Discovery of potent, orally bioavailable in vivo efficacious antagonists of the TLR7/8 pathway

Article abstract of DOI:10.1016/j.bmcl.2020.127366

Antagonism of the Toll-like receptors (TLRs) 7 and TLR8 has been hypothesized to be beneficial to patients suffering from autoimmune conditions. A phenotypic screen for small molecule antagonists of TLR7/8 was carried out in a murine P4H1 cell line. Compound 1 was identified as a hit that showed antagonistic activity on TLR7 and TLR8 but not TLR9, as shown on human peripheral blood mononuclear cells (hPBMCs). It was functionally cross reactive with mouse TLR7 but lacked oral exposure and had only modest potency. Chemical optimization resulted in 2, which showed in vivo efficacy following intraperitoneal administration. Further optimization resulted in 8 which had excellent in vitro activity, exposure and in vivo activity. Additional work to improve physical properties resulted in 15, an advanced lead that had favorable in vitro and exposure properties. It was further demonstrated that activity of the series tracked with binding to the extracellular domain of TLR7 implicating that the target of this series are endosomal TLRs rather than downstream signaling pathways.

Full text of DOI:10.1016/j.bmcl.2020.127366

Journal Pre-proofs
Discovery of potent, orally bioavailable in vivo efficacious antagonists of the
TLR7/8 pathway
Phil B. Alper, Deane Jonathan, Claudia Betschart, David Buffet, Géraldine
Collignon Zipfel, Perry Gordon, Janice Hampton, Stuart Hawtin, Maureen
Ibanez, Tao Jiang, Tobias Junt, Thomas Knoepfel, Bo Liu, Jillian Maginnis,
Una McKeever, Pierre-Yves Michellys, Daniel Mutnick, Bishnu Nayak,
Satoru Niwa, Wendy Richmond, James S. Rush, Peter Syka, Yi Zhang,
Xuefeng Zhu
PII:
S0960-894X(20)30477-7
DOI:
Reference:
https://doi.org/10.1016/j.bmcl.2020.127366
BMCL 127366
To appear in:
Bioorganic & Medicinal Chemistry Letters
Received Date:
Revised Date:
Accepted Date:
23 April 2020
17 June 2020
20 June 2020
Please cite this article as: Alper, P.B., Jonathan, D., Betschart, C., Buffet, D., Collignon Zipfel, G., Gordon, P.,
Hampton, J., Hawtin, S., Ibanez, M., Jiang, T., Junt, T., Knoepfel, T., Liu, B., Maginnis, J., McKeever, U.,
Michellys, P-Y., Mutnick, D., Nayak, B., Niwa, S., Richmond, W., Rush, J.S., Syka, P., Zhang, Y., Zhu, X.,
Discovery of potent, orally bioavailable in vivo efficacious antagonists of the TLR7/8 pathway, Bioorganic &
Medicinal Chemistry Letters (2020), doi: https://doi.org/10.1016/j.bmcl.2020.127366
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover
page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version
will undergo additional copyediting, typesetting and review before it is published in its final form, but we are
providing this version to give early visibility of the article. Please note that, during the production process, errors
may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
© 2020 Published by Elsevier Ltd.

Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com
Discovery of potent, orally bioavailable in vivo efficacious antagonists of the
TLR7/8 pathway
Alper, Phil B.a*; Deane Jonathana*; Betschart, Claudiab; Buffet, Davidb; Collignon Zipfel, Géraldine b;
Gordon, Perrya; Hampton, Janicea; Hawtin, Stuartb; Ibanez, Maureena; Jiang, Taoa; Junt, Tobiasb;
Knoepfel, Thomasb; Liu, Boa; Maginnis, Jilliana; McKeever, Unab; Michellys, Pierre-Yvesa; Mutnick,
Daniela; Nayak, Bishnua; Niwa, Satorub; Richmond, Wendya; Rush, James S.b; Syka, Petera; Zhang, Yia;
Zhu, Xuefenga
aGenomics Institute of the Novartis Research Foundation 10675 John J. Hopkins Dr,. San Diego, CA 92121, USA
bDepartment of Autoimmunity, Transplantation and Inflammation, Novartis Institutes of Biomedical Research, 4002 Basel, Switzerland
*) To whom correspondence should be addressed
ARTICLE INFO
ABSTRACT
Article history:
Received
Revised
Accepted
Available online
Antagonism of the Toll-like receptors (TLRs) 7 and TLR8 has been hypothesized to be beneficial
to patients suffering from autoimmune conditions. A phenotypic screen for small molecule
antagonists of TLR7/8 was carried out in a murine P4H1 cell line. Compound 1 was identified as
a hit that showed antagonistic activity on TLR7 and TLR8 but not TLR9, as shown on human
peripheral blood mononuclear cells (hPBMCs). It was functionally cross reactive with mouse
TLR7 but lacked oral exposure and had only modest potency. Chemical optimization resulted in
2, which showed in vivo efficacy following intraperitoneal administration. Further optimization
resulted in 8 which had excellent in vitro activity, exposure and in vivo activity. Additional work
to improve physical properties resulted in 15, an advanced lead that had favorable in vitro and
exposure properties. It was further demonstrated that activity of the series tracked with binding
to the extracellular domain of TLR7 implicating that the target of this series are endosomal TLRs
rather than downstream signaling pathways.
Keywords:
TLR7
TLR8
antagonist
immunomodulation
medchem
2009 Elsevier Ltd. All rights reserved.
Toll-like Receptors (TLRs) are critical components of the
innate immune response to Danger Associated Molecular Patterns
(DAMPs)1, which are molecular signatures of infection. For
example, endosomal hypomethylated CpG rich DNA and single-
stranded RNA (ssRNA) can indicate the presence of infection by
viruses or bacteria. As such, TLR7 and TLR8 are capable of
detecting endosomal ssRNA, and TLR9 is capable of detecting
endosomal CpG-rich DNAs, to initiate innate responses against
pathogens. In recent years, several lines of evidence suggested
that in some autoimmune diseases, the body’s own RNA or DNA
may gain access to endosomes, aberrantly activate TLR7/8/9 and
support the development of chronic inflammation.2 Therefore,
inhibiting endosomal TLRs has been proposed as a novel
therapeutic approach for autoimmune diseases.2 There have since
been significant efforts in this space.3-8 Here we describe the
identification and characterization of a novel series of low
molecular weight inhibitors of the TLR7/8 pathway.
P4H19, for the ability to inhibit IL-6 production induced by the low
molecular weight TLR7 agonist, imiquimod.10 Compounds
showing activity of <10 M and not interfering with cell viability
were further studied on P4H1 cells for inhibition of IL-6
production induced by the TLR4 agonist, Lipopolysaccharide
(LPS).11 Only compounds without activity in this assay were
further pursued.
N
N
N
N
N
N
F
F
H2N
H2N
N
N
O
O
N
N
NH
1
2
(Hit compound)
mDC (TLR7): 0.17 M
mDC (Tox): 0.90 M
mSplenocyte (TLR7): 1.0 M
mSplenocyte (TLR9): 2.0 M
hPBMC (TLR7): 0.31 M
hPBMC Tox: 30 M
hPBMC (TLR7): 0.053 M
THP-1 (TLR8): 0.026 M
hPBMC (TLR9): 2.3 M
mSplenocyte (TLR7): 0.097 M
mSplenocyte (TLR9): 1.3 M
In search of small molecule TLR7/8 inhibitors, two million
compounds were screened on the murine dendritic cell (mDC) line

Figure 1. A screening hit that was cross reactive with human cells and a
derivative used in an in vivo experiment. Values are IC50.
assess whether the hits were crossreactive to human TLR7. For
this purpose, human peripheral blood mononuclear cells
(hPBMCs) were stimulated with the TLR7/8 ligand ssRNA40,
with the TLR7-dependent cytokine IFN-α as a readout It was
found that most of the hits from the mDC screen were not cross
reactive to hPBMCs.
Next, the antagonist activity of the TLR4-inactive hits derived
from the above TLR7 screen was confirmed on mouse splenocytes
using ssRNA40 as TLR7 agonist and IL-6 as a readout. Selectivity
was ascertained on mouse splenocytes using the TLR9 agonist
CpG1585 (TLR9) and IL-6 as a readout. The next step was to
Table 1. Structure activity relationship of the basic amine bearing tail groupa
H2N
N
N
N
N
F
N
H
N
H
N
NH2
NH2
NH2
H
OH
O
Assayb,c
3
4
5
6
7
8
9
0.20
0.075
5.3
hPBMC (TLR7)
THP-1 (TLR8)
hPBMC (TLR9)
0.17
0.06
3.4
0.095
0.088
1.2
0.50
0.01
6.5
0.071
0.02
2.2
0.025
0.005
1.4
0.006
0.003
2.7
mSplenocyte (TLR7)
mSplenocyte (TLR9)
aValues are IC50 in µM.
0.22
2.0
0.12
1.2
0.15
1.2
0.11
1.9
0.038
0.84
1.2
1.2
2.8
bDetails of assay conditions are in Supplemental Information.
cCellular toxicity was run as a negative control, and was not observed up to the highest level of compound tested (10 M)
Compound 112 shown in Figure 1 was found to be an exception.
This hit not only confirmed in the screening assay using the
synthetic TLR7 agonist imiquimod but also demonstrated
moderate antagonist activity in mouse splenocytes using ssRNA
as a TLR7 agonist. Importantly, TLR7 antagonist activity was
maintained in hPBMC’s, and this without an indication of cellular
toxicity. Therefore 1 was selected for further optimization.
THP-1 cells14 , were stimulated with the synthetic TLR7/8 agonist
resiquimod (R848)10 and the TLR8-dependent cytokine TNFα was
read out. To investigate TLR9 activity in human cells, hPBMCs
were stimulated with the TLR9 agonist ODN2216 reading out
IFNα.
It was found that 2 demonstrated 97 nM activity on TLR7 in
mouse splenocytes with substantial selectivity over TLR9. In
human cells, the potency on TLR7 and TLR8 was even slightly
better with a high degree of selectivity over TLR9. Although 2 did
not have adequate oral exposure (Table 3), exposure following i.p.
dosing was considered sufficient to demonstrate in vivo efficacy.
With the substitution of the methyl piperazine for piperidine, 2
was obtained. At this point, two additional assays were used to
enhance the characterization suite. Mice do not express a
functional TLR813, so to investigate TLR8 activity in human cells
Table 2. Stucture Activity Relationships for the rest of the scaffolda
H2N
H2N
N
H
H
N
N
N
N
O
N
O
N
N
NH2
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
F
F
O
O
O
O
NH2
NH2
NH2
NH2
NH2
OH
Assayb
10
11
12
0.045
0.018
3.7
13
14
15
hPBMC (TLR7)
THP-1 (TLR8)
<0.001
0.001
3.3
0.046
0.040
3.4
<0.001
<0.001
1
0.003
<0.001
>5
<0.001
0.002
ND
hPBMC (TLR9)
mSplenocyte (TLR7)
mSplenocyte (TLR9)
ND
0.093
0.46
0.30
0.80
0.004
1.2
0.055
>10
ND
ND
ND
aValues are IC50 in µM., ND: No Data.
bDetails of assay conditions are in Supplemental Information.

Figure 2 shows that i.p. administration of 2 at 30 mg/kg to
female 129/Sv mice, blocked both IFN-α and IL6 secretion in
serum after a challenge with ssRNA006. The secretion of the same
cytokines in blood in response to CpG1585 was only modestly
affected. The exposure of 2 at 2 hours was 1.5 µM, consistent with
the in vitro IC50 (Figure 2).
secretion, whereas in species with both TLR7 and TLR8 one
expect that ssRNA would drive IFN-α and IL6 via TLR7 and
TLR8, respectively. The secretion of the same cytokines in
response to CpG1585 was only modestly affected. The exposure
2 at 2 hours was 300 nM, making the in vivo data consistent with
the in vitro derived activity data.
Based on these results, the SAR for the amine containing
portion of the molecule was further explored (Table 1).
Although the activity of 8 was encouraging, there were
challenges with the physical properties of this molecule, especially
the solubility (high throughput solubility of the HCl salt at pH 6.8:
0.023 µM). The team decided to pursue an exploration of
derivatives that were meant to enhance the properties of the
molecule. Table 2 shows some key compounds in this effort.
Shortening the linker to the piperidine to 1 carbon afforded 3
which was slightly less active but did exhibit very modest but
measurable p. o. exposure in mouse. Metabolite ID studies pointed
to robust metabolism of the piperidine ring so modifications were
pursued. Introduction of a bicyclic bridge afforded 4 which
maintained activity. The spirocyclic derivative 5 lost some
activity while moving the amine function out of the ring in 6 and
7 maintained or slightly improved activity. None of these changes
were effective in improving p. o. exposure in mouse
(Supplementary Material).
Figure 3: Compound 8 inhibits (A) TLR7/ssRNA006- but not (B)
TLR9/CpG1585-mediated cytokines in mice
Bars represent means of n=5 ± SD; corresponding blood exposures ± SD are
shown in blue above the respective bars. *, p<0.05, ****, p<0.0001, unpaired
t test.
Figure 2: Compound
2 inhibits (A) TLR7/ssRNA006- and (B)
Replacement of the fluorophenyl group with
a 3-methyl
TLR9/CpG1585-mediated cytokines in mice
pyridine afforded 10 which gave a substantial boost to potency as
well as allowing the formation of soluble salts. Head group
exploration afforded 11 which retained activity. Moving the
attachment point to the 4 position of the pyrazolo-pyrimidine ring
system afforded 12 which similarly maintained activity.
Additional exploration of the bicyclic head group combined with
the 3-methyl pyridine result afforded 13 which had very potent
activity in human and mouse cell lines. This optimized skeleton
now allowed the substitution of an alcohol for an amine to retain
activity (14). It was found that 13 was a substrate for aldehyde
oxidase (Supplementary Material).15 This would complicate
development since this enzyme is only expressed in higher species
and can result in highly compromised oral exposure once dosed in
these species. For this reason, the putative site of metabolism was
blocked with a methyl group on the 6-position of the pyrazolo-
pyridine ring. Additionally, it was found that blocking of the NH
was beneficial for p. o. exposure and the advanced lead 15 was
synthesized which demonstrated a lack of aldehyde oxidase
metabolism along with outstanding in vitro activity in both human
and mouse cells.
Bars represent means of n=5 ± SD; corresponding blood exposures ± SD are
shown in blue above the respective bars. **, p<0.01, ****, p<0.0001, unpaired
t test.
The [2,2,2] bicyclic derivative 8 was made to introduce steric
bulk next to the amine functionality with the hypothesis that this
would mitigate metabolism and fortuitously resulted in a
significant boost in activity both in human and mouse cellular
assays. Gratifyingly, 8 also had very good oral exposure when
dosed in suspension to Balb/c male mice (Table 3). Up to this
point, it had not been possible to identify any derivatives in this
series without an amine. As an exception, 9 was found to still have
an appreciable level of antagonism for TLR7 and TLR8, though
not in mouse cells.
When 8 was dosed as the HCl salt p. o. at 5 mg/kg to female
129/Sv mice (Figure 3), both IFN-α and IL6 secretion were
completely blocked in blood 2 h after challenge with ssRNA006.
As mentioned above, TLR8 is not functional in mice, which would
indicate that TLR7 is the major driver of both IFN-α and IL6

Table 3 shows the mouse PK parameters for selected
compounds. It should be noted that in the case of 8, a specific
crystal polymorph needed to be used in order to get reproducible
data. While both 8 and 10 exhibited roughly similar exposure
parameters, switching to the pyrrolo pyridine (13) head group
resulted in markedly improved exposure. The Vss was also
reduced from 15 to 5.1, a marked improvement. Substitution of
the primary amine for and alcohol (14) maintained excellent
exposure and further reduced the Vss to 3.1. Methylation of the
pyrazolo-pyridine head group (15) maintained good p. o. exposure
and excellent human potency, but was not optimal in IV clearance
and Vss.
ECD was a strong indication that these compounds mediate
inhibition by interaction with TLR7 itself.
In summary, we have demonstrated how a pathway based
screen served to find inhibitors of the human TLR7/8 pathway in
mouse dendritic cells. The team started with the hit 1 which had
modest activity and no oral exposure. After showing in vivo
efficacy with the tool compound 2, systematic SAR exploration
led to both increased on target potency and vastly improved p. o.
exposure culminating in the advanced lead 15. The team also
showed that the receptor itself was likely the target of the lead
series in a series of cell-based domain swapping experiments.
Acknowledgements
Parameter
2
0
8
10
13
14
36300c
4370c
11
15
28900
1400
44
p. o. AUC
(h*nM)
The authors would like to acknowledge all of our team
collaborators working at Novartis at the time of these efforts.
3046 2983 29600
p. o. Cmax (nM)
0
506
98
354
96
3230
23
References and notes
IV CL
(mL/min/kg)
165
1.
Vidya, M. K.; Kumar, V. G.; Sejian, V.; Bagath, M.; Krishnan, G.
Bhatta, R. Int. Rev. Immunol. 2018, 37, 20.
IV Vss (L/kg)
IV T1/2 (h)
F (%)
29
5.9
0
15
4.1
36
15
4.1
34
5.1
4.0
75
3.1
4.7
43
13.8
4.9
2.
3.
Junt, T.; Barchet, W. Nat. Rev. Imunol. 2015, 15(9), 529.
Lipard, G. B.; Zepp, C. M. Quinazoline Derivative Useful as Toll-
Like Receptor Antagonist. WO 2008152471, 2008.
Boivin, R.; Carlson, E.; Endo, A.; Hansen, H.; Hawkins, L. D.;
Ishizaka, S.; Mackey, M.; Narayan, S.; Satoh, T.; Schiller, S.
Tetrahydropyrazolopyrimidine compounds. US20130324547,
2013.
Carlson, E.; Hansen, H.; Mackey, M.; Schiller, S.; Ogawa, C.
Davis. H. Selectively Substituted Quinoline Compounds.
WO2015057659 A1.
Boivin, R.; Hansen, H., Ishizaka, S.; Mackey, M.; Schiller, S.;
Ogawa, C.; Narayan, S.; Bertinato, P.; Burger, G. Selectively
Substituted Quinoline Compounds. WO2015057655, 2015.
Hawkins, L. D. E6887: A novel and selective inhibitor of toll-like
receptors 7 and 8. Presented at the 252nd National Meeting,
American Chemical Society, Philadelphia, PA, MEDI 203, August
20-25, 2016.
Padilla-Salinas, R.; Anderson, R.; Sakaniwa, K.; Zhang, S.;
Nordeen, P.; Lu, C.; Shimizu, T.; Yin, H. H. Discovery of Novel
Inhibitors Targeting Toll-like Receptor 7 and 8. J. Med. Chem.
2019, 62, 22, 10221–10244.
150
4.
Table 3. PK parameters in mousea,b,c
aBalb\c male mice, 3 animals per cohort. 5mg/kg IV, 20 mg/kg p. o..
PEG300/D5W, 3:1 formulation.
5.
6.
7.
bp.o. – orally dosed, AUC- area under the curve, Cmax – maximum observed
concentration, IV CL – observed clearance when dosed intravenously, IV Vss
– observed steady state volume when dosed intravenously, IV T1/2 –
observed terminal half life when dosed intravenously, F – observed
bioavailability.
cp. o. formulation was 2.5 mg/mL suspension in 0.5%MC/0.5% Tween
Although the overall parameters of 15 looked compelling, there
were additional efforts ongoing to further increase the sp3
character of the scaffold16,17 which ultimately delivered the
preferred candidate. These efforts will be reported in a separate
publication.
8.
9.
Lee, A. M.; Cruite, J.; Welch, M. J.; Sullivan, B; Oldstone, M. B.
A. Virology 2013, 442, 114.
Concomitantly with the SAR efforts based on functional
readouts, the team was conducting target identification activities
for this series of TLR7/8 antagonists. It was hypothesized that
selective inhibition of TLR7 and TLR8 over TLR9 would involve
a signaling node unique to these TLRs. Alternatively, the
inhibition could be mediated through binding to the TLRs
themselves. To clarify this point, chimeric TLR domain swap
mutants were expressed in HEK29318 cells where the TLR7
extracellular domain (ECD) was fused to the TLR9 intracellular
domain (ICD) and vice versa. The cells were then stimulated with
either ssRNA40 for the TLR7 ECD or R848 for the TLR9 ECD
and NFκB was read out. Table 4 shows the activity of 13 and 14
on these mutants.
10. Hemmi, H.; Kaishi, T.; Takeuchi, O.; Sato, S.; Sanjo, H.; Hoshino,
K.; Horiuchi, T.; Tomizawa, H.; Takeda, K.; Akira, S. Nat.
Immunol. 2002, 3(2), 196.
11. Hoshino, K.; Takeuchi, O.; Kawai, T.; Sanjo, H.; Ogawa, T.;
Takeda, Y.; Takeda, K.; Akiro, S. J. Immunol. 1999, 162, 3749.
12. WO2007/134828, 2007.
13. Barrat, F. J. J. Exp. Med. 2018, 215, 2964–2966
14. Chanput, W.; Mes, J. J.; Wichers, H. J. Int. Immunopharm. 2014,
23, 37.
15. Dalvie, D.; Di, L. Pharmocol. and Therapeutics 2019, 201, 137.
16. Lovering, F.; Bikker, J.; Humblet, C. J. Med. Chem. 2009, 52, 6752.
17. Lovering, F. Med. Chem. Commun. 2013, 4, 515.
18. Thomas, P.; Smart, T. G. J Pharmocol. and Tox. Methods 2005, 51,
187.
Supplementary Material
Assay
13
14
Detailed synthetic procedures for all of the newly synthesized
compounds as well as biochemical and in vivo and ex vivo assay
procedures are provided in the Supplementary material.
TLR7 ECD, TLR9 ICD
TLR9 ECD, TLR7 ICD
< 0.001 µM
>5 µM
< 0.001 µM
1.5 µM
Table 4. TLR7/9 chimeric receptor activity
The fact that TLR7 potency tracked with the presence of the
ligand binding TLR7 ECD and not the chimera with the TLR9
Click here to remove instruction text...

Graphical Abstract
To create your abstract, type over the instructions in the template box below.
Fonts or abstract dimensions should not be changed or altered.
Discovery of potent, bioavailable in vivo
Leave this area blank for abstract info.
efficacious modulators of the TLR7/8 pathway
Alper, Phil B.*; Deane Jonathan*; Betschart, Claudia; Buffet, David; Collignon Zipfel, Géraldine; Gordon, Perry; Hampton, Janice;
Hawtin, Stuart; Ibanez, Maureen; Jiang, Tao; Junt, Tobias; Knoepfel, Thomas; Liu, Bo; Maginnis, Jillian; McKeever, Una; Michellys,
Pierre-Yves; Mutnick, Daniel; Nayak, Bishnu; Niwa, Satoru; Richmond, Wendy; Rush, James S.; Syka, Peter; Zhang, Yi; Zhu, Xuefeng
N
N
N
N
N
F
H2N
N
N
N
O
O
N
N
NH2
1
15
micromolar TLR7/8 inhibitor
with no PO exposure
nanomolar TLR7/8 inhibitor
with good PO exposure