Rational Drug Design of Topically Administered Caspase 1 Inhibitors for the Treatment of Inflammatory Acne

Article abstract of DOI:10.1021/acs.jmedchem.8b00067

The use of an interleukin β antibody is currently being investigated in the clinic for the treatment of acne, a dermatological disorder affecting 650M persons globally. Inhibiting the protease responsible for the cleavage of inactive pro-IL1β into active IL-1β, caspase-1, could be an alternative small molecule approach. This report describes the discovery of uracil 20, a potent (38 nM in THP1 cells assay) caspase-1 inhibitor for the topical treatment of inflammatory acne. The uracil series was designed according to a published caspase-1 pharmacophore model involving a reactive warhead in P1 for covalent reversible inhibition and an aryl moiety in P4 for selectivity against the apoptotic caspases. Reversibility was assessed in an enzymatic dilution assay or by using different substrate concentrations. In addition to classical structure-activity-relationship exploration, topical administration challenges such as phototoxicity, organic and aqueous solubility, chemical stability in solution, and skin metabolic stability are discussed and successfully resolved.

Full text of DOI:10.1021/acs.jmedchem.8b00067

Subscriber access provided by UNIV OF ALABAMA BIRMINGHAM
Rational Drug Design of Topically Administered Caspase
Inhibitors for the Treatment of Inflammatory Acne
1
Jean-Francois Fournier, Laurence Clary, Sandrine Chambon, Laurence Dumais, Craig Steven
Harris, Corinne Millois-Barbuis, Romain Pierre, Sandrine Talano, Etienne Thoreau, Jérome
Aubert, Michèle Aurelly, Claire Bouix-Peter, Anne Brethon, Laurent Chantalat, Olivier Christin,
Catherine Comino, Ghizlane El-Bazbouz, Anne-Laurence Ghilini, Tatiana Isabet, Claude
Lardy, Anne-Pascale Luzy, Céline Mathieu, Kenny Mebrouk, Danielle Orfila, Jonathan Pascau,
Kevin Reverse, Didier Roche, Vincent RODESCHINI, and Laurent François Hennequin
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b00067 • Publication Date (Web): 12 Apr 2018
Downloaded from http://pubs.acs.org on April 12, 2018
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 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 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.
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 84
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Rational Drug Design of Topically Administered
Caspase 1 Inhibitors for the Treatment of
Inflammatory Acne
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
,†
†
†
†
Jean-François Fournier,* Laurence Clary, Sandrine Chambon, Laurence Dumais, Craig
†
†
†
†
†
Steven Harris, Corinne Millois, Romain Pierre, Sandrine Talano, Étienne Thoreau, Jérome
†
†
†
‡
†
Aubert, Michèle Aurelly, Claire Bouix-Peter, Anne Brethon, Laurent Chantalat, Olivier
†
†
†
†
§
Christin, Catherine Comino, Ghizlane El-Bazbouz, Anne-Laurence Ghilini, Tatiana Isabet,
‡
†
†
†
†
Claude Lardy, Anne-Pascale Luzy, Céline Mathieu, Kenny Mebrouk, Danielle Orfila,
†
†
‡
‡
Jonathan Pascau, Kevin Reverse, Didier Roche, Vincent Rodeschini, Laurent François
†
Hennequin
Dedicated to Professor E. J. Corey on the occasion of his 90th birthday
† Nestlé Skin Health R&D, 2400 Route des colles, BP 87, 06902 SophiaꢀAntipolis Cedex,
France
‡
§
Edelris, 115 Avenue Lacassagne, 69003 Lyon, France
Synchrotron Soleil, L'Orme des Merisiers, SaintꢀAubin ꢀ BP 48, 91192 GifꢀsurꢀYvette Cedex,
France
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 84
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
ABSTRACT: The use of an interleukin beta antibody is currently being investigated in the clinic
for the treatment of acne, a dermatological disorder affecting 650M persons globally. Inhibiting
the protease responsible for the cleavage of inactive proꢀIL1β into active ILꢀ1β, caspaseꢀ1, could
be an alternative small molecule approach. This report describes the discovery of uracil 20, a
potent (38 nM in THP1 cells assay) caspaseꢀ1 inhibitor for the topical treatment of inflammatory
acne. The uracil series was designed according to a published caspaseꢀ1 pharmacophore model
involving a reactive warhead in P1 for covalent reversible inhibition and an aryl moiety in P4 for
selectivity against the apoptotic caspases. Reversibility was assessed in an enzymatic dilution
assay or by using different substrate concentrations. In addition to classical structureꢀactivityꢀ
relationship exploration, topical administration challenges such as phototoxicity, organic and
aqueous solubility, chemical stability in solution and skin metabolic stability are discussed and
successfully resolved.
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
Acne is a dermatological disorder affecting 40ꢀ50 million Americans and 650 million
1
globally. Lesions typically appear in the face, upper back and chest area and can lead to
permanent scarring as well as low selfꢀesteem. Current treatment consists of the use teratogenic
retinoids such as Roaccutane™. Therefore, drugs with novel mechanisms of action and improved
2
safety profiles would benefit patients.
The pathogenesis of acne can be divided in different steps: 1) (over)production of sebum by
the sebaceous gland; 2) colonization of the hair follicle by P. Acnes; 3) hyperkeratinisation of the
3
follicle; and 4) the release of inflammatory mediators into the skin. Proꢀinflammatory cytokines,
including TNFα, ILꢀ1α and ILꢀ1β are considered to be involved for follicular hyperkeratinisation
and the inflammatory lesions characteristic of acne. This cytokine usually exists in its proꢀform
ACS Paragon Plus Environment
2

Page 3 of 84
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
which is cleaved to its active form by the aspartic cysteine protease caspaseꢀ1 upon inflammatory
4
stimulus. Recently, French and coꢀworkers have demonstrated that the active form of ILꢀ1β is
5
abundant in human acne lesions. Moreover, they demonstrated that the production induced by P.
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
Acnes of active ILꢀ1β in monocytic lineage and in mice is dependant of the NLRP3/Caspaseꢀ1
inflammasome by P. Acnes exposed cells. This data indicates that ILꢀ1β as a novel potential
therapeutic target in acne, and therefore, we postulate that inhibiting caspaseꢀ1, could have a
beneficial effect in inflammatory acne.
Most dermatological diseases such as acne are nonꢀlife threatening, and as such, treatments are
subject to a very low tolerance to risk. Thus, a common strategy employed is ensuring the
concentration of drug in the target organ, skin, is significantly higher than that in blood. Because
6
the free drug hypothesis precludes accomplishing this via the systemic route, a local topical
7
administration of drugs showing high systemic clearance is often favored (soft drug ).
Furthermore, with dermatological diseases which affect only a small portion of the body surface
area (BSA), this can also lead to dose minimization and therefore to a reduced risk. This could be
8
of particular interest in the case of targets which are typically treated with covalent drugs.
Caspaseꢀ1 is a wellꢀdocumented target for which a few clinical candidates have already been
9
evaluated, including Vertex’ belnacasan and pralnacasan. However, previously disclosed
inhibitors have been optimized for the systemic rather than the topical route and are not
10
necessarily appropriate for use in dermatology. We have therefore initiated a chemistry
program to identify a caspaseꢀ1 inhibitor better suited to the topical treatment of inflammatory
acne.
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 84
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
RESULTS AND DISCUSSION
Hit Generation and Reversibility Issues
Most published structures of caspaseꢀ1 inhibitors follow a few general rules derived from the
elucidation of the protein’s tertiary structure by crystallography, and from the key binding
1
1
elements of the tetrapeptide AcꢀYVADꢀCHO, an early tool compound. As shown in Figure 1a,
a reactive warhead and (L)ꢀaspartic acid mimic in P1 is key for potency and selectivity vs other
cysteine proteases, key Hꢀbond donors and acceptors in P2 and P3 and an aromatic residue in P4
11
for selectivity vs the apoptotic caspases. Consequently, about a dozen different new scaffolds
were designed in accordance to these guidelines and were evaluated in an enzymatic assay using
two different concentrations (K and 20K ) of substrate, a fluorescent proꢀILꢀ1β analog (Figure
M
M
1
b). The aim was for these peptidomimetic scaffolds to present a chemical topology that would
allow the formation of the same Hꢀbonds with caspaseꢀ1 in P2 and P3 as depicted in Figure 1a.
Namely, the two NHs flanking R and R and, between these, one acceptor (carbonyl or
1
4
pyridine). In addition, the outgoing vectors for R and R should be consistent with those from
1
4
ZVADꢀCHO to provide the other binding elements: the covalent bond and salt bridge in P1 and
the πꢀstacking in P4. To standardize this phase, each synthesized scaffold was combined at the
very least with the ubiquitous aspartyl aldehyde warhead in P1 and a 2ꢀnaphthyl (similar to
pralnacasan’s 1ꢀisoquinolyl) group in P4.
a)
ACS Paragon Plus Environment
4

Page 5 of 84
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
b)
O
O
R1 =
O
R4 =
A= NH, CH2, CH2CH2
OH
O
R4
NH
A
R4 NH
O
N
O
O
R1
O
N
H
R1
H
N
N
N
R1
H
N
R4
N
OH
H
O
ABO
P(y)DO
Pyridine
pIC50 Km 7.8
pIC50 Km <5-6.0
pIC50 Km 5.9
pIC50 20Km 7.0
pIC50 20Km <5-5.9
pIC50 20Km <5
H
O
O
N
N
O
R4
O
O
R1
H
N
O
N
N
R1
R4
N
R1
H
R4
N
N
H
N
H
N
H
O
O
A
Uracil
Macrocycle
pIC50 Km 5.3
pIC50 20Km <5
CADC
pIC50 Km 6.3
pIC50 Km <5-6.6
pIC50 20Km 5.1
pIC50 20Km <5-5.8
Figure 1. (a) ZVADꢀCHO bound to caspaseꢀ1 shows the key binding elements that need to be
reproduced in novel scaffolds: covalent bond with CYS285, ionic interactions with ARG179,
GLN283 and ARG341, Hꢀbonds with SER339’s carbonyl and ARG341’s carbonyl and NH ; (b)
Some examples of novel scaffolds designed to reproduce key binding elements with caspaseꢀ1
and tested for inhibitory activity
In principle, the IC s of competitive inhibitors following MichaelisꢀMenten kinetics should be
5
0
1
2
right shifted about 10ꢀfold at the higher substrate concentration (K = IC /(S/K + 1)). By
i
50
M
testing at two different substrate concentrations, we were able to get a first crude idea of the
binding kinetics of our inhibitors. Indeed, since our inhibitors contained a reactive warhead
which after an initial preꢀcomplexation step makes a covalentꢀreversible bond with the protein’s
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 84
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
13
catalytic cysteine, we anticipated that some of our inhibitors would not reach equilibrium under
our assay conditions (Figure 2).
O
O
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
O
O
H
N
O
O
O
N
H
O
O
N
N
N
H
N
N
O
OH
H
O
OH
1
2
O
O
O
O
O
O
O
O
H
N
O
N
N
N
N
N
H
O
OH
H
H
OH
O
3
4
O
O
NH
O
O
N
N
O
O
N
O
NH
O
NH
OH
N
H
O
O
O
OH
5
(PGE-3935199)
6
O
O
NH
O
O
O
O
H
H
N
N
O
N
F
O
O
N
H
N
O
O
OH
N
N
H
H
O
OH
O
7 (Z-YVAD-FMK)
8
OH
(a)
(b)
ACS Paragon Plus Environment
6

Page 7 of 84
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 2. (a) Correlation between the pIC measured at K vs 20K substrate concentration for
50
M
M
early compounds. For competitive inhibition following MichaelisꢀMenten kinetics, a one log unit
difference in pIC values would be expected (green line). (b) Reversibility profiles as evaluated
5
0
by the recovery of enzymatic reaction rate in a kinetic dilution experiment. Comparing results
from both graphs suggest that reversibility of the inhibition – under the condition of the assay –
can be quickly evaluated using the ratio of IC50 values: fully reversible inhibitors (green
line/bars) have a 10ꢀfold difference whereas pseudoꢀirreversible inhibitors (red line/bars) have
equal IC s irrespective of the substrate concentration. Under this assumption, uracil compound 8
5
0
is fully reversible.
For a few selected compounds displaying various profiles, a kinetic experiment was done to
1
4
assess the reversibility of the inhibitors (Figure 2b). In this assay, the enzyme is preꢀincubated
with inhibitors at 10x their IC50 values before being diluted a 100 fold with the substrate.
Recovery of the enzymatic activity is then monitored and the extent of reaction speed recovery
(from the slope of the curve) indicates whether the inhibitor is displaying a fully reversible
profile, a pseudoꢀirreversible (irreversible in the timescale of the assay) or partiallyꢀreversible
1
2
profile. Since the results fully corroborated the ratio of the IC s with substrate concentrations
5
0
ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 84
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
at 20K vs K , this higher throughput method was thereafter used to monitor the reversibility of
M
M
1
5
the covalent bond formation with the catalytic cysteine. Consequently, this methodology gave
1
6
us confidence that we could properly rank our compounds and conduct a normal SAR.
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
From Hit to Lead
This article focuses on the work around the uracil hit 8, but other scaffolds were also
1
7
investigated in parallel. Analogs with varying substituents in the P2 to P4 subpockets were
prepared in a few SAR iteration cycles with most of the focus on R4, identifying lead compound
1
0 (pIC50 Km 8.3; pIC50 20Km 7.5) bearing Sunesis' quinoxaline in P4 and which was isolated as a
1
8
19
yellow solid. Conducting a retrospective binding efficiency index (BEI) analysis as proposed
by Hajduck showed that although the BEI slowly crept up during the hit to lead (H2L) phase for
the new “concept compounds”, a linear relationship between Mw and pIC could be found with
5
0
2
0
a slope of 48 amu/pIC unit (Figure 3). These findings are in line with Hajduk’s retroanalysis
5
0
across 15 Abbott projects, and points towards the same conclusion: when adding new atoms on a
hit or lead, if only the optimal ones are retained, then the molecular weight and potency of the
starting point (hit) should be projectable to those of the end point (candidate). This analysis thus
strengthened our belief the uracil series could be optimized to deliver a candidate.
ACS Paragon Plus Environment
8

Page 9 of 84
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 3. Binding efficiency index plot of the H2L phase of the uracil series. Starting from hit 8,
a 10ꢀfold improvement in potency can be achieved by the addition of 48 amu. Potency gains that
come at the cost of larger molecular weight contributions should be rejected.
However, the lead uracil compounds were subject to a number of liabilities specific to topical
drug development, namely color of the drug substance, solvent solubility, phototoxicity and
chemical stability in solution. These were, therefore, the parameters chosen as the main focus of
the lead optimization phase.
Lead Optimisation
Solvent Solubility. In the case of orally administered drugs, good aqueous solubility (>20ꢀ
4
00 µM for compounds with medium permeabilities) is necessary to achieve good absorption in
2
1
the gut. As per Yalkowsky’s General Solubility Equation (GSE), this can be achieved either by
decreasing the melting point (approximation of the crystalline lattice energy) or the LogD of
2
2
compounds. However, in the case of topically administered drugs, it is often necessary to
2
3
achieve orders of higher magnitude solubilities (>20 mM) for formulation purposes. This is
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 84
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
usually achieved by the use of nonꢀaqueous solvents such as ethanol, which means that the main
driver for solubility will be its crystallinity/Mp (the LogD component in the GSE serves to
“account for the difference between an ideal solution and an aqueous solution”; it is, therefore,
2
2,24
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
specific to aqueous solubility).
Considerable efforts were therefore invested in getting
meaningful solubility measurements: semiꢀstandardized crystallization processes were
elaborated, crystallinity and melting points was assessed by DSCs with large enthalpies (>50 J/g)
and thermodynamic solubilities were measured in a routine assay directly from crystalline
material after 24 h in a 30/70 mixture of PG/EtOH. In the case of the uracil series, compounds
2
5
with high solubility had lower melting points and higher (not lower!) ChromLogD (Figure 4).
Figure 4. Solubility in PG/EtOH is a function of the melting point in the uracil series with no
positive impact from lowering ChromLogD.
3
We surmised we might disrupt the crystalline lattice of our uracils by increasing their sp
2
6
fraction, and more specifically, by increasing the substitution pattern of the carbon atoms
ACS Paragon Plus Environment
10

Page 11 of 84
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
directly connected to the uracil core. This strategy proved fruitful as the isopropyl substituted
uracil 11 showed a 10ꢀfold improvement on PG/EtOH solubility while maintaining aqueous
2
7
solubility vs 10 (Table 1). Inversely, removing the methyl substituent at R resulted in a
2
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
compound with decreased solubilities (12 vs 11) despite the small drop in lipophilicity. Finally,
switching the quinoxaline ring for a naphthyl ring also led to significant improvements in both
types of solubility (12 vs 13). The fact that the aqueous solubility increased 6ꢀfold, in spite of the
ꢁChromLogD, confirms the importance of this structural elements on crystallinity and
28
solubility. Moreover, this switch also resulted in colorless compounds as opposed to the yellow
2
9
quinoxaline analogs. Unfortunately, the LipE took a double hit as the potency decreased at the
same time the ChromLogD increased. This behavior was generally observed across many
matched pairs.
Table 1. Optimisation of PG/EtOH 30/70 solubility. Replacing the quinoxaline for a naphthalene
group (13) and steric perturbation of the Hꢀbond donors/acceptors (14) give rise to solubility
improvements.
SolPG/ꢀ
EtOH
IC50
KM
(nM)
IC50
THP1
(nM)
Chrom
LogD
SolAq
(µM)
LipE
KM
LipE
THP1
Cpd
A
R1
Me
iPr
iPr
iPr
iPr
R2
(mM)
1
1
1
1
1
0
1
2
N,N
N,N
N,N
H,Me
H,Me
H,H
1.9
2.4
2.3
3.2
2.4
0.89
11
96
86
8.3
52
87
5.6
5.6
3
48
50
6.4
5.9
6.2
4.5
6.0
5.5
4.9
5
5.2
36
53
3 CH,CH
N,N
H,H
21
4.7
140
27
3.7
5.2
4
ꢀCH
2
CH
2
ꢀ
25
ACS Paragon Plus Environment
11

Journal of Medicinal Chemistry
Page 12 of 84
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
Next we thought the addition of a second substituent alpha to the aliphatic amide might further
improve solubility as well as stereochemical stability. Furthermore, inspection of the XꢀRay
structure of 10 in Caspaseꢀ1 revealed that the pocket formed by Val338 and Trp340 might
accommodate a small substituent and we hoped we could improve potency at the same time
Indeed, the overlay of the docking of cyclopropyl 14 over the Xꢀray structure of 10 shows that
the cyclopropyl group fits nicely in the shallow S2 lipophilic pocket with no perturbation of the
surrounding side chain, while the 3 essential backbone Hꢀbonds of the ligand are also maintained
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). Cyclopropane 14 was therefore synthesized and we were rewarded with small but
reproducible improvements in solubility, cellular potency and LipE vs 11. Figure 6 shows the
potency boost was specific to the cyclopropyl moiety and reproducible across many matched
pairs.
Figure 5. Docking studies of 14 (brown) vs XꢀRay structure of 10 (green; PDB ID 6F6R)
suggest the cyclopropyl moiety in S2 is optimally fits a small lipophilic pocket.
ACS Paragon Plus Environment
12

Page 13 of 84
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 6. Molecular matched pairs on the R1 substituents (connected dots) vs celluar potency:
cyclopropyl unit gives a consistent increase in potency vs nonꢀsubstituted, monoꢀmethyl, monoꢀ
ethyl and gemꢀdimethyl substituents.
Phototoxicity. Following its topical application, a drug substance is typically found in much
higher concentration in skin than what occurs after oral administration. Furthermore, following
UV irradiation by the sun, some benign compounds may become irritant or even toxic. This
phenomenon is called phototoxicity. Because skin irritation is often times incompatible with the
desired therapeutic effect, phototoxicity can be a liability and needs to be screened early on in
the development. Photocytotoxicity potential is assessed by evaluating cell viability after 24h
exposure with caspaseꢀ1 inhibitors in the presence and absence of UV light. A phototoxicity
irritation factor (PIF) value is determined by the ratio of LC50 values and compounds are
classified as follows: PIF<2 is non phototoxic, 2<PIF<5 has low phototoxic potential and PIF>5
3
0
is phototoxic. A good practice when evaluating phototoxicity SPR is to keep an eye on the
specific LC values for more subtle effects (provided in supplementary material, Table 1S). In
5
0
ACS Paragon Plus Environment
13

Journal of Medicinal Chemistry
Page 14 of 84
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 case of the present discussion, we find that simply taking into account the overall conclusion
suffices.
When we initially switched our focus from the quinoxaline to naphthalene substituents in P4
pocket, we observed a higher proportion of phototoxic products. For example, 13 was phototoxic
whereas 12 was not. A more detailed analysis revealed a clear dependency on lipophilicity
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) which gave an upper limit of ChromLogD of about 3.5 to 4 for the design of new
3
1
inhibitors. Because of their better suited physical properties and despite their inherent higher
lipophilicity, the naphthalene subseries was pursued since a design window remained and other
strategies to mitigate the phototoxicity risk, for instance breaking conjugation between πꢀ
systems, remained (see below).
Figure 7. Phototoxicity in the uracil series is highly dependent on ChromLogD: 100% of the
compounds with ChromLogD lower than 2 were photosafe vs only 18% when it was higher than
4
.
ACS Paragon Plus Environment
14

Page 15 of 84
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
For example, structural elements such as the addition of an orthoꢀmethyl on the phenyl ring of
phototoxic 13 resulted in an inhibitor with only a low risk of phototoxicity (15) along with an
improvement in solubility and potency (Table 2). The likely explanation is that the orthoꢀ
substituent twists the acyl moiety out of plane and breaks the conjugation with the phenyl ring,
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
32
an important factor in light absorption.
Table 2. Final Optimisation: cyclopropane 20 combines all the beneficial chemical features
which translate into a compound with the required profile in terms of solubility, potency, safety
and stability.
Chrom Sol
LogD
IC50 c
SolAq THP1
Photoꢀ
toxicity
3ꢀMo Skin S9
PG/ꢀ
a
b
d
e
Cpd
A
R1
H
R2
n
0
Stab
CLint
EtOH
1
3
CH
H,H
3.2
36
52
140 Phototoxic
Low
82
<0.82
1
5
CH Me
H,H
0
3.5
51
72
66
phototox
potential
85
72
(3.9)
16
17
18
CH Me
H,Et
H,Et
H,Et
0
0
1
4
n/a
49
40
2
76
n/a
(4.9)
2 (7.7)
2.9 (13)
N
Me
4.5
3.8
200 Phototoxic
100
84
CH Me
n/a
42
210
n/a
Low
phototox
potential
<0.82
(
1
9
0
CH
H
ꢀCH
2
CH
2
ꢀ
0
3.5
31
66
23
99
2.1)
Non
phototox
0.92
(4.6)
2
CH Me ꢀCH
2
2
CH
CH
2
ꢀ
ꢀ
0
1
3.7
3.4
43
70
63
38
100
88
21 CH Me ꢀCH
n/a
360
3.5
2
Non
ACS Paragon Plus Environment
15

Journal of Medicinal Chemistry
Page 16 of 84
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
phototox
(84d)/
a
b
c
d
Solubility in mM Solubility in µM Potency in nM In percent of recovered parent
compound. S9 CLint in µL/min/mg of protein. For numbers in parentheses, CLint values
e
6
measured in NHEK cells in µL/min/10 cells.
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
3
3
Chemical and Skin Metabolic Stability. The chemical stability of drug substances is
paramount in topical delivery medicinal chemistry projects as APIs are not formulated in solid
forms, but rather dissolved in various excipients and need to have a shelf life stability of at least
a year. From the start of the project, this parameter was judged to be at high risk due to the
electrophilic nature of the aspartyl aldehyde warhead and was, therefore, constantly monitored.
To this effect, stability of compounds with the aspartyl aldehyde warhead was evaluated in
34
9
6% ethanol at 23 °C for 12 weeks, but SPR information could be used after 4 weeks by
3
5,36
projecting stability results to 12 weeks using the appropriate (best fit) kinetic order equations.
It was found that a significant proportion of our inhibitors were not stable at 23 °C for more than
3
months. Consequently, significant efforts went into the synthesis of novel and hopefully more
3
7
stable electrophilic warheads.
These novel warheads should feature an electrophilic moiety susceptible to thiolate attack by
caspaseꢀ1’s CYS285 as well as an acidic residue for the ionic interactions with caspaseꢀ1’s
ARG179, GLN283 and ARG341. Herein lays the challenge: to be reactive, yet stable in solution
despite the combination of an electrophile with an acid. Different acids, electrophiles, linker
length and masking groups were evaluated. The results, however, were quite disappointing as
FreeꢀWilson analysis of the contributions of many warheads on chemical stability and potency
showed that most were much weaker binders in the enzymatic assay (2 log units loss) and/or did
3
8
not offer improvements in stability (Figure 8). Most of the remaining warheads were shifted in
ACS Paragon Plus Environment
16

Page 17 of 84
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
39
the cellular assay to even lower potencies, leaving the homoꢀaspartyl aldehyde warhead as the
4
0
sole candidate for further exploration.
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
a)
b)
ACS Paragon Plus Environment
17

Journal of Medicinal Chemistry
Page 18 of 84
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
a
b
a
Figure 8. (a) Stability vs potency of various electrophilic warheads. Numbers in boxes
represent the FreeꢀWilson contribution on chemical stability in ethanol at 23°C after three
months as compared to the standard aspartyl aldehyde warhead A. For example, molecular
matched pairs with warhead J give, on average, analogs with 33% less parent product after three
b
months. Grey boxes: no stability measurements were made.
The yꢀcoordinate of boxes
represent the binned FreeꢀWilson contribution on enzymatic pIC50 (at KM concentration of
substrate) as compared to the standard aspartyl aldehyde warhead A. For example, molecular
matched pairs with warheads F through H give, on average, analogs that are 10 to 30 fold less
potent compounds (ꢀ1 to ꢀ1.5 log unit). (b) Structure of warheads used in FreeꢀWilson plot.
Aspartylaldehyde warhead A is the optimal warhead when taking into account both potency and
chemical stability.
ACS Paragon Plus Environment
18

Page 19 of 84
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Indeed, homoꢀaspartyl aldehyde 18 showed a modest increase in 3ꢀmonth stability vs the
related aspartyl aldehyde 16 (84% vs 72%) with minimal drop in potency (IC THP1 210 nM vs
5
0
76 nM), however skin metabolism was also almost threeꢀfold higher (13 vs 4.9 µL/min/mg
protein in normal human epidermal keratinocytes). Furthermore, whereas the increase in the skin
S9 fraction or NHEK clearance was observed across many molecular match pairs, the increase in
chemical stability was not (Table 2).
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
Surprisingly, we observed that pyridine 17 was much more chemically stable than the
analogous phenyl 16 (100% vs 72%). The role of this pyridine in the compound’s stability was
unclear as the only degradation products observed for this chemical series involved the warhead.
4
1
However, the pyridine analogs also displayed increased skin clearance and lipophilicity along
with lower potency. This approach was therefore not pursued.
Fortunately, cyclopropane 19 displayed a good mix of solubility, potency, chemical and skin
4
2
metabolic stability, albeit with a low phototoxicity potential. As described previously, addition
of the orthoꢀmethyl to break πꢀconjugation gave 20, a nonꢀphototoxic compound that retained all
the qualities of 19. This compound was therefore selected for further evaluation.
Synthesis
The general strategy towards our novel P2/P3 peptidomimetic uracil fragment relied upon
selective Nꢀ3ꢀprotection of 5ꢀnitrouracil with the benzhydryl group followed by Nꢀ1 alkylation to
4
3
prepare 23 in acceptable yield. Subsequent deprotection of the benzhydryl group under
4
3b
hydrolytic conditions gave access to 24 for exploration from the N-3 vector (Scheme 1).
Scheme 1. General synthesis strategy employed for our uracilꢀbased P2/P3 peptidomimetic
fragments building blocks exemplified by the synthesis of 24.
ACS Paragon Plus Environment
19

Journal of Medicinal Chemistry
Page 20 of 84
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
Reagents and conditions: (a) Ph CHBr, N,OꢀBis(trimethylsilyl)acetamide, cat. I , acetonitrile,
2
2
reflux, yield 85%; (b) K CO , acetonitrile, 23 °C, yield 64%; (c) TFAꢀTfOH (1:1), 23 °C, yield
2
3
9
8%.
For analogs bearing a quaternary C centres at Nꢀ1, we were obliged to develop a novel 3MCR
multiꢀcomponent reaction) to prepare the gem dimethyl and cyclopropyl analogues (Scheme
(
4
4
2
). Briefly, aminocarbonylation of the amino acid residues followed by protection of the
carboxylic acid residue afforded primary urea precursors 26. Condensation of 26 with ethyl
nitroacetate in the presence of triethylorthoformate afforded a 6:4 E/Z ratio of eneꢀurea
intermediates 27. Finally cyclisation to the urea scaffolds was carried out in the presence of a
twoꢀfold excess of cesium carbonate to afford 28 (Scheme 2).
Scheme 2. Preparation of P2/P3 uracil peptidomimetic fragments building blocks with
quaternary C centres at Nꢀ1.
ACS Paragon Plus Environment
20

Page 21 of 84
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
Reagents and conditions: (a) KOCN, water, reflux, yield 78%; (b) Allyl bromide, DIPEA,
THF/H O (10:1), 60 °C, 18 h, yield 80%; (c) ethyl nitroacetate, triethyl orthoformate, toluene,
2
reflux, yield 92%; (d) Cs CO , MeCN, reflux, yield 81%.
2
3
Intermediates 24 and 28 were then converted to caspaseꢀ1 inhibitors following one of the three
strategies enumerated below. The most common approach was to begin by alkylation at Nꢀ3 to
afford intermediate 41 (Scheme 3b). Then, installation of the P4 fragment was carried out by
4
5
reduction of the nitro group under dissolving metal conditions and subsequent amide bond
formation using COMU (1ꢀcyanoꢀ2ꢀethoxyꢀ2ꢀoxoethylidenaminooxy)dimethylaminoꢀ
morpholinoꢀcarbenium hexafluorophosphate) as the coupling agent and acid 40 (Scheme 3a) as
4
6
47
the coupling partner to give 43. After saponification of ester 43 to acid 44, final compound
48
assembly was terminated by coupling the warheads 33 or 38 (Scheme 3a) after in situ
deprotection of the Alloc group with the preꢀactivated Pfp (pentafluorophenol) ester of 44
followed by hydrolysis of the acetals 45 or 46 to the hemiꢀacetals 20 or 21, respectively.
ACS Paragon Plus Environment
21

Journal of Medicinal Chemistry
Page 22 of 84
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
Scheme 3. General synthesis strategy employed to our uracilꢀbased caspaseꢀ1 inhibitors (a) P1
and P4 building blocks (b) and final assembly
a)
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
O
CO tBu
CO tBu
CO tBu
2
2
2
a
d
e
O
O
O
O
AllocHN
AllocHN
AllocHN
H N
2
CO H
2
X
O
29: X = OH
30: X = NCH OCH
31: X = H
b
c
32
33
3
3
CO tBu
2
CO tBu
CO tBu
CO tBu
2
2
2
f
i
j
R
N
H
AllocHN
AllocHN
FmocN
O
CO H
O
N
2
O
H
O
O
37
38
3
3
3
4: R = Fmoc
5: R = H
6: R = Alloc
g
h
O
O
O
O
O
k
O
l
OH
R'
OH
N
H
N
Br
H
39
40
b)
ACS Paragon Plus Environment
22

Page 23 of 84
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
Reagents and conditions: (a) NaHCO , Allyl chloroformate, THF/H O (1:3), 23 °C, yield 80%;
3
2
(
b) MeONHMe·HCl, NMM, EDCI, CH Cl , 23 °C, yield 86%; (c) LiAlH , THF, ꢀ78 °C, yield
2 2 4
83%; (d) (EtO) CH, 3Å molecular sieves, pꢀTsOH, EtOH, 23 °C, yield 92%; (e) TFA, CH Cl ,
3
2
2
2
3 °C, yield 71% (f) MeONHMe·HCl, NMM, EDCI, CH Cl , 23 °C, yield quantitative (g) DBU,
2 2
CH Cl , 23 °C, yield 68%; (h) NaHCO , Allyl chloroformate, THF/H O (1:3), 23 °C, yield 88%;
2
2
3
2
(
i) LiAlH , THF, ꢀ78 °C, yield 95%; (j) (EtO) CH, 3 Å molecular sieves, pꢀTsOH, EtOH, 23 °C,
4
3
yield 81%; (k) 2ꢀaminonaphthalene, Pd(OAc) , Xantphos, Cs CO , CH Cl , µwave 120 °C, yield
2
2
3
2
2
8
9%; (l) LiOH(aq), MeOH, 50 °C, yield 92% (m) 2ꢀiodopropane, K CO , DMF, 23 °C, yield
2 3
4
6%; (n) Fe(s), NH Cl, THF/EtOH/H O (1:1:1), 70 °C, yield 98%; (o) 40, COMU, NMM, DMF,
4
2
23 °C, yield 90%; (p) Pd(PPh ) , 1,3ꢀdimethylbarbituric acid, CH Cl , 23 °C, yield 60%; (q) i.
3
4
2
2
ACS Paragon Plus Environment
23

Journal of Medicinal Chemistry
Page 24 of 84
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
HATU, CH Cl /DMF (1:1), 23 °C; ii. 33, Pd(PPh ) , 1,3ꢀdimethylbarbituric acid, CH Cl /DMF
2
2
3 4
2
2
(1.4:1), 23 °C, yield quantitative; (r) HCl(aq), acetonitrile, 23 °C, yield 65%; (s) i. HATU,
CH Cl /DMF (0.7:1), 23 °C; ii. 38, Pd(PPh ) , 1,3ꢀdimethylbarbituric acid, CH Cl /DMF (1.4:1),
2
2
3 4
2
2
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
2
3 °C, yield 95%; (t) TFA, CH Cl , 0ꢀ23 °C, yield 80%.
2 2
The second strategy was developed to permit Nꢀ3 alkylation at the end of the sequence in order
to probe SAR at this position. Reduction of 47 – obtained using the methodology developed for
28 – by hydrogenation afforded amine 48 in poor yield (Scheme 4). Coupling of the P4 fragment
using COMU followed by deprotection of the tertꢀbutyl ester afforded 50. The preꢀcyclised
hemiacetal warhead 33 was next deprotected and coupled in situ using T3P as the coupling agent
4
9
to afford 51. Finally, exploration at Nꢀ3 was best achieved through Mitsunobu alkylation
chemistry and hydrolysis of the acetal proꢀdrug to the hemiacetal drug molecules was achieved
using aqueous HCl to afford 20.
Scheme 4. Strategy employed for Nꢀ3 alkylation exploration based on a lateꢀstage Mitsunobu
alkylation.
H
N
H
N
H
N
O
O
O
O
O
O
O
N
a
N
b
N
c
O N
2
O
H N
2
O
R
N
O
H
O
O
O
4
7
48
49
H
H
O
O
O
N
O
N
O
N
O
O
O
O
O
O
O
O
e
d
N
N
N
R
N
OH
R
N
N
R
N
N
H
H
H
O
H
H
O
O
OR'
O
50
51
45: R' = Et
20: R' = H
f
*
=
R
N
H
ACS Paragon Plus Environment
24

Page 25 of 84
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Reagents and conditions: (a) H , PdꢀC, EtOAc, 23 °C, yield 32%; (b) 40, COMU, NMM, DMF,
2
8
0 °C, yield 79%; (c) 4M HCl in dioxane, 23 °C, yield 78%; (d) i. DIPEA, T3P, DMF, 23 °C; ii.
33, Pd(PPh ) , 1,3ꢀdimethylbarbituric acid, CH Cl /DMF (1:1), 23 °C, yield 73%; (e) iꢀPrOH,
3
4
2
2
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
DIAD, PPh , CH Cl , 0 °C to 23 °C, yield 11%; (f) 1M HCl (aq), MeCN, 23 °C, 65%.
3
2
2
Finally, we also developed an efficient library route to probe the P4 region (Scheme 5).
Alkylation of 24 followed by saponification of ester 52 afforded the acid 53. Unfortunately, 53
underwent partial epimerization during the saponification step but we decided to continue with
this route to build up our SAR knowledge of this key selectivity pocket. Coupling of 53 with the
preꢀcyclised warhead 33 via the Pfp ester methodology followed by reduction afforded the key
library synthon 55. Final amide coupling through Schotten-Baumann conditions or using COMU
followed by hydrolysis of the warhead acetal 56 afforded hemiacetal final compounds 57 in
moderate to good yields.
Scheme 5. Strategy employed for the P4 region exploration based on a lateꢀstage amide coupling
reaction.
ACS Paragon Plus Environment
25

Journal of Medicinal Chemistry
Page 26 of 84
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
H
N
O
N
O
O
N
O
O
O
O
O
N
N
N
a
b
c
O N
O
O N
O
O N
OH
2
2
2
O
24
52
53
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
O
O
O
N
O
O
N
O
O
O
O
O
O
N
O
O
O
O
N
N
d
e
O N
N
H
H N
N
N
2
2
R
N
N
H
O
O
H
H
O
OR'
54
55
5
5
6: R' = Et
7: R' = H
f
Compounds 14 and 19 employed this strategy.
Reagents and conditions: (a) MeꢀI, K CO , DMF, 23 °C, yield 88%; (b) LiOH, THF/MeOH
2
3
(
10:1), 50 °C, yield 92% (S/R, 86:14); (c) i. PfpꢀOH, EDCI, CH Cl /DMF (4:1), 23 °C; ii. 33,
2 2
Pd(PPh ) , 1,3ꢀdimethylbarbituric acid, CH Cl /DMF (2:1), 23 °C, yield 60%; (d) H , PdꢀC,
3
4
2
2
2
EtOAc, 23 °C, yield 80%; (e) RCOCl, DIEA, THF or RCOOH, HATU, NMM, THF, 23 °C or
RCOOH, COMU, NMM, DMF, 80 °C, yield 34ꢀ95%; (f) AcOH/H O (1:1), 40 °C or 1M HCl(aq),
2
acetonitrile, 23 °C, yield >70%.
Pharmacological profiling
Our lead compound cyclopropyl analog 20 was evaluated in a caspase selectivity panel and we
were pleased by the results indicating complete (>1000 fold) selectivity vs the apoptotic caspases
3, 6, 7 and 8. In addition, selectivity vs inflammatory caspase 5 was 300ꢀfold.
Table 3. Cellular potency and biochemical selectivity vs apoptotic and inflammatory caspases of
uracil 20 shows the compound is potent and selective.
ACS Paragon Plus Environment
26

Page 27 of 84
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2
0
a
a
a
a
a
a
IC Caspaseꢀ1
9.8
>9999
2980
>9999
>9999
>9999
38
5
0
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
IC Caspaseꢀ3
5
0
IC Caspaseꢀ5
5
0
IC Caspaseꢀ6
5
0
IC Caspaseꢀ7
5
0
IC Caspaseꢀ8
5
0
b
IC THP1
5
0
b
IC Macrophage
160
5
0
a
b
Biochemical IC evaluated in nM at K concentration of substrate. Cellular IC evaluated
5
0
M
50
in nM.
Ex-Vivo and In Vitro Pharmacokinetics
Next we wanted to evaluate the pharmacokinetic profile of 20. Topical drugs should provide
high tissue exposure in the target skin layer, the dermis in this case, with minimal plasma
exposure to provide a high safety margin. A high skin flux is needed to achieve high exposure in
2
the deeper skin layers such as the dermis (in our experience, at the very least 1 µmol/cm /h, but
2
10ꢀ100 µmol/cm /h is a safer range), but since this is vascularized, high hepatic clearance is also
50
needed to keep plasmatic exposure low. Skin exposure was measured at micromolar levels in
dermis after a single topical application both in mice PK and in a cryogenically preserved human
skin explant (Table 4). Furthermore, the maximum systemic exposure in mice following topical
6
administration was only 1 nM and the intrinsic clearance in human hepatocytes is 9.8 µL/min/10
cells, thus minimizing risk of adverse effects.
Table 4. Skin penetration profile of 20: >3000ꢀfold tissue selectivity between dermis and plasma.
ACS Paragon Plus Environment
27

Journal of Medicinal Chemistry
Page 28 of 84
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
2
0
4.3
4.2
a
Flux
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
b
b
Mouse dermis Cmax
Mouse plasma Cmax
0.0013
4.5
Human dermis
concentration
c
Human hepatocyte
9.8
d
clearance
a
2
51 b
Flux in µmol/cm /h calculated using Fick’s law and Potts and Guy K .
topical (0.1 µmol/cm applied using a 10 mM PG/EtOH 3/7 solution) mouse PK. Concentration
in µM after topical (0.2 µmol/cm applied using a 40 mM PG/EtOH 3/7 solution) on a
cryogenically preserved human skin explant. Intrinsic clearance in µL/min/10 cells.
CONCLUSION
C
in µM after
max
p
2
c
2
d
6
1
Acne is a skin disease affecting the quality of life of 650 million people globally. Inhibition of
cysteineꢀaspartic protease Caspaseꢀ1 was identified as a potential new approach to treat
inflammatory acne due to its role in the formation of ILꢀ1β. Structureꢀbased drug design was
therefore used to identify several novel scaffolds with an electrophilic warhead. The uracil series
was thus identified and optimized to deliver 20 as a potent, soluble and stable inhibitor of
caspaseꢀ1. Key structural improvements included a) the addition of cyclopropyl substituent in
P2, which boosted cellular potency, solubility and chemical stability, and b) the use of an oꢀtolyl
derivative in P4 to break πꢀconjugation and eliminate the risk of phototoxicity. Further profiling
of 20 revealed it was selective vs apoptotic caspases, was able to expose the mouse and human
ACS Paragon Plus Environment
28

Page 29 of 84
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
dermis at micromolar levels after topical administration and yet did not give rise to a systemic
exposure.
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
EXPERIMENTAL SECTION
Caspase-1 Enzymatic Assay
Human recombinant Caspaseꢀ1, assay buffer (50mM HEPES pH=7.4, 100mM NaCl, 0.1%
CHAPS, 1mM EDTA, 10% Glycerol, 10mM DTT) and proꢀfluorescent peptidic substrate (ACꢀ
WEHDꢀAMC) were purchased from Enzo life sciences. 4µL of 200µM or 10µM final
concentration of substrate (corresponding to its K ) was dispensed in a 384ꢀwell microplate
M
together with 4µL of appropriate dilutions of the evaluated compounds (10 000ꢀ0.04nM, 1%
DMSO). 2µL of enzyme was then added at a final concentration allowing for the conversion of
less than 20% of substrate after 20 minutes at room temperature. Free AMC was measured by
fluorescence detection (λex=340nm, λem=450nm). Percent inhibition was calculated from the
averaged values of the uninhibited and the uncatalyzed controls, and IC50 values were
determined by fitting the concentrationꢀresponse data with a fourꢀparameter logistic equation.
Caspase Panel Enzymatic Assays
Human recombinant Caspase proteins (ie Caspaseꢀ3, Caspaseꢀ5, Caspaseꢀ6, Caspaseꢀ7 and
Caspaseꢀ8), assay buffer (50mM HEPES pH=7.4, 100mM NaCl, 0.1% CHAPS, 1mM EDTA,
1
0% Glycerol, 10mM DTT) and proꢀfluorescent peptidic substrates (ACꢀWEHDꢀAMC for
Caspaseꢀ1 / ꢀ5, AcꢀDEVDꢀAMC for Caspaseꢀ3 / ꢀ7, AcꢀVEIDꢀAMC for Caspaseꢀ6 and Acꢀ
IETDꢀAMC for Caspaseꢀ8) were purchased from Enzo life sciences. 4µL of Caspaseꢀspecific
ACS Paragon Plus Environment
29