple, the 40-methoxy group in 2n can increase the electron
density on the carbonyl oxygen through resonance
(Fig. 5), thereby enhancing this group’s ability to act
as a HBA. Reducing the alkene linker will prevent this
movement of electron density and thus weaken any
NI, no inhibition.
NT, not tested.
a Mean of at least two experiments.
hydroxyl groups in these compounds are acting as
hydrogen-bond donors (HBDs).
Next, we replaced rhodanine with different heterocyclic
moieties (Table 3). Whilst thiazolidinedione derivatives
6a and 6b were equipotent with their rhodanine coun-
terparts, pseudothiohydantoin 7 was 3-fold more ac-
tive than 2n and some 58-fold more potent than
original lead 2a. Interestingly, all three compounds
were inactive against PDE3. The N-benzyl derivative
Figure 2. Dipole orientation in 2a and 2s.
Figure 4. Proposed binding of 3i with PDE4 active site.

2036
M. W. Irvine et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2032–2037
Figure 5. Movement of electron density in 2n.
Wollin, L.; Hatzelmann, A.; Beume, R. J. Pharmacol.
Exp. Ther. 2001, 297, 280.
6. Lipworth, B. J. Lancet 2005, 365, 167.
7. IUPAC name for rhodanine is 2-thioxothiazolidin-4-
one.
hydrogen bond between the carbonyl oxygen in 12 and
the PDE4 active site. The conjugation effect in question
is clearly observed by comparing the carbonyl stretching
frequencies for the two compounds. For example, the
tC@O for 2n appears at 1688 cmꢀ1 with the correspond-
ing absorption for 12 occurring some 40 cmꢀ1 higher at
8. (a) Boschelli, D. H.; Connor, D. T.; Kuipers, P. J.; Wright,
C. D. Bioorg. Med. Chem. Lett. 1992, 2, 705; (b) Grant, E.
B.; Guiadeen, D.; Baum, E. Z.; Foleno, B. D.; Jin, H.;
Montenegro, D. A.; Nelson, E. A.; Bush, K.; Hlasta, D. J.
Bioorg. Med. Chem. Lett. 2000, 10, 2179; (c) Whitesitt, C.
A.; Simon, R. L.; Reel, J. K.; Sigmund, S. K.; Phillips, M.
L.; Shadle, J. K.; Heinz, L. J.; Koppel, G. A.; Hunden, D.
C.; Lifer, S. L.; Berry, D.; Ray, J.; Little, S. P.; Liu, X.;
Marshall, W. S.; Panetta, J. A. Bioorg. Med. Chem. Lett.
1996, 6, 2157; (d) Sing, W. T.; Lee, C. L.; Yeo, S. L.; Lim,
S. P.; Sim, M. M. Bioorg. Med. Chem. Lett. 2001, 11, 91;
(e) Fujishima, H.; Tsubota, K. Br. J. Ophthalmol. 2002,
86, 860; (f) Orchard, M. G.; Neuss, J. C.; Galley, C. M. S.;
Carr, A.; Porter, D. W.; Smith, P.; Scopes, D. I. C.;
Haydon, D.; Vousden, K.; Stubberfield, C. R.; Young, K.;
Page, M. Bioorg. Med. Chem. Lett. 2004, 14, 3975; (g)
Cutshall, N. S.; O’Day, C.; Prezhdo, M. Bioorg. Med.
Chem. Lett. 2005, 15, 3374.
1728 cmꢀ1
.
In conclusion, we have identified a series of rhodanine
derivatives as novel inhibitors of PDE4. Structures 2n
and 3i displayed the most significant activity of the
compounds synthesized, being some 20- and 24-fold
more potent than lead compound 2a. Replacing the
rhodanine ring system with different heterocycles also
generated active inhibitors with the most potent of
these being pseudothiohydantoin 7. Further research
into PDE4 inhibitors structurally related to compound
7 is currently on-going and will be published in due
course.
9. IUPAC name for pseudothiohydantoin is 2-thioxo-imi-
dazolidin-4-one.
Acknowledgments
10. Zimmer, H.; Armbruster, D. C.; Trauth, L. J. J. Hetero-
cyclic Chem. 1965, 2, 171.
11. Phillips, M. L.; Berry, D. M.; Panetta, J. A. J. Org. Chem.
1992, 57, 4047.
12. Brown, F. C.; Bradsher, C. K.; McCallum, S. G.; Potter,
M. J. Org. Chem. 1950, 15, 174.
We thank Prof. Donald Cairns, Dr. Ciaran Ewins and
Mr. Daniel Spinks for their helpful comments and
meaningful discussions during the preparation of the
manuscript.
13. Gibson, L. C. D.; Hastings, S. F.; McPhee, I.; Clayton, R.
A.; Darroch, C. E.; Mackenzie, A.; Mackenzie, F. L.;
Nagasawa, M.; Stevens, P. A.; Mackenzie, S. J. Eur. J.
Pharmacol. 2006, 538, 39.
14. Marivet, M. C.; Bourguignon, J.-J.; Lugnier, C.; Mann,
A.; Stoclet, J.-C.; Wermuth, C.-G. J. Med. Chem. 1989,
32, 1450.
15. Brown, F. C.; Bradsher, C. K.; Bond, S. M. Ind. Eng.
Chem. 1953, 45, 1030.
16. Scott, A. W.; Robbins, T. E. Ind. Eng. Chem., Anal. Ed.
1942, 14, 206.
References and notes
1. (a) Torphy, T. J. Am. J. Respir. Crit. Care Med. 1998, 157,
351; (b) Souness, J. E.; Aldous, D.; Sargent, C. Immuno-
pharmacology 2000, 47, 127.
2. (a) Cavalla, D.; Frith, R. Curr. Med. Chem. 1995, 2, 561;
(b) Lombardo, L. J. Curr. Pharm. Design 1995, 1, 255; (c)
Dal Paiz, V.; Giovannoni, M. P. Eur. J. Med. Chem. 2000,
35, 463; (d) Campillo, N.; Paez, J. A. Expert Opin. Ther.
Patents 2002, 12, 53; (e) Burnouf, C.; Devillers, I.;
Pruniaux, M.-P. Frontiers Med. Chem. Online 2004, 1, 293.
3. (a) Norman, P. Expert Opin. Ther. Patents 2002, 12, 93;
(b) Montana, J. G.; Dyke, H. J. Expert Opin. Invest. Drugs
2002, 11, 1.
17. Grundon, M. F.; Perry, H. J. H. J. Chem. Soc. 1954,
3531.
18. Campaigne, E.; Kreighbaum, W. E. J. Org. Chem. 1961,
26, 1326.
19. Wong, E. J. Org. Chem. 1963, 28, 2336.
20. Stafford, J. A.; Feldman, P. L.; Marron, B. E.; Schoenen,
F. J.; Valvano, N. L.; Unwalla, R. J.; Domanico, P. L.;
Brawley, E. S.; Leesnitzer, M. A.; Rose, D. A.; Dougherty,
R. W. Bioorg. Med. Chem. Lett. 1994, 4, 1855.
21. The dipole orientation in 2a and 2s was investigated by
generating the global minimum energy conformations of
these compounds within the Insight-II molecular modeling
package (Accelrys, Cambridge, UK).
4. (a) Barnette, M. S.; Christensen, S. B.; Essayan, D. M.;
Grous, M.; Prabhakar, U.; Rush, J. A.; Kagey-Sobotka,
A.; Torphy, T. J. J. Pharmacol. Exp. Ther. 1998, 284, 420;
(b) Griswold, D. E.; Webb, E. F.; Badger, A. M.; Gorycki,
P. D.; Levandoski, P. A.; Barnette, M. A.; Grous, M.;
Christensen, S.; Torphy, T. J. J. Pharmacol. Exp. Ther.
1998, 287, 705.
5. (a) Hatzelmann, A.; Schudt, C. J. Pharmacol. Exp. Ther.
2001, 297, 267; (b) Bundschuh, D. S.; Eltze, M.; Barsig, J.;

M. W. Irvine et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2032–2037
2037
22. (a) Huai, Q.; Wang, H.; Sun, Y.; Kim, H.-W.; Lui, Y.;
Ke, H. Structure 2003, 11, 865; (b) Card, G. L.;
England, B. P.; Suzuki, Y.; Fong, D.; Powell, B.; Lee,
B.; Luu, C.; Tabrizizad, M.; Gillette, S.; Ibrahim, P. N.;
Artis, D. R.; Bollag, G.; Milburn, M. V.; Kim, S.-H.;
Schlessinger, J.; Zhang, K. Y. J. Structure 2004, 12,
2233; (c) Xu, R. X.; Rocque, W. J.; Lambert, M. H.;
Vanderwall, D. E.; Luther, M. A.; Nolte, R. T. J. Mol.
Biol. 2004, 337, 355.
23. Villemin, D.; Alloum, A. B. Phosphorus, Sulfur Silicon
Relat. Elem. 1993, 79, 33.
24. Andreasch, R.; Zipser, A. Monatsh. Chem. 1904, 25, 159.
25. Yang, D.-H.; Chen, Z.-C.; Chen, S.-Y.; Zheng, Q.-G.
J. Chem. Res., Synop. 2003, 6, 330.