Design and synthesis of tetrahydropyridopyrimidine based Toll-Like Receptor (TLR) 7/8 dual agonists

Article abstract of DOI:10.1016/j.bmcl.2018.08.015

In a continuing effort to discover novel TLR agonists, herein we report on the discovery and structure–activity relationship of novel tetrahydropyridopyrimidine TLR 7/8 agonists. Optimization of this series towards dual agonist activity and a high clearance profile resulted in the identification of compound 52a1. Evaluation in vivo revealed an interferon stimulated response (ISG) in mice with limited systemic exposure and demonstrated the potential in antiviral treatment or as a vaccine adjuvant.

Full text of DOI:10.1016/j.bmcl.2018.08.015

Accepted Manuscript
Design and Synthesis of Tetrahydropyridopyrimidine Based Toll-Like Receptor
(TLR) 7/8 Dual Agonists
David C. McGowan, Florence Herschke, Mourad D. Khamlichi, Mari Luz
Rosauro, Sara M. Pérez Benedicto, Frederik Pauwels, Bart Stoops, Vineet
Pande, Annick Scholliers, Bertrand Van Schoubroeck, Wendy Mostmans, Kris
Van Dijck, Tine Thoné, Helen Horton, Gregory Fanning, Tim H.M. Jonckers,
Pierre Raboisson
PII:
S0960-894X(18)30679-6
https://doi.org/10.1016/j.bmcl.2018.08.015
BMCL 25995
DOI:
Reference:
Bioorganic & Medicinal Chemistry Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
9 March 2018
29 July 2018
13 August 2018
Please cite this article as: McGowan, D.C., Herschke, F., Khamlichi, M.D., Luz Rosauro, M., Benedicto, M.P.,
Pauwels, F., Stoops, B., Pande, V., Scholliers, A., Van Schoubroeck, B., Mostmans, W., Van Dijck, K., Thoné, T.,
Horton, H., Fanning, G., Jonckers, T.H.M., Raboisson, P., Design and Synthesis of Tetrahydropyridopyrimidine
Based Toll-Like Receptor (TLR) 7/8 Dual Agonists, Bioorganic & Medicinal Chemistry Letters (2018), doi: https://
doi.org/10.1016/j.bmcl.2018.08.015
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. 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.

Design and Synthesis of Tetrahydropyridopyrimidine Based
Toll-Like Receptor (TLR) 7/8 Dual Agonists
David C. McGowan,a,* Florence Herschke,a Mourad D. Khamlichi,b Mari Luz Rosauro,b Sara M. Pérez
Benedicto,b Frederik Pauwels,a Bart Stoops,a Vineet Pande,a Annick Scholliers,a Bertrand Van
Schoubroeck,a Wendy Mostmans,a Kris Van Dijck,a Tine Thoné,a Helen Horton,a Gregory Fanning,a Tim
H. M. Jonckers,a and Pierre Raboissona
aJanssen Infectious Diseases, Turnhoutseweg 30, 2340 Beerse, Belgium
bVillapharma Research, Parque Tecnológico de Fuente Álamo. Ctra. El Estrecho-Lobosillo, Km. 2,5- Av.
Azul 30320 Fuente Álamo de Murcia, Murcia, Spain
Abstract. In a continuing effort to discover novel TLR agonists, herein we report on the discovery and
structure-activity relationship of novel tetrahydropyridopyrimidine TLR 7/8 agonists. Optimization of this
series towards dual agonist activity and a high clearance profile resulted in the identification of compound
52a1. Evaluation in vivo revealed an interferon stimulated response (ISG) in mice with limited systemic
exposure and demonstrated the potential in antiviral treatment or as a vaccine adjuvant.
Key words: Immunomodulator, Toll-Like Receptor, TLR7/8, tetrahydropyridopyrimidine.
Toll-like receptors (TLRs) have the important role of recognizing molecular patterns that are present in
foreign pathogens and to subsequently activate an immune response.1 Single-stranded viral RNA is
recognized by TLR7 and 8, the agonism of which induces a robust TH1-type immune response. TLR 7 is
expressed in plasmacytoid dendritic cells (pDCs) in humans, inducing the production of endogenous IFN,
and on B cells, to a lesser extent, inducing the proliferation and secretion of antibodies. TLR 8 however, is
principally expressed in monocytes and myeloid dendritic cells (mDCs), and activation also results in a

strong TH1 response primarily due to secretion of IL-12.2,3,4 Therefore, TLR7/8 agonists show potential in
the treatment of viral infections and other indications, such as potential adjuvants for therapeutic vaccines.
5,6,7
The objective of this study was to identify an agonist with equal activity on both TLRs 7 and 8, based on
the data generated from HEK293 cells transfected with hTLR7 or 8, and be structurally different to
previously reported dual agonists. As a verification of cytokine production, a second in vitro assay was
employed on human peripheral blood mononuclear cells (hPBMCs) where activation of human TLR7 on
plasmacytoid dendritic cells (pDCs) effected the production of IFN and other cytokines, after which the
conditioned media was added to the HCV replicon system measuring antiviral activity. Data for both assays
are reported as least effect concentrations (LEC) defined as the concentration that induces an effect at least
two-fold above the standard deviation of the assay (see supplementary data). Finally, a high first pass effect
was targeted in vivo, to display a pharmacodynamic effect and limit risk factors associated with systemic
exposure of TLR agonists.8,9
NH2
N
NH2
H
N
H
N
O
N
N
O
N
O
N
O
N
N
NH
HCl
1
2
Figure 1
Structure of selected TLR agonists containing an aliphatic amine.
GS-9620 (1, Figure 1) is a selective TLR 7 agonist that has been studied in clinical trials for the treatment
of chronic HBV.10 It is a potent bicyclic agonist, with a pendent benzyl group further substituted by a
methylenepyrrolidine, and displays a high first pass effect. The hydroxypurine series are described as TLR7
selective agonists.11,12 An exception to this selective series was noted when the purine scaffold and pendant
piperidine were separated by one methylene unit (2), that afforded potency on both TLR7 and 8 (Figure
1).13 It was hypothesized that the position of the amine may be a factor influencing the selectivity for TLR7

vs 8. In the conceptual design of a new scaffold, targeting both TLR7 and 8, the pyrimidine core structure
from previously described TLR7/8 agonist series (3, 4)14,15 was combined with a piperidine ring fused across
the 5,6-bond of the pyrimidine scaffold to afford novel agonist 5 (Figure 2).15b The piperidine is thus
included into the bicyclic scaffold to provide a more rigid structure with greater control over the
conformation to empirically learn about TLR 7 vs. 8 selectivity following varied substitution.
NH2
N
NH2
N
N
H
N
N
R2
N
N
6
5
H
H
R1
X
X
1
5
, X = N-Pg or CH2
3
, R = OEt, R2 = H
1
New Design
4,
R , R2= CH3
Figure 2.
Novel series of tetrahydropyridopyrimidine TLR 7/8 agonists derived from described pyrimidine series.14,15
NH2
Ph
O
O
X
N
N
COOCH3
COOCH3
N
O
N
X
H
X
X
X = N-Pg or CH2
Figure 3.
Retrosynthetic analysis of novel TLR7/8 agonists. Pg; protecting group.
Tetrahydropiperidinopyrimidines are described as drug-like scaffolds concerning a variety of drug targets,
such as calcium-calmodulin protein kinase inhibitors,16 as phosphodiesterase 10A inhibitors for the
treatment of schizophrenia,17 and towards the potential treatment of cystic fibrosis.18 Retrosynthetically, the
tetrahydropyridopyrimidine scaffold can be derived from the corresponding oxo-piperidinecarboxylates,
which in turn could be formed by a Dieckman condensation from a protected aminodiester (Figure 3). The
synthesis pathway was analogous to methods described in the literature.18,19 Initial exploration in potency
and selectivity originated from a library of electrophiles on the piperidine nitrogen. Further modifications

to the carbons of the fused piperidine ring, if needed, could arise from alteration of the precursors or be
introduced using other methods.19b
NH2
N
NH2
N
NH2
N
O
N
N
N
COOEt
HCl
ii
i
iii
XB
HO
Cl
N
XA
H
XB
XB
XB
XA
6a,b
XA
7a,b
XA
A
A
, X = NBn, XB = CH2
, X = NBn, XA = CH2
, X = NBn, XB = CH2
, X = NBn, XA = CH2
, X = NH, XB = CH2
, X = NH, XA = CH2
5a
5b
8a
8b
B
B
iv
iv
A
9a
9b
B
NH2
NBoc2
NBoc2
N
N
N
N
N
N
iv
v
N
8a,b
N
Boc
N
H
ZB
Boc
YB
XB
ZA
YA
XA
A
A
, Y = NH, YB = CH2
, Z = NCH3, YB = CH2
A
, X = NBn, XB = CH2
11a
11b
12a
10a
B
B
, Y = NH, YA = CH2
, Z = NCH3, YA = CH2
B
, X = NBn, XA = CH2
vi, ix
12b
13a
13b
14a
14b
15a
15b
16a
16b
10b
A
, Z = N(CH2)2Ph, YB = CH2
vi, ix
B
, Z = N(CH2)2Ph, YA = CH2
NH2
A
, Z = NAc, YB = CH2
B
, Z = NAc, YA = CH2
vii, ix
viii, ix
N
N
N
A
, Z = NBz, YB = CH2
N
B
, Z = NBz, YA = CH2
H
vii, ix
A
, Z = NMs, YB = CH2
B
, Z = NMs, YA = CH2
17-33
(Table 2)
R
NH2
N
N
N
H
N
34-47
O
(Table 3)
R
Scheme 1. Synthesis of TLR agonists. Reagents and conditions: (i) guanidine carbonate, EtOH, reflux, 18h (ii) POCl3,
100ºC, 4h (iii) Al2O3, 1,4-dioxane, n-butylamine 120ºC, 18h (iv) H2, 10% Pd/C, CH3OH, rt, 2h (v) Boc2O, DMAP,
THF, 80 ºC, 4h (vi) RCHO, NaHB(OAc)3, DCE, rt, 18h (vii) RCOCl, DMAP, Et3N, rt, 16h (viii) MsCl, DMAP, Et3N,
0ºC to rt, 18h (ix) 4M HCl in dioxane, rt, 18h.
Regardless of the electrophile substitution on the nitrogen of regioisomers A and B, agonists 8-16 showed
a varied degree of TLR8 selectivity (Table 1) and less potency on TLR7 in comparison with in-house data
on Resiquimod (TLR7 LEC = 0.1 µM, TLR8 LEC = 0.3 µM). The most potent analogs were indeed the

more basic benzyl, methyl and phenethyl analogs (8a, 12a, 13a, 13b respectively). Although these agonists
provided intriguing potency, the desired 1:1 ratio of TLR7:8 agonism was not observed. Furthermore,
concern was raised about agonists containing a basic piperidine moiety, a potential cationic center that may
have an impact on physicochemical properties and pharmacokinetics, in particular, an increase in the
volume of distribution, and the possible risk of phospholipidosis.20,21 This could result in a longer terminal
half-life, potentially risking toxicity associated with systemic cytokine activation.8 The greatest differences
in TLR8 activity between regioisomers A and B were the unsubstituted scaffold (9a vs 9b) and the
methanesulfonamide (16a vs 16b). Finally, the benzamide substituent 15a of the regioisomer A subseries
gave the desired equal ratio of TLR7 to 8 activities, while its congener 15b, in the regioisomer B subseries,
was five-fold more selective for TLR8. Thus, exploration continued solely on the regioisomer A subseries.
Table 1
Electrophile scan on regioisomers A and B
NH2
N
NH2
N
N
N
N
N
B
A
N
R
H
H
N
R
LECa
hTLR7
(µM)
1.6
11
>25
8.9
9.6
10
1.8
6.5
4.9
5.8
1.5
3.9
>25
20
LECa
hTLR8
(µM)
0.06
0.4
7.2
0.6
0.03
0.2
0.2
0.2
0.5
0.4
1.4
0.8
9.9
0.8
Regio-
isomer
Entry*
R
8a
8b
9a
9b
Bn
Bn
H
H
CH3
CH3
A
B
A
B
A
B
A
B
A
B
A
B
A
B
12a
12b
13a
13b
14a
14b
15a
15b
16a
16b
Ph(CH2)2-
Ph(CH2)2-
Ac
Ac
Bz
Bz
Ms
Ms
*All compounds had CC50 > 24 µM
aLEC; least effective concentration (see supplementary data)

Before initiating further exploration into benzamide analogs of 15a in the regioisomer A subseries, benzyl
substituents were briefly examined because of the numerous possibilities for variation from readily
available reagents. Indeed, with benzyl substituents, one can influence the conformation of the ring in
space by groups on the 2-phenyl position, and influence the electronics by changing the ring substituents.
The library approach provided facile identification of possible protein interactions, via substituent effects,
or -stacking, that could affect TLR potency and selectivity, or could reveal differences in
physicochemical properties.
The synthesis of a library of benzylamine products on regioisomer A, started from Boc-protected
intermediate 10a (Scheme 1). The variable benzyl group was installed via reductive alkylation employing
the corresponding aldehyde in DCE, using sodium triacetoxyborohydride as a reducing agent, to afford
agonists 17-23 (Scheme 1). Overall, the activity of the benzylamine regioisomer A subseries had greater
selectivity for TLR8 (Table 2). Compounds containing a substituent at the 2-position of the benzene ring
(17-20) showed the best potency on this subseries but were not able to influence the selectivity vs. that of
the unsubstituted benzene ring (8a). Substitution at the benzylic-4-position (e.g. 22) was deleterious to
activity on TLR7, especially with larger groups (23-26), but left little effect on TLR 8 agonist potential.
The crowded 1-napthyl analog (27) shared the same selectivity as 8a. Heterocyclic ring analogs (28-33)
also led to decreased activity versus 8a, and showed no compounds having equal ratio of TLR7 vs 8
agonism. Attention thus returned to forming a library of amides on regioisomer A to identify additional
agonists with equal potency on TLR7 vs 8.
Table 2
Activity of the benzyl substituents on regioisomer A
NH2
N
N
N
H
A
N
R
LECa
hTLR7
Entry*
R
LECa
LECa

(µM)
hTLR8
(µM)
0.05
0.04
0.18
0.14
0.23
0.08
0.24
0.23
0.66
0.50
0.43
0.07
0.13
0.29
0.14
0.08
0.22
hPBMC
(µM)
0.15
0.05
0.25
0.19
0.53
0.17
0.52
0.54
2.79
4.49
0.43
0.14
0.15
0.50
0.16
0.07
0.18
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
2-chlorobenzyl
1.5
1.0
1.5
2.4
3.5
2.1
4.7
2.8
19
2-chloro-3-methoxybenzyl
2,3-dichlorobenzyl
2,6-dichlorobenzyl
3,5-difluorobenzyl
4-methoxybenzyl
4-(trifluoromethoxy)benzyl
4-trifluoromethylbenzyl
4-t-butylbenzyl
4-phenyl-benzyl
1-napthyl-CH2-
2-pyridyl-CH2-
3-pyridyl-CH2-
4-pyridyl-CH2-
14
1.3
3.3
3.0
5.2
2.5
3.7
>20
2-thiophen-CH2-
2-furyl-CH2-
3-indole-CH2-
*All compounds had CC50 >24 µM
aLEC; least effective concentration (see supplementary data).
The library of amide analogs on regioisomer A (34-47) were prepared by reacting the variable acid
chloride with 10a under basic conditions in dichloromethane with DMAP at ambient temperature
(Scheme 1, Table 3). Alkyl amides (34-36) proved active but TLR8 selective. The 2 or 4-pyridyl analogs
(37, 38) led to less active agonists, and the same for the pyrazine congeners 39, and 40. The thiophene
amide 41 proved to be more potent than the isosteric benzamide (15a), had a 1:1 ratio of agonist activity,
and was the most potent of the series. The 4-thiazole derivative 42, showed desirable activity. However,
the addition of a 2-amino group to the heterocyclic amide, as in 43, led to a significant loss in potency on
both TLR7 and 8. Activity was regained in the 2-methyl-4-thiazole analog 44. The opposite observation
in relative activity was made in the 5-thiazole amides (45 vs 46) where the amino-substituent proved more
active than the methylthiazole analog. Heterocyclic amides 47-49 displayed activity above 1 µM, and
were found to be nearly equipotent on the PBMC assay. In summary, 41 and 44 had the desired 1:1 ratio
of agonist activity but showed insufficient selectivity over the hERG antitarget (hERG IC50 = 0.9 and 1.3
M respectively).22 While low plasma concentrations of 41 and 44 (99 and 88 percent metabolized in
mouse liver microsomes after 15 minutes at 1M concentration, respectively) were expected after oral

administration, however, an effort was made to remove this off-target activity. Further derivatives were
made to address the potential cardiovascular liability of 41 and 44.
Table 3
Activity of the aromatic amide series on regioisomer A.
NH2
N
N
N
H
A
N
R
O
LECa
hTLR7
(µM)
2.3
8.1
0.8
9.5
>25
3.7
11
0.2
1.2
8.0
2.0
2.1
13
0.9
0.7
0.8
LECa
hTLR8
(µM)
0.4
1.6
0.4
2.9
>25
2.2
6.8
0.4
0.5
6.9
1.8
4.4
5.6
0.1
0.2
0.4
LECa
hPBMC
(µM)
0.53
1.80
0.23
2.21
3.56
2.38
1.07
0.07
0.17
1.05
0.53
0.38
2.23
0.15
0.17
0.22
Entry*
R
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Et
Ipr
cyclobutyl
2-pyridyl
4-pyridyl
2-pyrazinyl
2-pyrazinyl-5-methyl
2-thiophene
4-thiazole
2-amino-4-thiazole
2-methyl-4-thiazole
2-amino-5-thiazole
4-methyl-5-thiazole
1,2,3-thiadiazole
4-oxazole
3-methylisoxazole
aLEC; least effective concentration (see supplementary data). All compounds had CC50 >24 µM
It was previously described in a related series of TLR agonists that the presence of a branched (S)-
configured aminoalcohol aided in the reduction of off-target activity.14 Thus, the same strategy was
applied to this scaffold. The favored 2-thiophene (41), and 2-methyl-4-thiazole (44) amides were held
constant in the next phase of exploration where aminoalcohol derivatives were formed to explore their
potential to reduce hERG binding.

NH2
N
NH2
N
NH2
N
O
N
HO
N
N
COOEt
HCl
(S)
iii, iv
i, ii
v
Cl
N
HO
A
H
n
N
N
R
N
N
H
Bn
O
R
O
52a1 : R = 2-methyl-4-thiazole; n = 1
52a2 : R = 2-methyl-4-thiazole; n = 2
52b1 : R = 2-thiophene; n = 1
5a
50
51a : R = 2-methyl-4-thiazole
51b : R = 2-thiophene
52b2 : R = 2-thiophene; n = 2
Scheme 2. Synthesis of TLR agonists with an aminoalcohol variation. Reagents and conditions: (i) guanidine
carbonate, EtOH, reflux, 18h, (ii) H2, 10%Pd/C, methanol, rt, 2h (iii) R-COCl, Et3N, DMAP, CH2Cl2, rt, 16h (iv)
POCl3, 100ºC, 4h (v) 3-aminohexanol or 3-aminoheptanol, IPA, 120ºC, 18h.
Akin to the chemistry described in scheme 1, 5a was converted to 50 in two steps (Scheme 2). The
amines were acylated with the preferred acid chlorides, and the 4-chloro species 51a and 51b were
afforded after treatment with phosphorousoxychloride at reflux. 51a was then reacted with 3-
aminohexanol or 3-aminoheptanol to afford 52a1 and 52a2, respectively. Likewise, 51b was also reacted
with the two -aminoalcohols to afford 52b1 and 52b2. The resulting products had acceptable agonist
potential and showed a clear reduction in hERG potential. The aminoalcohol modification in amide
products 52a1, and 52a2, resulted in over 20-fold reduction in hERG binding compared to the n-
butylamine congener 44. Analogs containing the larger (S)-3-aminoheptanol (52a2, 52b2) trended toward
TLR8 selectivity when compared to the (S)-3-aminohexanol congeners (52a1, 52b1). Noteworthy, all
four compounds shared the same potency in the PBMC assay (Table 4). In summary, 52a1 represented
nearly equal agonism on both TLR7 and 8, with a potentially reduced cardiovascular risk. To validate
these in vitro results, compound 52a1 was examined in vivo for its pharmacokinetic and
pharmacodynamic properties. 52a1 was administered orally to healthy C57Bl/6 mice at a dose of 5mg/kg.
52a1 was found to be readily absorbed (Tmax = 0.5h) and rapidly cleared, in line with our target product
profile, resulting in low plasma concentrations but considerably higher compound concentration in the
liver. Despite the low Cmax seen in mice, pharmacodynamic results showed a response in interferon γ-
induced protein (IP-10), in plasma and in the liver, where the Tmax was measured at 4h. This data suggests
that compound 52a1 induced an antiviral innate immune response. This is reassuring regarding the safety

of 52a1 in that at low dose antiviral cytokines could be observed without the negative aspects of systemic
IFN.23
Table 4
Activity of selected aromatic amides with an amino alcohol modification
NH2
HO
N
N
(S)
n
N
H
A
N
R
O
LECa
hTLR7
(µM)
1.5
1.0
0.2
LECa
hTLR8
(µM)
3.4
6.6
1.1
LECa
hPBMC
(µM)
0.2
LMb
(m,h)
hERG IC50
(µM)
Entry
n
R
52a1
52a2
52b1
52b2
1
2
1
2
2-methyl-4-thiazole
2-methyl-4-thiazole
2-thiophene
22
>50
7.6
6.9
33, 21
71, 30
73, 22
90, 37
0.2
0.2
0.2
2-thiophene
0.2
1.7
aLEC; least effective concentration (see supplementary data). All compounds had CC50 >24 µM
bLM; liver microsome stability (percent turnover at 1 M, 15 min.). m; mouse. h; human.
Table 5
Mouse pharmacokinetic and pharmacodynamic parameters after oral administration of 5 mg/kg.
AUC 0−last
(ng·h/mL)
Tmax
Cmax
mIFNα
IP-10
(h)
Entry
Plasma
(ng/mL)
Liver
(ng/g)
Plasma
Liver
Plasma Cmax Liver Cmax Plasma Cmax Liver Cmax
(pg/mL)
Tmax
(h)
Plasma
(ng/mL) (ng/g)
(pg/g)
(pg/mL)
(pg/g)
29
711
20
± 7
75
± 21
52a1
0.5
70
1410
0
0
4
± 9.0 ± 657
a

b
Figure 4. Compound 52a1 modeled in the Resiquimod binding pocket on (a) monkey TLR7 dimer interface (PDB
ID:5GMH) & (b) human TLR8 dimer interface (PDB ID: 3W3N). The two monomers are colored orange and green,
respectively. The surface representation is on the left, labelled key interactions in the middle, and an overlay with
Resiquimod on the right. The carbon atoms of 52a1 are colored in cyan, while the carbon atoms of Resiquimod are in
gray.
Figure 4 shows potential binding modes of compound 52a1 in the TLR7 and TLR8 dimer interfaces.
Analogous to Resiquimod, 52a1 forms hydrogen-bonding interactions with aspartate residues (ASP555 in
TLR7 and ASP543 in TLR8) and backbone atoms (THR586 in TLR7 and THR574 in TLR8), in both TLR7
and 8. Additionally, the pendant alkyl chain of 52a1, makes a tight fit into the deep hydrophobic pocket,
composed of residues including mainly valine, phenylalanine and tyrosine at the dimer interface in both
TLR7 and TLR8. The alcohol chain, while having van der waals interactions, remains solvent exposed at
the distal end, in both cases. Despite a good ligand based overlay with Resiquimod, there is an uncertainty
in the orientation of the 2-methyl-4-thiazole amide substituent, as well as corresponding receptor residues
around this moiety, due to observed clashes in the receptor crystal structures bound to directly overlaid
conformations of 52a1. Although in the current model, such clashes are relieved by selecting different
lower-energy rotamers of tyrosine residues (Figure 4; TYR356 in TLR7 and TYR353 in TLR8) stacked
with the thiazole moiety, it is possible that the conformations of the receptors are different, especially in
the loops containing these tyrosine residues or some other form of conformational plasticity in this region,
for instance, in the relative orientation of the TLR monomers. Overall, this model of 52a1 in complex with
the receptor, provides a partial structural basis for the dual agonistic nature, as it shows how this compound
engages conserved residues in both TLR7 and 8.

In conclusion, a series of dual TLR7/8 agonists were described, with many compounds showing inherent
TLR8 selectivity. Among these, 52a1, of the amide subseries on regioisomer A, was identified as a novel
and potent dual agonist. The pharmacokinetic profile of 52a1 in mice confirmed the low systemic exposure,
and target engagement was demonstrated by the induction of IP-10. In addition, 52a1 showed low plasma
protein binding (30% bound in mice plasma) no inhibition of major CYP450 isozymes (CYP450 >10 M:
3A4, 2C8, 2C9, 2D6, 1A2, 2C19) and lacked off-target activity across a multireceptor panel (>10 M
against histamine, dopamine, and serotonin subtypes). This series warrants further exploration as potential
immunomodulating agents.
Acknowledgements
The authors would like to thank Lieve Dillen, Ludo Quirynen, and Petra Gysemberg of the bioanalysis
group at Janssen.
Supplementary data
Supplementary data associated with this article can be found, in the online version, at ...
References
1. Kawai T, Akira S. TLR signaling. Semin. Immunol. 2007; 19: 24-32. DOI:
10.1016/j.smim.2006.12.004 (b) Akira S, Takeda K, Kaisho T. Toll-like receptors: critical proteins
linking innate and acquired immunity. Nat. Imm. Rev. 2001; 2: 675-680. doi:10.1038/90609
2. Pasare C, Medzhitov R. Toll-like receptors: linking innate and adaptive immunity. Microbes and
Infection 2004; 6: 1382–1387. DOI: 10.1016/j.micinf.2004.08.018
3. Diebold S, Kaisho T, Hemmi H, Akira S, Reis e Sousa C. Innate antiviral responses by means of TLR7-
mediated recognition of single-stranded RNA. Science 2004; 303: 1529-1531. DOI:
10.1126/science.1093616.
4. Lund JM, Alexopoulou L, Sato A, Karow M, Adams NC, Gale NW, Iwasaki A, Flavell RA.
Recognition of single-stranded RNA viruses by Toll-like receptor 7. Proc. Natl. Acad. Sci. U. S. A.
2004; 101: 5598–5603. doi: 10.1073/pnas.0400937101.
5. Schurich A, Pallett LJ, Lubowiecki M, Singh HD, Gill US, Kennedy PT, Nastouli E, Tanwar S,
Rosenberg, W, Maini M. The third signal cytokine IL-12 rescues the anti-viral function of exhausted
HBV-specific CD8 T cells. PLoS Pathog. 2013; 9: e1003208. doi: 10.1371/journal.ppat.1003208
6. Hammerbeck DM, Burleson GR, Schuller CJ, Vasilakos JP, Tomai M, Egging E, Cochran FR, Woulfe
S, Miller RL. Administration of a dual toll-like receptor 7 and toll-like receptor 8 agonist protects
against influenza in rats. Antiviral Res. 2007; 73: 1-11. Doi: 10.1016/j.antiviral.2006.07.011
7. Chang J, Guo JT. Treatment of chronic hepatitis B with pattern recognition receptor agonists: Current
status and potential for a cure. Antiviral Res. 2015; 121:152-159. Doi: 10.1016/j.antiviral.2015.07.006

8. Pockros PJ, Guyader D, Patton H, Tong MJ, Wright T, McHutchison JG, Meng T-C. Oral
Resiquimod in chronic HCV infection: safety and efficacy in 2 placebo-controlled, double-blind
phase IIa studies. J. Hepatol. 2007; 47: 174–182. doi.org/10.1016/j.jhep.2007.02.025
9. Delaney S, Biffen M, Maltby J, et al. Tolerability in man following inhalation dosing of the selective
TLR7 agonist, AZD8848. BMJ Open. Resp. Res. 2016; 3: e000113. doi:10.1136/bmjresp 2015-
000113.
10. Roethle PA, McFadden RM, Yang H, Hrvatin P, Hui H, Graupe M, Gallagher B, Chao J,
Hesselgesser J, Duatschek P, Zheng J, Lu B, Tumas DB, Perry J, Halcomb RL. Identification and
optimization of pteridinone toll-like receptor 7 (TLR7) agonists for the oral treatment of viral
hepatitis. J. Med. Chem. 2013; 56: 7324-7333. doi: 10.1021/jm400815m.
11. Biggadike K, Ahmed M, Ball D, Coe D, Wilk DA, Edwards C, Gibbon RH, Hardy CJ, Hermitage
SA, Hessey JO, Hillegas AE, Hughes SC, Lazarides L, Lewell, XQ, Lucas A, Mallett DN, Price MA,
Priest FM, Quint DJ, Shah P, Sitaram A, Smith SA, Stocker R, Trivedi NA, Tsitoura DC, Weller V.
Discovery of 6-Amino-2-{[(1S)-1-methylbutyl]oxy}-9-[5-(1-piperidinyl)pentyl]-7,9-dihydro-8H-
purin-8-one (GSK2245035), a highly potent and selective intranasal toll-like receptor 7 agonist for
the treatment of asthma. J. Med. Chem. 2016; 59: 1711–1726. DOI: 10.1021/acs.jmedchem.5b01647.
12. Isobe Y, Kurimoto A, Tobe M, Hashimoto K, Nakamura T, Norimura K, Ogita H, Takaku H.
Synthesis and biological evaluation of novel 9-substituted-8-hydroxyadenine derivatives as potent
interferon inducers. J. Med. Chem. 2006; 49: 2088-2095. DOI: 10.1021/jm051089s.
13. Bazin HG, Li Y, Khalaf JK, Mwakwari S, Livesay MT, Evans JT, Johnson DA. Structural
requirements for TLR7-selective signaling by 9-(4-piperidinylalkyl)-8-oxoadenine derivatives. Bio.
Org. Med. Chem. Lett. 2015; 25: 1318-1323. doi: 10.1016/j.bmcl.2015.01.037.
14. McGowan DC, Herschke F, Pauwels F, Stoops B, Last S, Pieters S, Scholliers A, Thoné T, Van
Schoubroeck B, De Pooter D, Mostmans W, Khamlichi MD, Embrechts W, Dhuyvetter D, Smyej I,
Arnoult E, Demin S, Borghys H, Fanning F, Vlach J, Raboisson P. Novel pyrimidine toll-like receptor
7 and 8 dual agonists to treat hepatitis B virus. J. Med. Chem. 2016; 59: 7936–7949. doi:
10.1021/acs.jmedchem.6b00747.
15. (a) Beesu M, Salyer AC, Brush MJ, Trautman KL, Hill Justin K, David SA. Identification of high-
potency human TLR8 and dual TLR7/TLR8 agonists in pyrimidine-2,4-diamines. J. Med. Chem.
2017; 60: 2084-2098. doi: 10.1021/acs.jmedchem.6b01860. (b) See WO2013117615 and supplementary
data.
16. Asano S, Komiya M, Koike N, Koga E, Nakatani S, Isobe Y. 5,6,7,8-Tetrahydropyrido[4,3-
d]pyrimidines as novel class of potent and highly selective CaMKII inhibitors. Bio. Org. Med. Chem.
Lett. 2010; 20: 6696-6698. doi: 10.1016/j.bmcl.2010.09.005.
17. Raheem I, Breslin MJ, Fandozzi C, Fuerst J, Hill N, Huszar S, Kandebo M, Kim SH, Ma B,
McGaughey G, Renger JJ, Schreier JD, Sharma S, Smith S, Uslaner J, Yan Y Coleman PJ, Cox CD.
Discovery of tetrahydropyridopyrimidine phosphodiesterase 10A inhibitors for the treatment of
schizophrenia. Bio. Org. Med. Chem. Lett. 2012; 22: 5903-5908. doi.org/10.1016/j.bmcl.2012.07.072
18. Pesci E, Bettinetti L, Fanti P, Galietta LJ, La Rosa S, Magnoni L, Pedemonte N, Sardone GL,
Maccari L. Novel hits in the correction of ΔF508-cystic fibrosis transmembrane conductance
regulator (CFTR) protein: synthesis, pharmacological, and ADME evaluation of
tetrahydropyrido[4,3-d]pyrimidines for the potential treatment of cystic fibrosis. J. Med. Chem. 2015;
58: 9697–9711. DOI: 10.1021/acs.jmedchem.5b00771
19. (a) Connolly T, Matchett M, Sarma K. Process development and scale-up of a selective α1-
adrenoceptor antagonist. Org. Proc. Res. & Dev. 2005; 9: 80-87. DOI: 10.1021/op0498114. (b) Beng,
TK, Gawley RE. Catalytic dynamic resolution applied to the synthesis of 2,6-disubstituted
piperidines: preparation of (+)-lupetidine and (-)-epidihydropinidine. Heterocycles, 84(2), 697-718;
2012
20. Avdeef A. Physicochemical profiling (solubility, permeability and charge state). Curr. Top. Med.
Chem. 2001; 1: 277-351.

21. Lamphier M, Zheng W, Latz E, Spyvee M, Hansen H, Rose J, Genest M, Yang H, Shaffer C, Zhao Y,
Shen Y, Liu C, Liu D, Mempel TR, Rowbottom C, Chow J, Twine NC, Yu M, Gusovsky F, Ishizaka
ST. Novel small molecule inhibitors of TLR7 and TLR9: mechanism of action and efficacy in vivo.
Mol. Pharmacol. 2014; 85: 429-440.
22. hERG potassium ion channel 3H-dofetilide binding in vitro.
23. Marcellin P, Lau GK, Zeuzem S, Heathcote EJ, Pockros PJ, Reddy KR, et al. Comparing the safety,
tolerability and quality of life in patients with chronic hepatitis B vs chronic hepatitis C treated with
peginterferon alpha-2a. Liver Int. 2008; 28: 477-485. doi: 10.1111/j.1478-3231.2008.01696. x.
Highlights
 A new series of TLR 7/8 dual agonists based on tetrahydropyridopyrimidines.
 A desired high first pass effect was observed to limit systemic exposure and cytokine
activation.
 Target engagement was demonstrated by the induction of IP-10 after oral administration
in mice.
24.