Topical ‘dual-soft’ glucocorticoid receptor agonist for dermatology

Article abstract of DOI:10.1016/j.bmcl.2020.127402

Steroidal glucocorticoids (GR agonists) have been widely used for the topical treatment of skin disorders, including atopic dermatitis. They are a very effective therapy, but they are associated with both unwanted local effects in the skin (skin thinning/atrophy) and systemic side effects. These effects can limit the long-term utility of potent steroids. Here we report on a topically delivered non-steroidal GR agonist, that has the potential to deliver high efficacy in the skin, but due to rapid metabolism in the blood & liver (“dual-soft”) it should have greater systemic safety than existing treatments. In addition, compared to less selective steroidal GR agonists, the new non-steroidal Selective Glucocorticoid Agonists (SEGRAs) have the potential to avoid the skin atrophy observed with existing topical steroids. Due to its potential for reduced skin atrophy and low systemic exposure, LEO 134310 (17) may be suitable for long term topical treatment of skin diseases such as atopic dermatitis and psoriasis.

Full text of DOI:10.1016/j.bmcl.2020.127402

Bioorganic&MedicinalChemistryLetters30(2020)127402
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Topical ‘dual-soft’ glucocorticoid receptor agonist for dermatology
⁎
Kevin N. Dacka, , Patrick S. Johnsona, Krister Henrikssona, Stefan Eirefelta, Martin A. Carnerupa,
Martin Stahlhuta, Anna K. Ollerstamb
a LEO Pharma A/S, Industriparken 55, 2750 Ballerup, Denmark
b AstraZeneca, Laboratory Animal Sciences, Drug Safety and Metabolism, IMED Biotech Unit, Pepparedsleden 1, Mölndal, Sweden
A R T I C L E I N F O
A B S T R A C T
Keywords:
Steroidal glucocorticoids (GR agonists) have been widely used for the topical treatment of skin disorders, in-
cluding atopic dermatitis. They are a very effective therapy, but they are associated with both unwanted local
effects in the skin (skin thinning/atrophy) and systemic side effects. These effects can limit the long-term utility
of potent steroids.
Dermatology
Non-steroidal
SEGRA
Soft-drug
Here we report on a topically delivered non-steroidal GR agonist, that has the potential to deliver high
efficacy in the skin, but due to rapid metabolism in the blood & liver (“dual-soft”) it should have greater systemic
safety than existing treatments. In addition, compared to less selective steroidal GR agonists, the new non-
steroidal Selective Glucocorticoid Agonists (SEGRAs) have the potential to avoid the skin atrophy observed with
existing topical steroids.
Glucocorticoid receptor agonist
Due to its potential for reduced skin atrophy and low systemic exposure, LEO 134310 (17) may be suitable for
long term topical treatment of skin diseases such as atopic dermatitis and psoriasis.
For more than 70 years steroidal glucocorticoids (GR agonists) have
been widely used for the topical treatment of skin disorders, including
atopic dermatitis.1 They are a very effective therapy but are associated
with both unwanted local effects in the skin (skin thinning/atrophy)
and systemic side effects. These effects can limit the long-term utility of
potent steroids.2
Substitution on the N1-aryl can access a region in the GR that is
often referred to as the ‘meta-channel’,6 that has proven beneficial for
obtaining greater GR selectivity over related steroid receptors. A good
example of this is AZD7594, (Fig. 1) an inhaled development candi-
date.7
In contrast to the small inhaled doses, a topically applied product
must avoid photosensitivity and/or phototoxicity, especially when ap-
plied to sun exposed skin on face, hands and limbs. The ICH guidelines8
require further testing if the compound absorbs in the range > 290 nm
and most N1-aryl-indazoles have strong UV absorbance in that range
and may require further de-risking of parent and metabolites in vitro
and in in vivo.
The glucocorticoid receptor (GR) is a transcription factor belonging
to the nuclear hormone receptor family. Ligand activation triggers both
positive and negative regulation of gene transcription and subsequently
pleiotropic effects on metabolism and the immune system.
Glucocorticoids are widely prescribed for the treatment of in-
flammatory and autoimmune conditions and are highly effective, but
their therapeutic use can be limited due to adverse effects following
chronic administration.2
Therefore, we initially designed and investigated isosteric novel
templates based around pyridyl and benzoyl amides, which could use
the carbonyl oxygen as an isostere for the indazole N-2 position. These
isosteres have a smaller chromophore, than an indazole, and conse-
quently a lowered risk of phototoxicity. We theorized that maintaining
coplanarity between the carbonyl and the aryl would give the closest
mimetic of the indazole. In particular, the pyridyl amide is expected to
be almost planar due to the attraction between the amide NH and
pyridyl N, whereas, the benzamide, due to steric clash between the
amide NeH and ortho Ar-H of the benzoic acid core, is significantly
twisted out of plane. (Fig. 2).
Numerous attempts to dissociate anti-inflammatory effects from
systemic metabolic side effects, using new non-steroidal Selective
Glucocorticoid Agonists (SEGRAs), have been described3 wherein the
rigid steroidal core is replaced by alternative non-steroidal scaffolds. An
alternative scaffold of note is the suitably substituted indazole system,4
wherein the N2 position of the indazole ring system has been shown, by
X-ray crystallography,5,7 to be a mimic of the A-ring ketone of steroidal
agonists, which can form hydrogen bonds to Gln570 within the GR
binding site.
⁎ Corresponding author.
E-mail address: kevin.dack@leo-pharma.com (K.N. Dack).
https://doi.org/10.1016/j.bmcl.2020.127402
Received 12 May 2020; Received in revised form 6 July 2020; Accepted 7 July 2020
Availableonline13July2020
0960-894X/©2020ElsevierLtd.Allrightsreserved.

K.N. Dack, et al.
Bioorganic&MedicinalChemistryLetters30(2020)127402
Table 2
"Meta"
Channel
Optimisation of piperidine and lactone moieties for potency and blood in-
stability.
AZD7594
Phase II
COPD & Asthma
(inhaled)
Fig. 1. AZD7594.
6-13
Entry Xa Lactone hPBMCb IC50 (nM) Bloodc T½ (mins.) KCd T½ (mins)
H-bond
acceptor
H-bond
acceptor
H-bond
acceptor
6
S
5-R
5-R
5-S
5-S
4-R
4-R
4-S
4-S
7
> 120
51
> 360
nd
7
R
S
90
73
107
7
8
83
nd
Indazole
Pyridyl amide
"Planar"
Benzamide
"Twisted"
9
R
S
R
S
44
nd
10
11
12
13
< 10
< 10
102
17
> 360
> 360
> 360
nd
21
8
Fig. 2. Indazole isosteres.
R
34
Table 1
a
b
X = Piperidine enantiomer.
Indole Isosteres.
Inhibition of LPS-induced TNF-α release in peripheral blood mononuclear
cells (PBMCs) – geometric mean of minimum of 2 independent experiments.
c
d
Stability half-life in human whole blood.
Stability half-life in cultured human keratinocytes.
molecules enter the systemic circulation, thus limiting the systemic
adverse effects. Crucial to this concept is having assays to determine
both the stability in the skin (good keratinocyte stability) and rapid
metabolism in whole blood to less active metabolites.
1
2-4
Entry
A
R1
hPBMCa IC50 (nM)
2
Comment
A series of compounds were prepared in which the Boc-protecting
group of 5 was replaced by a set of esters and lactones to explore the
balance of GR agonist potency and ‘soft drug’ potential. It was chal-
lenging to find esters that could achieve a suitable balance of potency
with both rapid hydrolysis by esterases in blood and stability in kera-
tinocytes, however, the lactones 6–13 (Table 2) provided some ex-
amples that satisfied these criteria. The (S)-piperidine enantiomers (6,
10 and 12) had full agonist PBMC IC50 < 10 nM and were more potent
GR agonists than their R- enantiomers. In addition, the (4R)-lactones
(10 & 11) were shown to be very unstable in human whole blood and
rapidly hydrolyzed to the open hydroxy acids in < 10 min but were also
sufficiently stable in human keratinocytes (half-life > 360 min). The
lactone hydrolysis was mainly paraoxonase (PON) mediated, as the
addition of EDTA stabilized12 the lactones in assays. Compound 10
(LEO 131928A) had an encouraging, balanced profile and was in-
vestigated further.
1
2
3
4
Indazole
CH
N
Phenyl
~5000
Partial agonist
Full agonist
Full agonist
Phenyl
26
9
N
Cyclohexyl
a
Inhibition of LPS-induced TNF-α release in peripheral blood mononuclear
cells (PBMCs) – geometric mean of minimum of 2 independent experiments.
A small range of initial analogues (2–4, Table 1) were made to ex-
plore the potential for achieving agonist activity in a hPBMC assay9
stimulated with LPS to secrete TNFα. The benzamide 2 was a weak,
partial agonist, but the more planar pyridyl amide 3 gave full agonism
and was only 10-fold less potent than an indazole comparator (1). Sa-
turating the pendant phenyl to a cyclohexyl 4 improves the potency
further and removes a potential aniline toxicophoric structural alert.
The pyridyl amides were further explored by making a larger range
of amides (data not shown) to look for vectors to access the ‘meta-
channel’. Amongst the numerous analogues prepared, one of the most
promising was the 3-(N1-Boc) piperidine amide 5 (Fig. 3), with a full-
agonist IC50 of 8 nM
LEO 131928 (10) was rapidly metabolized (Fig. 4) in blood and
hepatocytes to give the hydroxy acid 14 as the major metabolite. This
was consistent across all major species (human, dog, pig, rat and
mouse). The acid 14 was > 100-fold less active as a GR agonist in
human PBMCs and therefore LEO 131928 fits the profile of a ‘dual-soft’
molecule.
We next turned our attention to introducing a metabolically soft
spot. The concept of ‘soft drugs’ design10 is to rapidly convert an active
parent molecule by metabolism to a less active derivative to reduce
systemic effects. In a recent publication,11 LEO scientists described
‘dual-soft’ PDE4 inhibitors that can act locally when applied topically
and are then metabolized extra-hepatically in the blood as the
In addition, 10 was sufficiently stable towards chemical hydrolysis
at pH 5.5 (> 95% remaining after 16 h at 37 °C) but slowly hydrolyzed
at pH 7 (82% remaining after 16 h at 37 °C). The photostability was
excellent and 10 was (as predicted) also non-phototoxic in 3 T3 cells
(NRU Assay)13 with viability IC50 > 125 μg/mL in both the presence
and absence of UV light.
"Meta"
Channel
With an optimized meta-channel moiety with favorable dual-soft
Hepatocytes
& Whole
Blood
5
14
10 (LEO 131928)
Fig. 3. Compound 5.
Fig. 4. Whole blood and hepatic metabolism of cmpd. 10.
2

K.N. Dack, et al.
Bioorganic&MedicinalChemistryLetters30(2020)127402
Table 3
Selected SAR.
b
a
A
Taniguchi
Lactone
B
Fig. 6. Route to key chiral building blocks A and B: Steps: a. RuCl3, NaIO4,
CCl4, MeCN, H2O, RT (70%) b. HATU, DIPEA, CH2Cl2, RT, 3-(S)-hydro-
xypiperidine 3 h (90%).
10-17
Entry
R1
R2
A/X
hPBMCa IC50 (nM)
10
15
16
17
18
19
CF3
Me
Me
Me
cPr
cPr
cPr
N /NH
7
a
b, c
CF3
CH / O
CH / NH
CH / O
CH / O
CH / O
2
CF3
79
2
2R-THF
2S-THF
CF3
X
22
1
a
Inhibition of LPS-induced TNF-α release in peripheral blood mononuclear
cells (PBMCs) – geometric mean of minimum of 2 independent experiments.
d, e
f
g
properties in hand, we then reconsidered alternatives to the pyridyl
amide as an indazole isostere and investigated benzoate esters as they
should also maintain coplanarity between the phenyl and carbonyl
(Table 3). The pyridyl esters analogues had insufficient chemical re-
sistance to aqueous hydrolysis and were not pursued further, however,
the benzoate analogue of 10 (i.e. 15) had increase GR agonist potency
and all other properties were similar to 10.
h, i, j
As previously noted with comparison of 2 and 3, the benzamide
analogue (16) was 10-fold weaker than 10 and supported the hypoth-
esis of attaining coplanarity between the aryl and the amide carbonyl.
To select a development candidate, we investigated a larger range of
analogues in both the pyridyl amides and benzoate esters, in which the
amide R1 and the aryl substituent R2 were varied to get the optimal
balance of properties such as potency, solubility and stability.14 From
these analogues, compound 17 was selected as a development candi-
date for further in vitro and in vivo profiling.
LEO 134310 (17)
Y
Mpt 146 o
C
Fig. 7. Discovery route to LEO 134,310 (17): Steps: a. 4-Bromo-phe-
nylmagnesium bromide, THF, 0 °C-RT, 16 h (43%) b. Al(OiPr)3, IPA, 60 °C, 16 h
(97%) c. TFA, CH2Cl2, 0 °C-RT, 4 h (72%) d. 4-fluoro-benzonitrile, NaH, NMP,
60 °C, 1 h (81%) e. BOC2O, DMAP, THF, RT, 2 h (72%) f. cPropyl boronic acid,
(cHexyl)3P, Pd(OAc)2, Toluene/water (10:1), 120 °C (96%) g. 32% aq. NaOH,
EtOH, 65 °C, 72 h (92%) h. EDC, DMAP, CH2Cl2, B, RT, 18 h (94%) i. 4 M HCl
in dioxane, RT, 1 h (100%) j. HATU, DIPEA, DMF, (R)-Furoic acid (71%).
LEO 134310 (17) has low nM potency in the TNFα stimulated
PBMC assay, comparable to dexamethasone, and like 10 and most
analogues containing the (4R)-lactone moiety, it is rapidly hydrolyzed
to the less active (hPBMC IC50 = 260 nM) hydroxy acid 20 in whole
blood and hepatocytes of human and all main laboratory mammals
(Fig. 5). Importantly, the stability in skin keratinocytes was very good
(half-life > 6 h). Due to the extrahepatic metabolism via blood and
other tissues, the in vivo PK in rat and dog after intravenous dosing gave
systemic clearance greatly in excess of liver blood flow (500 mL/min/
kg in rat).
skin atrophy compared to betamethasone valerate when applied for
28 days to minipig skin in vivo.15
The Medicinal Chemistry route14 to LEO 134310 (17) is illustrative
of the synthesis towards most of the analogues discussed herein.
The key ‘dual-soft’ lactone acid A is prepared by oxidative cleavage
of the Taniguchi lactone17 which is then coupled to 3-(S)-hydro-
xypiperidine to give the lactone alcohol fragment B (Fig. 6).
The central beta amino alcohol X (Fig. 7) is prepared by disclosed
methods14 starting from an aryl-Grignard addition to the Weinreb
amide of Boc (S)-alanine and subsequent chelation controlled Meer-
wein-Ponndorf-Verley stereoselective reduction18 of the ketone fol-
lowed by deprotection. Arylation of the amino alcohol X is conducted
using sodium hydride in N-methyl pyrrolidinone to control O-arylation
over N-arylation, reprotection and palladium catalyzed cross-coupling,
with cyclopropyl-boronic acid, followed by nitrile hydrolysis gives the
benzoic acid building block Y.
The in vitro and in vivo safety and efficacy was investigated in
greater detail and is covered in greater depth in an article in press.15
Briefly, LEO 134310 was > 100-fold selective for GR over the related
AR, MR and PR (data not shown) and clean when assessed in all off-
target assays, hERG, Cyp and DDI assays and genetic toxicology testing.
It also demonstrated efficacy after topical application in a mouse TPA
induced ear inflammation model,16 with a greatly increased therapeutic
index over betamethasone valerate with regards to efficacy relative to
systemic adverse effects. In addition, LEO134310 demonstrated target
engagement and biomarker changes equivalent to betamethasone va-
lerate when applied to human skin in vitro and had significantly less
The benzoate ester linkage is formed between benzoic acid Y and
alcohol B using EDC and DMAP and finally, the Boc-protection is re-
placed with the (R)-Furoic amide to give LEO 134310 (17). Further
Process Chemistry improvements to the synthesis will be disclosed in a
future publication.
In conclusion, we have applied the concept of ‘dual-soft’ drug design
towards identifying LEO 134310 (17) as a potent and selective GR
agonist that shows target engagement and relevant GR mediated bio-
marker changes in mouse and pig after topical application. The po-
tential for the unwanted systemic side-effects of topical glucocorticoids
appears to be reduced preclinically by the rapid ‘deactivation’ in blood
to the hydroxy acid metabolite, with > 100-fold less GR potency. LEO
Whole
Blood
Metabolite (20)
LEO 134310 (17)
>100x reduced GR agonism
GR agonism hPBMC EC50 = 2 nM
Fig. 5. Whole blood metabolism of cmpd. 17.
3

K.N. Dack, et al.
Bioorganic&MedicinalChemistryLetters30(2020)127402
134310 was taken to Phase I studies19 to ascertain the safety, toleration
1016/j.bmcl.2010.12.121.
7. Hemmerling M, et al. J Med Chem. 2017;60:8591–8605. https://doi.org/10.1021/
acs.jmedchem.7b01215.
and efficacy – results will be disclosed in due course.
8. Photosafety evaluation of pharmaceuticals S10, https://www.fda.gov/regulatory-
information/search-fda-guidance-documents/s10-photosafety-evaluation-pharma-
ceuticals.
Declaration of Competing Interest
9. It was possible to distinguish partial and full agonists in the TNFα release assay in-
dicating that both transactivation and transrepression mechanisms contributed to the
cellular response. Unless noted, compounds mostly behaved as full agonists.
Reference SEGRA compounds (and hydrocortisone) known to mediate reduced GR
transactivation behaved as partial agonists in this assay. However, the mechanistic
details leading to their partial effects were not addressed. Experimental protocol is
included in the Supplementary Notes.
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
ence the work reported in this paper.
Acknowledgements
10. (a) Bodor N. Med Res Rev. 1984;4(4):449–469. https://doi.org/10.1002/med.
2610040402;
We would like to thank the numerous LEO Pharma colleagues in-
(b) Aprile S, et al. Drug Discovery Today. 2019;24:2234–2246. https://doi.org/10.
1016/j.drudis.2019.08.007.
volved in the discovery and early development of LEO 134310.
11. Larsen J, Lambert M, Pettersson H, Vifian T, Larsen M, Ollerstam A, Hegardt P,
Eskilsson C, Laursen S, Skak-Nielsen T, Hansen LM, Knudsen NØ, Stilou TG, Nielsen
SF. Discovery of LEO 39652 a Novel “Dual-soft” PDE4 Inhibitor for Topical
Treatment of Atopic Dermatitis, article in press.
Appendix A. Supplementary data
The synthesis of 17 and all intermediates used in its preparation are
described along with the protocols for the PBMC functional assay, and
the whole blood and keratinocyte stability assays. Supplementary data
to this article can be found online at https://doi.org/10.1016/j.bmcl.
2020.127402.
12. (a) Mackness MI. Biochem Biophys Res Commun. 1998;242. https://doi.org/10.1006/
bbrc.1997.7872;
(b) Draganov DI, et al. J Lipid Res. 2005;46:1239–1247. https://doi.org/10.1194/jlr.
M400511-JLR200.
13. Kim K. Toxicol. Res. 2015;31:97–104. https://doi.org/10.5487/TR.2015.31.2.097.
14. Johnson PS, Dack KN, Henriksson K. Preparation of aminoalkoxypyridinecarbonyl-
amino- piperidine carboxyate derivatives and analogs for use as non-steroidal glu-
cocorticoid receptor modulators. WO 2017046096 A1.
References
15. Eirefelt S, Stahlhut M, Svitacheva N, Ewald DA, Marstrand TT, Ollerstam A,
Norsgaard H. Non-clinical evaluation of a novel non-steroidal glucocorticoid agonist
with low systemic exposure and reduced potential for skin atrophy, article in press.
16. Stanley PL, Steiner S, Havens M, Tramposch KM. Skin Pharmacol.
1991;4(4):262–271. https://doi.org/10.1159/000210960.
1. Ahluwalia A. Mediators Inflamm. 1998;7(3):183–193. https://doi.org/10.1080/
09629359891126.
2. Hengge UR, et al. J Am Acad Dermatol. 2006;54(1):1–15. https://doi.org/10.1016/j.
jaad.2005.01.010.
3. Sundahl N. Pharmacol Therapeutics. 2015;152:28–41. https://doi.org/10.1016/j.
pharmthera.2015.05.001.
17. Kieseritzky V, et al. OPRD. 2014;18(5):643–645. https://doi.org/10.1021/
op500096j.
4. Hemmerling M. Bio & Med Chem Letts. 2016;26:5741–5748. https://doi.org/10.
1016/j.bmcl.2016.10.052.
18. Yin J, et al. J Org Chem. 2006;71(2):840–843. https://doi.org/10.1021/jo052121t.
19. Clinical Trial to Assess Safety, Tolerability and the Pharmacodynamic Effect of
Different Concentrations of a New Anti-inflammatory Substance in Subjects with
Chronic Plaque Psoriasis (NCT03669757).
5. Biggadike K, et al. J Med Chem. 2008;51:3349–3352. https://doi.org/10.1021/
jm800279t.
6. Biggadike K, et al. Bioorg Med Chem Lett. 2011;21:1126–1133. https://doi.org/10.
4