Discovery and Optimization of Thiazolidinyl and Pyrrolidinyl Derivatives as Inhaled PDE4 Inhibitors for Respiratory Diseases

Article abstract of DOI:10.1021/acs.jmedchem.7b01044

Phosphodiesterase 4 (PDE4) is a key cAMP-metabolizing enzyme involved in the pathogenesis of inflammatory disease, and its pharmacological inhibition has been shown to exert therapeutic efficacy in chronic obstructive pulmonary disease (COPD). Herein, we describe a drug discovery program aiming at the identification of novel classes of potent PDE4 inhibitors suitable for pulmonary administration. Starting from a previous series of benzoic acid esters, we explored the chemical space in the solvent-exposed region of the enzyme catalytic binding pocket. Extensive structural modifications led to the discovery of a number of heterocycloalkyl esters as potent in vitro PDE4 inhibitors. (S?,S??)-18e and (S?,S??)-22e, in particular, exhibited optimal in vitro ADME and pharmacokinetics properties and dose-dependently counteracted acute lung eosinophilia in an experimental animal model. The optimal biological profile as well as the excellent solid-state properties suggest that both compounds have the potential to be effective topical agents for treating respiratory inflammatory diseases.

Full text of DOI:10.1021/acs.jmedchem.7b01044

Subscriber access provided by Caltech Library
Article
Discovery and optimization of thiazolidinyl- and pyrrolidinyl-
derivatives as inhaled PDE4 inhibitors for respiratory diseases
Laura Carzaniga, Gabriele Amari, Andrea Rizzi, Carmelida Capaldi, Renato De Fanti,
Eleonora Ghidini, Gino Villetti, Chiara Carnini, Nadia Moretto, Fabrizio Facchinetti, Paola
Caruso, Gessica Marchini, Loredana Battipaglia, Riccardo Patacchini, Valentina Cenacchi,
Roberta Volta, Francesco Amadei, Alice Pappani, Silvia Capacchi, Valentina Bagnacani,
Maurizio Delcanale, Paola Puccini, Silvia Catinella, Maurizio Civelli, and Elisabetta Armani
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.7b01044 • Publication Date (Web): 04 Dec 2017
Downloaded from http://pubs.acs.org on December 4, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Discovery and optimization of thiazolidinyl- and
pyrrolidinyl- derivatives as inhaled PDE4 inhibitors for
respiratory diseases
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
,§
§
§
§
§
Laura Carzaniga,* Gabriele Amari, Andrea Rizzi, Carmelida Capaldi, Renato De Fanti,
§
ǁ
ǁ
ǁ
ǁ
Eleonora Ghidini, Gino Villetti, Chiara Carnini, Nadia Moretto, Fabrizio Facchinetti, Paola
ǁ
ǁ
ǁ
#
‡
Caruso, Gessica Marchini, Loredana Battipaglia, Riccardo Patacchini, Valentina Cenacchi,
‡
†
†
†
†
Roberta Volta, Francesco Amadei, Alice Pappani, Silvia Capacchi, Valentina Bagnacani,
§
‡
†
┴
,§
Maurizio Delcanale, Paola Puccini, Silvia Catinella, Maurizio Civelli, and Elisabetta Armani*
§
ǁ
‡
Chemistry Research and Drug Design, Pharmacology and Toxicology, Pharmacokinetics
†
#
Biochemistry and Metabolism, Analytics and Early Formulations, Project Leader Corporate Drug
┴
Development, Corporate Pre-Clinical R&D Director, Chiesi Farmaceutici S.p.A, Nuovo Centro
Ricerche , Largo Belloli 11/a, 43122 Parma, Italy
Laura Carzaniga, Corporate Preclinical R&D, Chemistry Research & Drug Design, Chiesi
Farmaceutici, Largo Belloli 11/a, 43122 Parma, Italy, phone +39 0521 279652, fax +39 0521 279880,
e-mail: l.carzaniga@chiesi.com
Elisabetta Armani, Corporate Preclinical R&D, Chemistry Research & Drug Design, Chiesi
Farmaceutici, Largo Belloli 11/a, 43122 Parma, Italy, tel. +39 0521 279477, Fax +39 0521 279880, e-
mail: e.armani@chiesi.com
RECEIVED DATE (to be automatically inserted after your manuscript is accepted if required
according to the journal that you are submitting your paper to)
ACS Paragon Plus Environment
1

Page 3 of 58
Journal of Medicinal Chemistry
ABSTRACT. Phosphodiesterase 4 (PDE4) is a key cAMP-metabolising enzyme involved in the
pathogenesis of inflammatory disease and its pharmacological inhibition has been shown to exert
therapeutic efficacy in chronic obstructive pulmonary disease (COPD). Herein we describe a drug
discovery program aiming at the identification of novel classes of potent PDE4 inhibitors suitable for
pulmonary administration. Starting from a previous series of benzoic acid esters, we explored the
chemical space in the solvent-exposed region of the enzyme catalytic binding pocket. Extensive
structural modifications led to the discovery of a number of heterocycloalkyl esters as potent in vitro
PDE4 inhibitors. (S*,S**)-18e and (S*,S**)-22e, in particular, exhibited optimal in vitro ADME and
pharmacokinetics properties and dose-dependently counteracted acute lung eosinophilia in an
experimental animal model. The optimal biological profile as well as the excellent solid state properties
suggest that both compounds have the potential to be effective topical agents for treating respiratory
inflammatory diseases.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
INTRODUCTION.
Phosphodiesterases (PDEs) are intracellular enzymes responsible for the degradation of cyclic
nucleotides (cAMP and/or cGMP) to their inactive forms through the cleavage of a phosphodiester
bond. PDEs have been classified in 11 families based on sequence homogeneity, inhibitor sensitivity
and biochemical properties. This large number of PDE splicing variants serve to fine-tune intracellular
levels of cyclic nucleotides which are key intracellular second messengers regulating a wide range of
cellular processes.1
PDE4 family, which consists of 4 subtypes (PDE4A, PDE4B, PDE4C and PDE4D) sharing a highly
conserved catalytic domain, has received particular attention due to the fact that represents the most
prominent cAMP-metabolising enzyme in immune cells, including neutrophils, monocytes,
macrophages and T cells, the main inflammatory cells involved in the pathophysiology of lung
inflammation underlying asthma and chronic obstructive pulmonary diseases (COPD). In particular,
PDE4 plays a key role in regulating the biosynthesis of several pro-inflammatory mediators and in
2
modulating chemotaxis of leukocytes to sites of inflammation. Moreover, PDE4 regulates the function
ACS Paragon Plus Environment
2

Journal of Medicinal Chemistry
Page 4 of 58
of structural lung cells such as fibroblasts, airway smooth muscle, airway epithelium, vascular
endothelium and airway sensory nerves.2
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
These potential beneficial actions of PDE4 inhibitors have been proven in various experimental
models of airway inflammation, thus generating a considerable interest in utilizing PDE4 inhibitors to
target inflammation underlying pulmonary diseases such as COPD and asthma.3,4
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Prototypical PDE4 inhibitors, designed for oral administration (Figure 1), suffered from the high
prevalence of gastrointestinal side effects such as nausea and emesis, which greatly limited their clinical
4
application. Furthermore, the disappointing results of the second generation PDE4 inhibitor cilomilast
(1) (Figure 1) Phase III clinical programme raised major concerns about the therapeutic utility of this
class of compounds in COPD. Indeed, it is likely that the highest tolerated dose of cilomilast (i.e. 15
mg bid) was on the leading edge of the dose–response curve and that the low therapeutic ratio of this
compound prevented efficacy from being realized.5
6
7
The latest generation of oral PDE4 inhibitors such as roflumilast 2 and apremilast 3 seems to have
a better therapeutic index than cilomilast. Indeed, cilomilast is 10-fold more potent in inhibiting PDE4D
8
in comparison with the other PDE4A, PDE4B, and PDE4C isoforms and such subtype selectivity has
9
been associated with emesis-inducing effects of PDE4 inhibitors. Unlike cilomilast, both roflumilast
and apremilast do not preferentially inhibit PDE4D subtype and this might contribute to their improved
therapeutic ratio when compared with cilomilast and possibly other predecessors.
Roflumilast has been licensed in EU and US for once-daily treatment of severe and very severe COPD
associated with chronic bronchitis and frequent exacerbations as add-on to standard of care, making
10
PDE4 inhibitors a reliable anti-inflammatory therapy for COPD, while apremilast has recently been
approved for the treatment of adult patients with active psoriatic arthritis. Furthermore, among oral
PDE4 inhibitors, tetomilast (4) has been reported to be under clinical development in COPD, although
nausea and vomiting are frequent side effects, but the results have not been published yet.11
ACS Paragon Plus Environment
3

Page 5 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 1. Structures of oral PDE4 inhibitors: cilomilast (1), roflumilast (2), apremilast (3) and
tetomilast (4)
However, even roflumilast is fraught with dose-limiting adverse effects like diarrhea, nausea, weight
loss, back pain, headache and psychiatric adverse events, resulting in 9% to 16% of patients
discontinuing therapy. At this stage, it is still unclear whether the dose limiting side effects of
1
2
roflumilast allow the use of doses that provide full therapeutic efficacy in COPD patients. Indeed,
there are no published clinical data on the use of roflumilast at doses higher than 500 µg/day in this
patient population.
This has prompted the search for PDE4 inhibitors with a better therapeutic ratio.
In the respiratory field, a logical way to improve the therapeutic index is to deliver drugs directly into
the lung via the inhalation route. This route of delivery could provide prolonged inhibition of the target
at the therapeutic site of action with limited systemic exposure and consequently lower potential for
target-related side effects. However, only compounds specifically designed for topical treatment can be
successfully administered using the inhalation route. Therefore, a successful medicinal chemistry
strategy aimed at identifying an inhaled PDE4 inhibitor with limited side effects potential should be
directed to obtain: 1) potent anti-inflammatory effects with convincing activity after topical
administration; 2) long persistence in the lung but no significant accumulation; 3) minimal systemic
exposure; 4) production of inactive metabolites.
Some inhaled PDE4 inhibitors that progressed to the clinic have been reported so far, these include
1
3
14
15
AWD 12-281 5 , tofimilast 6 and UK-500,001 7 (Figure 2). These drugs proved to be potent in
vitro PDE4 inhibitors and showed efficacy in preclinical pathology models but no positive effects were
reported in the clinic in patients suffering from moderate to severe COPD and have been discontinued.
ACS Paragon Plus Environment
4

Journal of Medicinal Chemistry
Page 6 of 58
GSK-256066 816,17 a highly potent and selective inhaled PDE4 inhibitor, demonstrated potent and long-
lasting anti-inflammatory effects in preclinical in vivo models. In a phase 2a study in mild asthmatics,
8 (GSK-256066) exhibited inhibition of forced expiratory volume (FEV1) decrease both in the late
asthmatic response (LAR) and in the early asthmatic response (EAR) after 1 week of treatment.
Nevertheless, 8 is no longer reported in the company’s pipeline. To date, CHF 6001 918,19 is the only
example of inhaled PDE4 inhibitor currently under active clinical development for COPD and its
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
efficacy and safety have been proved in a phase 2a study in mild asthmatics where 9 (CHF 6001)
significantly attenuated allergen challenge response.20
Figure 2. Structures of representative inhaled PDE4 inhibitors in clinical development.
More recently, Astra Zeneca and Merck reported the design and discovery of highly potent and pre-
clinically efficacious inhaled PDE4 inhibitors with low emetic potential21,22 as candidates for treatment
of COPD but no clinical development has been reported to date (Figure 3).
Figure 3. Structures of Astra Zeneca (10) and Merck (11) inhaled PDE4 inhibitors
Following a structure-based design we have synthetized and characterized a novel classes of inhaled
PDE4 inhibitors. These efforts have led to the identification of two potent compounds that effectively
ACS Paragon Plus Environment
5

Page 7 of 58
Journal of Medicinal Chemistry
counteracted allergen-induced airway inflammation and showed a pharmacokinetic profile suitable for
lung delivery, when tested upon intratracheal administration in the rat.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Rationale We previously disclosed a class of highly potent benzoic acid esters as PDE4 inhibitors,
2
3
of which 9 is the most representative example. This paper describes the search for a structurally
differentiated backup series to access previously unexplored region of the chemical space and to prepare
for possible unpredictable non-mechanism-based, chemotype-specific liabilities.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
New compounds were designed on the basis of the abovementioned series of esters showing potent
23
PDE4 inhibitory activity. In particular, the crystal structure of the human PDE4B catalytic domain in
2
4
complex with roflumilast was used as starting structure to perform molecular docking experiments of
2
3
the benzoic ester series. Analyzing the binding pose of the unsubstituted inhibitor (S)-14a (Scheme
1
) it was observed that a deep pocket was available for the insertion of substituents not only in the para
2
3
and meta positions, as seen for compound 9 (CHF6001) and its close analogues, but also in the ortho
position of the phenyl ring (Figure 4). Nonetheless, when ortho substitution was addressed in our former
program, we observed a decrease in the cell-based potency or even a complete loss of activity. We
envisaged that alternative central cores could provide the scaffolding of exit ortho-vectors in the
appropriate geometrics to achieve potency while avoiding any cellular drop-off. We therefore focused
our efforts on new structural classes of ortho-substituted compounds.
In additon to this, since we aimed at potential candidates for pulmonary administration, our strategy
targeted new scaffolds whose features could be able to minimize the systemic exsposure due to the
swallowed dose. Foreasmuch as previous benzoate esters23 were poorly metabolized in vitro, we
focused on obtaining molecules endowed with high intrinsic Clearance, a predictive tool for high
systemic Clearance, in addition to low free fraction in human plasma and reduced gut wall permeability.
With this purpose we pursued the replacement of benzoyl group gradually increasing the three-
dimensional morphology of the compound (Aryl  Heteroaryl  Heterocycloalkyl) to probe if it was
feasible maintaining potency while, at the same time, achieving an effectively turned-over series.
ACS Paragon Plus Environment
6

Journal of Medicinal Chemistry
Page 8 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 4. Docking pose of (S)-14a in the catalytic site of the PDE4B crystal structure (PDB ID:
1
XMU). A pocket for the insertion of substituents on the phenyl ring of the ligand is available in the
solvent exposed region of the catalytic binding site (grey).
At first we synthesized a new series of ester derivatives where structural modifications were achieved
functionalizing the ortho position of aromatic rings with sulfonyl and carbonyl substituents.
Encouraging results from these compounds prompted us to expand our exploration towards the
replacement of the aryl esters with 5- and 6-membered heterocycloalkyl rings, investigating the steric
and stereochemical requirements for optimal activity. Newly synthesized compounds were at first
assessed in a cell-free assay and structure-activity relationships of the most promising classes were
further investigated. Our efforts were focused on obtaining derivatives able to retain their activity in
the cell-based assay, overcoming any potency drop-off likely to occur in a cellular system, to identify
a suitable lead compound whose optimization could be profitably developed. In addition we looked for
a candidate able to exert sustained in vivo efficacy in animal models of inflammation through its
persistence in the pulmonary district along with high stability in human lung.
Chemistry. All compounds described in this paper were synthesized starting from common
intermediates (S)-12 and (R)-12, which were prepared as described in ref. 23.
ACS Paragon Plus Environment
7

Page 9 of 58
Journal of Medicinal Chemistry
Scheme 1. Preparation of Compounds 14a
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
aReagents and conditions: (a) EDC, DMAP, DCM or DMF, rt; (b) mCPBA, DCM, rt, 48h.
The synthesis of aryl esters (S)-14a-d is shown in Scheme 1. Esters (S)-14a-c were obtained by
coupling the intermediate (S)-12 with suitable carboxylic acids using EDC and DMAP as condensing
agents, while ester (S)-14d was achieved performing a further oxidation step using mCPBA. Aryl acids
13a-d where either commercially available or prepared as detailed in the Supporting Information.
Scheme 2. Preparation of Compounds 18 and 19a
aReagents and conditions: a) (S)-1-(tert-butoxycarbonyl)pyrrolidine-2-carboxylic acid or (R)-1-(tert-
butoxycarbonyl)pyrrolidine-2-carboxylic acid, EDC, DMAP, DMF, rt, 48 h; b) HCl/Dioxane 4 M, rt,
8
h c) Py or Py/DCM, rt; d) dimethylamine 2M in THF, CDI, DMF, rt.
ACS Paragon Plus Environment
8

Journal of Medicinal Chemistry
Page 10 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
The synthetic pathway for the synthesis of pyrrolidinyl derivatives 18a-f and 19 is depicted in Scheme
2. Chiral alcohol (S)-12 was condensed with either (R)- or (S)- enantiomerically pure Boc-protected
proline, then removal of tert-butoxycarbonyl group under acidic conditions led to the versatile
intermediates (S*,S**)-15 and (R*,S**)-15 which underwent N-sulfonylation to afford sulfonamides
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
8a-f. Further amidation of 18f using CDI as condensing agent afforded amide (S*,S**)-19.
Scheme 3. Preparation of Compounds 22 and 24a
aReagents and conditions: a) (S)-3-(tert-butoxycarbonyl)thiazolidine-2-carboxylic acid or (R)-3-(tert-
butoxycarbonyl)thiazolidine-2-carboxylic acid, EDC, DMAP, DMF, rt; b) HCl/AcOEt 5 M, rt; c) Py,
rt; d) EDC, DMAP, DCM or DMF, rt; e) alkylamine, CDI, DMF, rt.
An analogous synthetic route was followed for N-thiazolidinyl-sulfones 22b-m (Scheme 3).
Thiazolidinyl esters 20 were prepared by coupling either alcohol (S)-12 or (R)-12 with the suitable
thiazolidine-carboxylic acid isomer in the presence of EDC and DMAP followed by Boc cleaveage.
The subsequent sulfonylation in pyridine gave sulfonamides 22b-m. Carboxylic acid 22f was further
converted into amides 22m reacting the appropriate amine with CDI in DMF. Alternatively, amidation
of intermediate 21 with acids 23a-b afforded compounds 24a-b. Sulfonyl chlorides 17 and carboxylic
acids 23 were either commercially available or prepared as described in Supporting Information.
ACS Paragon Plus Environment
9

Page 11 of 58
Journal of Medicinal Chemistry
1
2
Scheme 4. Preparation of Compounds 28a
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
aReagents and conditions: a) (S)-3-(tert-butoxycarbonyl)thiazolidine-2-carboxylic acid, EDC, DMAP,
DCM, rt; b) HCl/AcOEt 4 M, rt; c) Py, DCM, rt; d) TFA, rt, overnight.
A synthetic pathway for the preparation of compounds 28 bearing a differently decorated catecholic
moiety is summarized in Scheme 4. Enantiomerically pure alcohols (S)-25a-b, obtained as described
in Supporting Information, were coupled with the suitable thiazolidine-carboxylic acid isomer in the
presence of EDC and DMAP, the subsequent Boc cleavage led to esters (S*,S**)-27a-b. Sulfonamides
(S*,S**)-28a-b were finally achieved via sulfonylation as described in Scheme 3. A modified procedure
was followed to obtain compound (S*,S**)-28c which was achieved via acidic cleavage of the
catecholic ether of thiazolidinyl ester (S*,S**)-22e, prepared as described in Scheme 3.
Scheme 5. Preparation of Compounds 35, 36 and 37a
ACS Paragon Plus Environment
1
0

Journal of Medicinal Chemistry
Page 12 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
aReagents and conditions: a) (R)-3-(tert-butoxycarbonyl)thiazolidine-4-carboxylic acid or (R)-1-(tert-
butoxycarbonyl)piperidine-3-carboxylic acid or (S)-1-(tert-butoxycarbonyl)piperidine-2-carboxylic
acid, EDC, DMAP, DMF, rt; b) HCl/Dioxane 4 M, rt; c) Py, rt; d) dimethylamine 2 M in THF, CDI,
DMF, rt.
Sulfonamides 35, 36 and 37 were obtained by sulfonylation of the corresponding
heterocycloalkylamino intermediates 32, 33, 34 according to Scheme 5. Intermediates 32, 33, 34 were
prepared by coupling of (S)-12 with suitable N-Boc heterocycloalkyl carboxylic acids followed by
acidic deprotection. (S*,S**)-36b underwent further amidation with dimethylamine to afford (S*,S**)-
36c.
RESULTS AND DISCUSSION
SAR Analysis.
Table 1. Structure and Inhibitory Activity in Cell-Free and Cell-Based Assays of Aryl Esters
Compd.
S)-14ac
R
Cell-free IC50
nM) ±SDa
(
(
0.072
ACS Paragon Plus Environment
1
1

Page 13 of 58
Journal of Medicinal Chemistry
(
S)-14d
S)-14b
0.323±0.085
0.320±0.094
1
2
3
4
5
6
7
8
9
S
O
O
O
(
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
S)-14c
O
0.825±0.043
S
N
a
b
Data are the mean ± SD of three to four experiments performed in duplicate. Data are presented as
c
IC50 values followed by their 95% confidence intervals obtained by nonlinear regression. compound
(S)-26a ref. 23
Our initial hypothesis, based on additional available space in the solvent-exposed region of the
catalytic binding pocket of PDE4, was confirmed by the potency achieved by the first series of
analogues. The phenyl ester (S)-14d and its thienyl (S)-14b and pyrrolyl (S)-14c bioisosteric analogues,
carrying a phenyl-sulfonyl or a phenyl-carbonyl substituent in the position adjacent to the ester function,
showed limited reduction of PDE4 inhibitory potency compared to unsubstituted derivative (S)-14a.
Table 2. Structure and inhibitory activity in cell-free and cell-based assays of pyrrolidinyl esters
featuring a substituent on the Nitrogen Atom.
Compd.
R
Cell-free IC50
nM)
PBMCs IC50
(nM)
(
(
R*,S**)-18c
0.06±0.011
0.4 (0.12 – 1.3)
ACS Paragon Plus Environment
1
2

Journal of Medicinal Chemistry
Page 14 of 58
(
S*,S**)-18c
0.109±0.045
0.34 (0.130 –
.900)
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
(S*,S**)-18a
0.5038±0.080
0.0442±0.006
0.0324±0.027
4.2 (1.500
2.000)
–
–
1
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
S*,S**)-18b
S*,S**)-18d
0.31 (0.21
.46)
0
(
0.36 (0.260 –
.510)
0
(
S*,S**)-19
0.0351±0.020
0.042
.130)
(0.014-
0
(S*,S**)-18e
(R*,S**)-18e
0.0378±0.022
0.0533±0.027
0.032 (0.014 –
0.070)
0.38 (0.200 –
0
.700)
a
b
Data are the mean ± SD of three to four experiments performed induplicate. Data are presented as
IC50 values followed by their 95%confidence intervals obtained by nonlinear regression.
To verify if a subtle geometrical variation in the arrangement of the ester portion could improve the
inhibitory potency, our exploration was then focused on the synthetically easily accessible saturated
analogues of (S)-14c. Therefore a series of pyrrolydinyl analogues was examined, in which ring
unsaturation led to a heterocycloalkanyl core containing an additional stereogenic centre. Both
diastereomers featuring the pyrrolydinyl moiety ((R*,S**)-18c, (S*,S**)-18c) were endowed with
higher potency than their corresponding aromatic analogue (S)-14c, irrespective of their
stereochemistry (Table 2). We then considered mono or bis substitutions at the para and meta position
of the aryl sulfonyl ring. Concurrently we prepared both diastereomers of the meta-dimethylamido
derivative to further study the impact of stereochemistry on potency: (S*,S**)-18e and (R*,S**)-18e.
ACS Paragon Plus Environment
1
3

Page 15 of 58
Journal of Medicinal Chemistry
They exerted similar potency in the cell-free test but the R configuration at the heterocycloalkyl ring
caused a 10-fold drop in the cell-based activity. This finding prompted us to select the stereochemical
configuration of (S*,S**)-18e as the one able to provide the best inhibition of LPS-induced TNF-α
release in the in vitro model of PBMCs. Preparation of (S*,S**)-18d and (S*,S**)-19 allowed us to
infer that, in this series, the m-dimethylaminocarbonyl substituent, coupled with S,S stereochemistry, is
able to enhance by 10-fold the potency in PBMCs. Replacement of the phenyl ring with a methyl group
in (S*,S**)-18a was detrimental, whereas compound (S*,S**)-18b with a longer methoxyethyl
substituent showed interesting PDE4 inhibitory activity.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The potent inhibitor (S*,S**)-18e was crystallized in the PDE4B catalytic site. Its binding mode is
similar to that of roflumilast in the PDB 1XMU crystal structure. Indeed, the catecholic oxygens form
two hydrogen bonds with the side chain of Gln443 residue in the hydrophobic pocket of the binding
site, while the pyridine N-oxide ring is located in the metal region and forms a hydrogen bond with the
backbone NH group of Met347 residue. The phenylsulfonyl-proline fragment is positioned in the
solvent-exposed region of the catalytic site and the carbonyl group of the dimethylaminocarbonyl
substituent interacts with a water molecule, which is indirectly linked to the water molecules
coordinated to the metals (Figure 5). The high PDE4 inhibitory potency of compound (S*,S**)-18b
may be related to its ability to undertake a hydrogen bond with the same water molecule bound to the
amide substituent of (S*,S**)-18e. (Figure 6). The wide surface of the solvent exposed region
accommodates structurally different fragments, which in most cases are able to maintain high binding
energy values.
ACS Paragon Plus Environment
1
4

Journal of Medicinal Chemistry
Page 16 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 5. Crystal structure of compound (S*,S**)-18e bound to the PDE4B catalytic binding site.
Hydrogen bonds are represented as yellow dashed lines.
ACS Paragon Plus Environment
1
5

Page 17 of 58
Journal of Medicinal Chemistry
Figure 6. Overlay of the docking binding pose for compound (S*,S**)-18b (magenta colour) on the
structure of (S*,S**)-18e (orange colour) bound in the PDE4B catalytic site. Hydrogen bonds are
represented as yellow dashed lines.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 3. Structure and inhibitory activity in cell free and cell based assays of thiazolidinyl esters
featuring -a substituent on the Nitrogen Atom.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Compd.
R
Cell-free IC50 (nM) PBMCs IC50 (nM)
O
O
0.043±0.02
0.028±0.01
0.176±0.131
0.078 (0.049 – 0.120)
0.860 (0.360 – 2.000)
0.260 (0.140-0.460)
(S*,S**)-22b
S
CH3
CH3
O
O
O
(S*,S**)-24a
(S*,S**)-22c
O
O
S
O
O
(
S*,S**)-22d
0.041±0.015
0.026±0.010
0.057 (0.010-0.280)
0.022 (0.011-0.041)
S
OCH3
OCH3
(S*,S**)-22m
O
O
O
O
O
S
S
N(CH3)2
OCH3
O
(
S*,S**)-22e
0.053±0.007
nd
0.019 (0.012 – 0.030)
N(CH3)2
O
O
0.570(0.150 – 2.270)
(S*,S**)-24b
N(CH3)2
a
b
Data are the mean ± SD of three to four experiments performed induplicate. Data are presented as
IC50 values followed by their 95%confidence intervals obtained by nonlinear regression. Nd: not
determined.
ACS Paragon Plus Environment
1
6

Journal of Medicinal Chemistry
Page 18 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
To further investigate the class of saturated heterocycles, we explored the SAR of a set of bioisosteric
compounds where we incorporated the thiazolidinyl core thus replacing the pyrrolydinyl one (Table 3).
An aliphatic sulfonyl substituent was well accepted in (S*,S**)-22b, increasing the PBMCs potency of
this compound by 10-fold with respect to its pyrrolydinyl analogues (S*,S**)-18a and (S*,S**)-18b,
whereas the aliphatic carboxamide in (S*,S**)-24a caused a drop in activity in the cell-based assay.
Successively an in-depth examination of the role of the N-aryl-sulfonyl ring decorations was carried
out. The unsubstitued benzensulfonylamide derivative (S*,S**)-22c served as baseline compound and
showed limited potency. The dimethoxy (S*,S**)-22d proved to be very active and, with the
introduction of the amido groups, the inhibitory potency was slightly further enhanced ((S*,S**)-22m,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(S*,S**)-22e). Replacement of the N-sulfonylamido with an N-carboxamido substituent as in (S*,S**)-
2
4b is detrimental for cell-based activity, as already proved by the aliphatic derivative (S*,S**)-24a.
Table 4. Structure and inhibitory activity in cell free and cell based assays of (S*,S**)-22e
analogues featuring different R at the phenolic Oxygen.
Compd.
R
Cell-free IC50 (nM) PBMCs IC50 (nM)
(
S*,S**)-22e
0.053±0.007
0.183±0.119
0.014±0.0004
0.021±0.023
0.019 (0.010
.030)
–
–
–
–
0
(S*,S**)-28c
H
1.950 (1.310
2.890)
(S*,S**)-28a
(S*,S**)-28b
CH
3
0.032 (0.0026
0.04)
0.040 (0.023
0
.070)
ACS Paragon Plus Environment
1
7

Page 19 of 58
Journal of Medicinal Chemistry
a
b
Data are the mean ± SD of three to four experiments performed in duplicate. Data are presented as
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
IC50 values followed by their 95%confidence intervals obtained by nonlinear regression.
(S*,S**)-22e emerged as one of the most potent compounds in the thiazolidinyl series and was chosen
as reference compound to investigate the impact of the alkylation of the catecholic hydroxyl group on
the inhibitory activity, as shown in Table 4. The contribution of cyclopropylmethyl group was clearly
evident in (S*,S**)-28c, where O-dealkylation led to a 3-fold less potent derivative in the cell-free assay
and 100-fold less potent in the cell-based assay, likely because of limited cell penetration. Conversely,
the alkyl moiety replacement in (S*,S**)-28a and (S*,S**)-28b, where methyl and cyclopentyl were
chosen by analogy with rolipram and cilomilast, respectively, restores the inhibitory activity to the
nanomolar range.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Table 5. Structure and inhibitory activity in cell free and cell based assays of the four
diastereomer of (S*,S**)-22e.
Compd.
R
Cell-free IC50
nM)
PBMCs IC50
(nM)
(
(S*,S**)-22e
0.053±0.007
0.019 (0.012
–
–
0
.030)
(R*,S**)-22e
0.047±0.022
0.047 (0.010
.230)
0
ACS Paragon Plus Environment
1
8

Journal of Medicinal Chemistry
Page 20 of 58
(
R*,R**)-22e
6.397±1.041
nd
nd
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(
S*,R**)-22e
3.006±0.482
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
b
Data are the mean ± SD of three to four experiments performed in duplicate. Data are presented as
IC50 values followed by their 95%confidence intervals obtained by nonlinear regression. Nd: not
determined.
We subsequently examined the effect of the stereochemical configuration in thiazolidinyl derivatives
against activity (Table 5). With this aim the three diastereoisomers of (S*,S**)-22e were synthesized.
(R*,S**)-22e maintained comparable affinity and biological activity to the ones exerted by its epimer
(
S*,S**)-22e. Conversely stereocenter inversion of the alcoholic portion ((R*,R**)-22e, (S*,R**)-22e)
caused a remarkable loss of potency: (R*,R**)-22e and (S*,R**)-22e showed 50 and 150-fold drop of
activity compared to their respective enantiomers, thus meaning that there is a stereochemical
2
3
preference for the interaction with PDE4, confirming previous findings, while the configuration of
the asymmetric carbon centre on the thiazolidinyl core is not relevant for the interaction with PDE4.
Accordingly in the thiazolidinyl series the R configuration at the thiazolidine carbon seems to affect
less PBMC potency than in the pyrrolidinyl series.
Table 6. Structure and inhibitory activity in cell free and cell based assays of thiazolidinyl esters
featuring a heteroarylsulfonyl Group on the Nitrogen Atom.
ACS Paragon Plus Environment
1
9

Page 21 of 58
Journal of Medicinal Chemistry
Compd.
R
Cell-free IC50 (nM) PBMCs IC50 (nM)
1
2
3
4
5
6
7
8
9
N
(
S*,S**)-22g
0.061±0.016
0.070 (0.040-0.120)
(
S*,S**)-22h
N
0.020±0.009
0.028 (0.009-0.085)
N
CH3
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
CH3
0.025±0.006
0.176±0.080
0.030 (0.020-0.050)
0.170 (0.080-0.360)
(
S*,S**)-22i
N
O
NHCH3
(S*,S**)-22l
H3C
S
CH3
a
b
Data are the mean ± SD of three to four experiments performed induplicate. Data are presented as
IC50 values followed by their 95%confidence intervals obtained by nonlinear regression.
In the thiazolidinyl series the replacement of the terminal phenyl ring with a set of heteroaryl
bioisosteres was considered, as shown in Table 6. The introduction of pyridine in (S*,S**)-22g and
five-member heteroaryls in (S*,S**)-22h and (S*,S**)-22i led to compounds endowed with higher
potency than the naked phenyl (S*,S**)-22c. In contrast to what observed for the N-phenylsulfonyl
series, no further contribution to the of potency was achieved by heteroaryl ring decoration in (S*,S**)-
2
2i. The decrease of potency for thienyl derivative (S*,S**)-22l observed in both assays may be related
to the steric hindrance exerted by the methyl groups.
Table 7. Structure and inhibitory activity in cell free and cell based assays of heterocycloalkyl
esters featuring an arylsulfonyl Substituent on the Nitrogen Atom.
Compd.
R
Cell-free IC50
nM)
PBMCs IC50
(nM)
(
ACS Paragon Plus Environment
2
0

Journal of Medicinal Chemistry
Page 22 of 58
(
R*,S**)-35
0.055±0.032
0.035
.073)
(0.017-
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
(S*,S**)-36a
0.056±0.013
0.031±0.0004
0.012
.050)
(0.003-
(0.005-
0
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(S*,S**)-36c
0.015
.050)
0
(S*,S**)-36d
0.094±0.041
0.031±0.011
0.203
.540)
(0.100-
(0.030-
0
(R*,S**)-37
0.120
.480)
0
a
b
Data are the mean ± SD of three to four experiments performed in duplicate. Data are presented as
IC50 values followed by their 95%confidence intervals obtained by nonlinear regression.
Finally, as part of our core scaffold modifications, we extended our investigations to additional
saturated cyclic systems which deserved a more extensive SAR analysis. We therefore performed a
heterocycloalkyl swap introducing a small subset of 5- and 6-atoms cyclic bioisosteres, while
maintaining almost unmodified the N-arylsulfonyl group (Table 7). No substantial changes in terms of
potency were observed for the thioprolinyl analogue in (R*,S**)-35 and piperidines (S*,S**)-36a and
(S*,S**)-36c, suggesting that neither the absence nor the position of the sulfur atom seems to affect
their biological activity. Preparation of (S*,S**)-36d confirmed that the contribution of the amide
decoration on the phenyl ring is fundamental to achieve the wanted potency range, especially in the
cellular system. Comparable activity to 2-substitued analogues is displayed by compounds (R*,S**)-
3
7 where the nitrogen was moved from position 2 to 3.
In vitro ADME Characterization.
Table 8. ADME profile of selected compounds.
ACS Paragon Plus Environment
2
1

Page 23 of 58
Journal of Medicinal Chemistry
Compd.
CACO-2 permeability
hPPBb Hepatic
Hepatocytesd hLung S9 hPlasma
stability stability
1
c
microsomes
2
3
4
5
6
7
8
AB
BA
Caco-2 bound Clearance
Clearance
60
remain.
(%)
min 60
min
^
6
(
nm/sec) (nm/sec) perm.
(%)
(L/min/mg) (L/min/10
remain.
(%)
a
Level
Low
Low
Low
Low
Low
cell)
(
S*,S**)-18e 4.8
101.8
124.9
76.7
31.8
8.2
99.65 590.56
99.68 365.35
99.83 360.36
99.79 287.41
99.60 200.19
95.50
56.67
47.56
82.45
53.76
47.62
71.1
71.6
71.2
70.2
50.7
63.5
108.40
101.06
98.78
9
1
1
1
(
1
1
1
1
1
1
1
2
2
0
1
2
3
4
(
S*,S**)-19
5
S*,S**)-22m 2.3
(5 S*,S**)-22e 4.2
97.96
6
(7 S*,S**)-22d 2.5
100.60
100.25
8
9( S*,S**)-22i 12.6
182.7
Low/M 99.84 202.10
edium
0
1
2
(2 S*,S**)-22h 2.2
39.5
44.6
734
684.3
nd
Low
98.50 281.17
66.86
25.91
64.57
33.16
nd
77.7
52.1
72.5
59.7
nd
101.10
93.99
104.50
88.22
nd
2
2
2
3
4
5
R*,S**)-35 28.0
Medium 99.83 267.87
(
2
6
7
(
S*,S**)-36a 217.4
High
High
nd
99.88 362.74
99.77 243.50
2
2
8
(
S*,S**)-36c 131.8
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
S)-32ke
S)-26pe
nd
nd
nd
nd
nd
nd
35.22
50.94
71.24
(
nd
nd
nd
nd
nd
(
S)-38ee
nd
nd
nd
nd
nd
a
General absorption classification for permeability values: <10 nm/s, low level; 10-50 nm/s, medium
b
c
level; >50 nm/s, high level. hPPB = human plasma protein binding. Compounds were incubated with
d
human liver microsomes at 1 μM for 60 min. Compounds were incubated with human hepatocytes at
e
1
μM for 90 min. Ref.23. Nd = not determined.
Through a funnel-like process, a subset of 10 compounds displaying excellent inhibitory activity on
PDE4 and in the PBMC assay was chosen for further characterization and their in vitro ADME profiles
are captured in Table 8. The selection was based on the criteria that both IC50 values on the free enzyme
and in cell-based test should be less than or equal to 0.05 nM. Almost all compounds met the desired
criteria to be progressed further as candidates for inhaled administration, showing i) adequate metabolic
stability in human lung S9 preparation, ii) low in vitro permeability in the Caco2 assay, except for the
piperidinyl-derivatives (S*,S**)-36a and (S*,S**)-36c that were susceptible to high efflux, and iii)
ACS Paragon Plus Environment
2
2

Journal of Medicinal Chemistry
Page 24 of 58
human plasma protein binding values higher than 99%, excluding (S*,S**)-22h. The aforementioned
ADME profile strongly reduces the likelihood that the free fraction in the systemic circulation, absorbed
through the lung compartment or by the GI tract as a consequence of the swallowed fraction, could
have an impact on araising unwanted side effects. Additionally, the selected compounds were stable in
human plasma and showed favorable intrinsic Clearance (CLint) values, allowing us to infer that such
metabolic liability is likely to translate into the desired sustained hepatic turnover, thus furnishing the
sought-after prerogatives for a minimized systemic exposure. A comparison between in vitro Clearance
data for the heterocycloalkyl series herein described and the benzoic ester series, whose values of
Clearance in hepatic microsomes for three representative compounds (S)-32k, (S)-26p and (S)-38e23
are listed in Table 8, clearly highlighted that we were able to reach the goal of obtaining effectively
turned over compounds while maintaining potency.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Since physico-chemical properties are critical for an inhalation drug, the choice to progress
compounds (S*,S**)-18e and (S*,S**)-22e to further in vivo investigations relied on their ability to
better perform in solid state studies. In particular, highly crystalline, anhydrous, not hygroscopic forms
were identified for both (S*,S**)-18e and (S*,S**)-22e by means of proper crystallisation screening
and experiments (data not shown). Both compounds were successfully micronised by jet mill with
acceptable yields, good physico-chemical integrity and allowing to obtain a particle size suitable for
inhalation. They were then blended with lactose at strengths suitable for animal administration.
In Vitro and In Vivo Characterization of Compounds (S*,S**)-18e and (S*,S**)-22e.
(S*,S**)-18e and (S*,S**)-22e, administered as dry-powder formulation 2 hours before the aerosolized
ovalbumin (OVA) challenge, were screened in a dose-response study in Brown Norway rat OVA-
induced airway inflammation model via intratracheal (IT) dosing (Figure 7). Both compounds
displayed a dose-dependent pattern and markedly inhibited inflammatory cells infiltration in the
bronchoalveolar lavage (BAL), exerting a 50% inhibition of eosinophils recruitment at 0.1 mol/kg,
while at 1 mol/kg (S*,S**)-22e produced a higher efficacy in reducing cell infiltration than (S*,S**)-
1
8e. Estimated values of ED50 of 0.10 and 0.04 mol/kg for (S*,S**)-18e and (S*,S**)-22e,
respectively, were calculated from the dose-response curves of inhibition of eosinophils recruitment.
ACS Paragon Plus Environment
2
3

Page 25 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 7. Dose-dependent inhibition of OVA-induced recruitment of eosinophils in BALF (broncho
alveolar lavage fluid) of rats treated with compound (S*,S**)-18e and (S*,S**)-22e. The reference
antinflammatory compound budesonide is reported for comparison.
Table 9. Structure and Inhibitory Activity in Cell-Based Assay of Metabolites obtained after 60
min incubation of (S*,S**)-22e (compounds 38-40) and (S*,S**)-18e (compounds 32, 35) with
human hepatocytes at a concentration of 10 M.
PBMC
%
of total drug-
Compd.
Structure
related material
IC50 nM [CI]
0.019
(S*,S**)-22e 51
[0.003-0.020]
4
79.6
3
3
8
9
12
11
[
288.1-798.7]
1
.95
[
1.31-2.89]
ACS Paragon Plus Environment
2
4

Journal of Medicinal Chemistry
Page 26 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
40
>1µM
0.032
(
S*,S**)-18e 78
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
79.6
3
4
8
1
9
[
288.1-798.7]
>1µM
The in vitro metabolite profiles of (S*,S**)-18e and (S*,S**)-22e were studied using human
hepatocytes and the main metabolites detected are depicted in Table 9. For (S*,S**)-18e two
metabolites were identified in a percentage higher than 10% (38 and 39) and assessed in the PBMCs
assay, exhibiting a 50000-fold and a 300-fold lower potency than the parent molecule, respectively.
Considering that alcohol 38 stems from parent compound ester hydrolysis, the carboxylic acid fragment
40, whose presence in vivo in an amount comparable to the alcoholic fragment can be arguably
assumed, was prepared and tested as well, proving to be inactive with an IC50 lower than 1 M.
Comparably to (S*,S**)-22e, compound (S*,S**)-18e generates a substantial amount of alcohol 38. Its
corresponding acid 41 showed to be unable to inhibit TNF- release in our cellular model (IC50 > 1
M).
Table 10. Pharmacokinetic parameters of (S*,S**)-18e and (S*,S**)-22e determined from lung
tissues following IT administration as dry powder in rats at 1 µmole/kg.
Parameter (S*,S**)-22e
(S*,S**)-18e
ACS Paragon Plus Environment
2
5

Page 27 of 58
Journal of Medicinal Chemistry
C
max
1
2
3
4
5
6
7
8
9
16.06
16.24
(nmol/g)
T
T
T
max (h)
last (h)
1/2 (h)
0.250
24
0.083
24
4.39
5.18
AUClast
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
65.27
49.65
4.52
.
.
(hr nmol/g)
MRTlast(h)
5.56
Table 11. Pharmacokinetic parameters of (S*,S**)-18e and (S*,S**)-22e determined from plasma
following IT administration as dry powder in rats at 1 µmole/kg.
Parameter
(S*,S**)-22e
(S*,S**)-18e
C
max
9.80
20.65
(
pmol/mL)
T
T
T
max (h)
last (h)
1/2 (h)
0.083
8
0.083
8
nc
nc
AUClast
hr pmol/mL)
1
0.38
33.27
1.08
..
(
MRTlast(h)
1.33
(
S*,S**)-18e and (S*,S**)-22e exhibited ideal in vivo pharmacokinetic profiles for an inhaled agent.
Lung levels, plasma levels and pharmacokinetic parameters following IT treatment with (S*,S**)-18e
and (S*,S**)-22e in Sprague-Dawley rats are shown in Figure 8, Table 10 and Table 11. (S*,S**)-18e
and (S*,S**)-22e showed a lung half-life of around 5 h which is in line with the corresponding mean
residence time (MRT) (Table 10). Both compounds displayed prolonged lung retention (0.128 and
0
.220 nmol/g concentration at 24 hours, for (S*,S**)-18e and (S*,S**)-22e , respectively), indicating
that both molecules could deliver a long-lasting protective effect. Additionally, being optimized for
inhalation, a key pharmacokinetic measure of success was having a high exposure ratio, at the same
T
last, between lung and plasma. In fact the exposures calculated until 8 hours clearly indicated that
(
S*,S**)-18e and (S*,S**)-22e were mainly distributed in lung compared to plasma (2261 and 5152-
fold ratio for (S*,S**)-18e and (S*,S**)-22e, respectively) (Tables 10 and 11).
ACS Paragon Plus Environment
2
6

Journal of Medicinal Chemistry
Page 28 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(S *,S **)-2 2 e (lu n g )
(
S *,S **)-1 8 e (lu n g )
(
S *,S **)-2 2 e (p la s m a )
1 0 0
(S *,S **)-1 8 e (p la s m a )
1
0 0
1 0
1
1
0
1
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
0
.1
0 .1
0
4
8
1 2
1 6
2 0
2 4
0
4
8
1 2
1 6
2 0
2 4
T im e (h )
T im e (h )
Figure 8. Lung levels (nmol/g) and plasma levels (pmol/ml) of (S*,S**)-22e (left) and (S*,S**)-18e
(right) after IT administration as dry powder in Sprague-Dawley rats at 1 µmole/kg.
Table 12. Isoform Selectivity Profile of (S*,S**)-18e and (S*,S**)-22e
Target
hPDE1A
hPDE2A
hPDE3A
hPDE4B
hPDE5A
Bov PDE6
hPDE7A
hPDE8A1
hPDE9A2
hPDE10A2
(S*,S**) 18ea (S*,S**) 22ea
25.6 µM
>30 µM
>30 µM
7.40 µM
>30 µM
6.21 µM
-5
-5
8.81x10 µM 2.78x10 µM
13.9 µM
>30 µM
3.46 µM
17.0 µM
>30 µM
4.03 µM
4.61 µM
5.78 µM
0.66 µM
3.56 µM
>30 µM
0.81 µM
ACS Paragon Plus Environment
2
7

Page 29 of 58
Journal of Medicinal Chemistry
hPDE11A4
>30 µM
>30 µM
1
2
3
a
Data are given as mean IC50 values (mol/l) calculated from concentration-inhibition curves by
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
nonlinear regression analysis from 2 independent experiments
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(S*,S**)-18e and (S*,S**)-22e were subsequently evaluated in safety assessment studies. There was
no evidence of inhibition of hERG potassium channels current at a concentration of 1 M (data not
shown). Both (S*,S**)-18e and (S*,S**)-22e proved to be highly selective inhibitors for PDE4
(>10,000 versus PDE1, 3, 5, 7, 8, 10 and > 100,000 vs PDE2, 6, 9, 11) (Table 12). The propensity of
(S*,S**)-18e and (S*,S**)-22e to bind biological target proteins unrelated to PDE family enzymes was
evaluated by interrogating a panel of 400 key receptor, enzyme, and ion channel proteins at 1µM
concentration (Eurofins selectivity panels). The panel identified only three hits showing > 50% binding
at the concentration of 1 M for (S*,S**)-18e (rat Acyl CoA-Cholesterol Acyltransferase (50%), human
CYP3A4 (73%) and human monoamine oxidase A (76%)), and five hits showing > 50% binding at the
concentration of 1 M for (S*,S**)-22e (human CYP2C19 (69%), human CYP3A4 (84%), human
monoamine oxidase A (94%), human UDP glucuronyltransferase 1A1 e rat imidazoline I2 receptor
(60%)).
CONCLUSIONS
Following a structure-based design we identified a novel class of PDE4 inhibitors endowed with
potent anti-inflammatory properties and suitable for topical pulmonary administration. Starting from a
previously described series of benzoic esters, we explored the chemical space in the solvent-exposed
region of the PDE4 catalytic pocket, focusing our efforts on new structural classes of ortho-substituted
compounds. Our SAR efforts led to the discovery of pyrrolidinyl- and thiazolidinyl- esters as potent in
vitro PDE4 inhibitors. The optimal in vitro ADME profiles (high plasma protein binding, high CLint
,
low Caco2 permeability level, high lung S9 stability) together with their ability to better perform in
solid state studies (highly crystalline, anhydrous, not hygroscopic) led to the selection of compounds
(S*,S**)-18e and (S*,S**)-22e for progression to in vivo preclinical studies. Both compounds
effectively counteracted OVA-induced airway inflammation in rat upon IT administration as dry
ACS Paragon Plus Environment
2
8