8

Subscriber access provided by Kaohsiung Medical University

Article

2,4-Diaminoquinazolines as Dual Toll Like Receptor (TLR) 7/8 Modulators for the Treatment of Hepatitis B Virus Werner Embrechts, Florence Herschke, Frederik Pauwels, Bart Stoops, Stefaan Last, Serge Pieters, Vineet Pande, Geert Pille, Katie Amssoms, Ilham Smyej, Deborah Dhuyvetter, Annick Scholliers, Wendy Mostmans, Kris Van Dijck, Bertrand Van Schoubroeck, Tine Thoné, Dorien De Pooter, Gregory Fanning, Tim H.M. Jonckers, Helen Horton, Pierre Jean Marie Bernard Raboisson, and David Craig McGowan J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b00643 • Publication Date (Web): 02 Jul 2018 Downloaded from http://pubs.acs.org on July 2, 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 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Medicinal Chemistry

2,4-Diaminoquinazolines as Dual Toll Like Receptor (TLR) 7/8 Modulators for the Treatment of Hepatitis B Virus Werner Embrechts,†* Florence Herschke,†Frederik Pauwels,†Bart Stoops,†Stefaan Last,† Serge Pieters,†Vineet Pande,† Geert Pille,† Katie Amssoms,† Ilham Smyej,† Deborah Dhuyvetter,†Annick

Scholliers,†Wendy

Mostmans,†Kris

Van

Dijck,†Bertrand

Van

Schoubroeck,†Tine Thone,†Dorien De Pooter,†Gregory Fanning,† Tim H. M. Jonckers,† Helen Horton,† Pierre Raboisson,† David McGowan† †

Janssen Infectious Diseases Diagnostics BVBA, Turnhoutseweg 30, 2340 Beerse, Belgium

ABSTRACT A novel series of 2,4-diaminoquinazolines was identified as potent dual Toll-like receptor (TLR) 7 and 8 agonists with reduced off-target activity. The stereochemistry of the aminoalcohol was found to influence the TLR7/8 selectivity with the (R) isomer resulting in selective TLR 8 agonism. Lead optimization toward a dual agonist afforded (S)-3-((2-amino-8-fluoroquinazolin4-yl) amino) hexanol 31 as a potent analog, being structurally different to previously described dual agonists.3 Pharmacokinetic and pharmacodynamic (PK/PD) studies revealed the desired high first pass profile aimed at limiting systemic cytokine activation. In vivo pharmacodynamic studies with lead compound 31 demonstrated production of cytokines consistent with TLR7/8 activation in mice and cynomolgus monkeys, and ex vivo inhibition of Hepatitis B virus (HBV).

1 ACS Paragon Plus Environment

Journal of Medicinal Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 43

INTRODUCTION

It is estimated that 250 million people worldwide are chronically infected with the hepatitis B virus (HBV).1 Patients with chronic hepatitis B are at increased risk of disease progression to cirrhosis, hepatic decompensation, and hepatocellular carcinoma (HCC).2 The treatment of chronic hepatitis B continues to be a challenge for physicians due to the limited efficacy of available therapy. Currently, there are two HBV treatment strategies, immunomodulator therapies such as interferon α (IFNα); and oral nucleoside analog therapy.3,4 IFNα and its derivatives have the possibility of obtaining, albeit at low rates, functional cure with a finite treatment in (HBsAg loss with or without seroconversion) leading to long-term clinical benefits in patients without or with well-compensated liver cirrhosis, though with substantial side effects. Nucleoside analog therapy is efficient at inhibiting HBV viral replication, therefore partially reducing the risk of progression to liver cancer, but life-long treatment is needed. However, HBsAg loss rates, leading to long-term clinical benefits, are extremely low.3,4 Thus, a clear need exists for alternative treatment approaches. It is believed that the low rates of functional cure are due to high viral antigen loads, inability of the immune system to control the virus and immunosuppressive environment of the liver.5,6 The significant induction of an HBVspecific CD8+ T cell response and breadth of epitope recognition, functional in the liver, is key in controlling HBV infection.7,8 However, most HBV-specific CD8+ T cells are exhausted in chronic patients, unable to proliferate and/or display their full cytotoxic activity.


Page 3 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


In humans, TLR7 is mainly expressed in plasmacytoid dendritic cells (pDC), inducing their activation and production of IFNα, and to a lower extent on B cells, inducing their proliferation and secretion of antibodies. Conversely, TLR8 is mainly expressed in monocytes and myeloid DCs (mDCs), inducing their activation toward a strong TH1 profile.9-11 TLR7/8 agonists strongly induced the expression of IL-12, CD40 and OX40L, which are to date among the very few immunostimulatory agents able to reinvigorate exhausted HBV-specific T cells ex vivo.12-14 TLR8 agonists were also the most efficient at inducing IFNγ from human liver nonparenchymal cells.15 Therefore, TLR 7 e.g. GS-9620 (1, Figure 1) and 8 agonists, have received attention as a potentially new treatment of chronic hepatitis B virus.16-19 We previously reported on the identification of compound 2 as a dual TLR7/8 agonist.3

Figure 1. Selected compounds reported as TLR7 agonists.

RESULTS AND DISCUSSION

Medicinal chemistry strategy. In continuation of our work, we identified a series of quinazoline based agonists which proved to be an order of magnitude more potent compared to 2 and retained dual agonist activity (Figure 2).20 As done previously, we opted to retain compounds with minimal off target activity (e.g. hERG inhibition, CYP inhibition). 3 ACS Paragon Plus Environment


Page 4 of 43

First, amine variation was explored on the unsubstituted scaffold (compounds 4-14, Table 2). Two aminoalcohol analogs (12, 13) displayed high potency and dual agonist activity. A second subseries focused on holding the (S)-3-aminohexanol and (S)-3-aminoheptanol motif constant while exploring variation on the fused benzene ring (compounds 15-32, Table 3). Early ADME and safety profiling led to the identification of 31, a potent dual agonist which was selected for evaluation in vivo.

In contrast to a traditional medicinal chemistry program, rapidly metabolized compounds were desired, limiting compound exposure and systemic cytokine production. In addition, there is literature evidence indicating that an anti-viral response can be achieved in non-human primates with less than daily dosing of a TLR7 agonist.21

Figure 2. Reported agonist (2) and new quinazoline TLR7/8 agonist scaffold.


Page 5 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Chemsistry. Novel quinazoline derivatives 4-14 were obtained via coupling of 2-amino-4hydroxyquinazoline

(3a)

with

various

amines

using

(benzotriazol-1-

yloxy)tris(dimethylamino)phosphonium hexafluorophosphate (BOP) with DBU in anhydrous DMF at ambient temperature (Scheme 1, Table 3).22 Scheme 1. Synthesis of 2,4-diaminoquinazolines.a

a

Reagents and conditions: (i) DBU, BOP, R-NH2, anhyd. DMF, rt, 16h.

Synthesis of the substituted 2-amino-4-hydroxyquinazolines was accomplished by the reaction of anthranilic acids or esters with cyanamide in concentrated HCl in ethanol at 100°C. Scheme 2 (3b-j, Table 1).20 Then, the γ-aminoalcohols were added to a solution of 3b-j in anhydrous DMF in combination with BOP and DBU at room temperature, to afford the second subseries 15-32 (Table 3).

Scheme 2. Synthesis of 2-amino-4-hydroxyquinazolines.a



Page 6 of 43

Table 1. 2-amino-4-hydroxyquinazolines. Entry

a

Reagents and conditions: (i) HCl, NH2CN, EtOH, 100°C, 16h, pressure vessel, 5-85% isolated yield (ii) DBU, BOP, R-NH2, anhyd. DMF, rt, 16h, 10-90% isolated yield.

3a 3b 3c 3d 3e 3f 3g 3h 3i 3j

R1 H CH3 F H H H H H H H

R2 H H H CH3 F OCH3 H H H H

R3 H H H H H H CH3 F OCH3 H

R4 H H H H H H H H H F

Structure-Activity Relationship. Compounds in this program were evaluated for activity and selectivity in a HEK cell based assay containing a wild type TLR7 or TLR8 expression vector. In addition, compounds were incubated with human peripheral blood mononuclear cells (hPBMCs), and the media was transferred to an HCV replicon system to measure antiviral activity resulting from IFN and cytokines induced. Further in vitro assessment on the more interesting compounds led to the identification of lead compound 31 that was further evaluated in vivo in mice, rat and cynomolgous monkeys. Table 2. Amine Variation on the Unsubstituted Quinazoline Scaffold. a

hTLR7 (LEC µM)

hTLR8 (LEC µM)

hPBMC (LEC µM)

0.16

0.06

0.03

5

3.90

0.19

2.30

6

0.46

0.01

0.04

7

0.24

0.02

0.04

Entry 4

R

n-butyl


Page 7 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


a

8

3.18

0.45

0.71

9

>25

0.57

2.36

10

22.5

0.05

0.16

11

1.34

0.08

0.14

12

0.2

0.11

0.02

13

0.08

0.08

0.02

14

1.83

0.69

0.55

All compounds showed CC50 > 24 µM in the HEK 293 TOX assay described in the supporting information. LEC

values were averaged when determined in two or more independent experiments.

The activity data of the amine variations (compounds 4-14) are reported in Table 2. The nbutylamine analogue 4, reported previously, showed potent dual agonism, however, hERG inhibition was observed (hERG IC50 = 5.62 µM).20 In the previously reported dual agonist series (compound 2), certain aminoalcohol derivatives offered lower off-target activity while maintaining a dual agonist profile.

3

Thus, several (S) and (R)-configured ß- and γ-

aminoalcohols were explored on the unsubstituted quinazoline scaffold. A comparison of the activity of the (S)-2-aminobutanol analog (5), with its homolog (S)-2-aminopentanol (6), revealed a correlation between the potency and length of the alkyl chain, where (S)-2aminohexanol derivative (7) was the most potent of the three on TLR7 and in the same potency 7 ACS Paragon Plus Environment


Page 8 of 43

range for TLR8. (S)-2-amino-4-methylpentanol derivative (8) in comparison with its congener (6) highlighted that the 4-methyl substitution was less tolerated. Overall, the (S)-ß-aminoalcohol homologous series exhibited greater selectivity for TLR8. A more pronounced selectivity for TLR8 was observed with the (R)-2-aminopentanol (9) and (R)-2-aminohexanol (10) analogues with the latter displaying exquisite potency and selectivity for TLR8. To the best of our knowledge, this sterochemical selectivity towards TLR8 has not been reported. Isoleucinol derivative 11 demonstrated that methyl group substitution at the three position was better tolerated compared to the 4-methyl substituted congener 8. Both (S)-3-aminohexanol (12) and (S)-3-aminoheptanol (13) analogues showed the desired dual agonist activity with exceptional potency.

However,

the

(S)-3-amino-5-methylhexanol

(14),

like

the

(S)-2-amino-4-

methylpentanol (8) congener, proved deleterious to activity. Despite the desired potency and dual agonist activity of 12 and 13, hERG inhibition was observed for both compounds (hERG IC50 = 6.8 and 2.7 µM, respectively). Although compounds 7-14 were highly metabolized (Supporting Information, Table S2), and in line with our target profile to limit systemic concentrations, we preferred to focus on derivatives without cardiovascular liability. Therefore, our attention shifted to the examination of small variations on the quinazoline scaffold, since larger substituents altered the selectivity profile away from dual agonism.20 In the next subseries, both (S)-3aminohexanol and (S)-3-aminoheptanol were held constant (Table 3). Table 3. Quinazoline Variation in Combination with (S)-γ-Aminoalcohols.a

Entry

R1

R2

R3

R4

n

hTLR7

hTLR8

hPBMC

(LEC µM)

(LEC µM)

(LEC µM)

MSb M

MSb H

hERG IC50 (µM)


Page 9 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


12 H H H H 1 0.20 0.11 0.38 72 23 6.8 13 H H H H 2 0.08 0.08 0.02 82 45 2.7 5.5 15 CH3 H H H 1 0.22 0.14 n.d. 88 44 16 CH3 H H H 2 0.14 0.06 0.02 82 67 0.6 17 F H H H 1 0.22 0.02 n.d. 47 62 8.9 18 F H H H 2 0.34 0.10 0.06 25 87 >10 19 H CH3 H H 1 0.54 0.45 0.39 97 21 7.2 20 H CH3 H H 2 0.44 0.25 0.14 99 24 5.4 21 H F H H 1 1.05 0.42 n.d. 47 22 9.3 22 H F H H 2 1.59 0.42 0.37 63 33 5.8 23 H OCH3 H H 1 11.8 2.97 n.d. n.d. n.d. n.d. 6.0 24 H OCH3 H H 2 1.85 2.81 0.88 97 25 25 8.1 H H CH3 H 1 0.06 0.05 0.39 87 44 26 H H CH3 H 2 0.03 0.06 0.02 98 70 1.8 27 H H F H 1 0.11 0.04 4.8 100 13 9.1 28 H H F H 2 0.04 0.03 0.04 91 39 4.9 1 >10 29 H H OCH3 H 0.20 0.13 0.5 65 35 30 H H OCH3 H 2 0.16 0.16 0.04 80 37 3.2 31 H H H F 1 0.15 0.16 0.25 93 43 >10 32 H H H F 2 0.11 0.16 0.05 98 76 4.47 a All compounds displayed CC50 > 24 µM in the HEK 293 TOX assay described in the supporting information. LEC values were averaged when determined in two or more independent experiments. See supplementary information for assay conditions. n.d.; not determined. b MS; Metabolic Stability, percent metabolized after 15 min. at 1µM in liver microsomes for Mice (M) and Human (H).

In general, 5-methylquinazolines (15, 16) were equipotent compared to the unsubstituted 12 and 13, and the same was observed for the 5-fluoroquinazolines 17 and 18, some showing higher potency on TLR8. A marked loss in agonist potential was observed for the 6-substituted quinazoline analogues 19 to 24. Activity was regained in the 7-substituted quinazoline series (2530) rivaling the potency of the unsubstituted 12, and 13. The potency of the 8-fluoroquinazolines 31 and 32 was in the same range as the unsubstituted congeners 12 and 13. Taking a closer look at the hERG binding showed that the nature and position of the substituent along with the aminoalcohol influenced the hERG inhibition and was not simply correlated to log P. Among the compounds made, 18, 29, and 31 displayed the lowest hERG binding affinities. All three analogs were found to be potent dual agonists, however the toxicity observed on Hep G2 cells 9 ACS Paragon Plus Environment


Page 10 of 43

(EC20 = 13.2 µM) for 18 and the off-target activity (Aa1A IC50 = 2.34 µM; D1 IC50 = 5.75 µM; 5HT2A IC50 = 1.78 µM) and CYP inhibition (CYP 2D6 IC50 = 0.6 µM) of 29 resulted in selection of 31 that displayed a better off-target profile (Aa1A, D1, 5HT2A IC50 > 10 µM; CYP 2D6 IC50 = 5.7 µM). The binding profile of 31 was tested at 10 µM against a panel of 80 receptors including G-protein-coupled receptors (GPCRs), nuclear receptors, voltage and ligand gated receptors, other receptors, transporters, and kinases showing no significant activity.

a)

b)


Page 11 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3. Compound 31 docked into the Resiquimod-binding pocket of (a) Monkey TLR7 dimer interface (PDB ID: 5GMH), (b) Human TLR8 dimer interface (PDB ID: 3W3N). The two protein monomers are colored orange and green, respectively with surface representation on the left and details of key interactions on the right, with the ligand carbon atoms colored in cyan.

Figure 3 shows potential binding modes of 31 in the TLR7 or TLR8 dimer interfaces. Analogous to Resiquimod (Supporting Information, Figure S1), 31 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 31, makes a tight fit into the deep hydrophobic pocket, composed of residues including mainly valine, phenylalanine and tyrosine (Figure 3) at the dimer interface in both TLR7 and TLR8, whereas the alcohol chain, while having Van der Waal’s interactions, remains solvent exposed in both cases. This atomic level model of 31 in complex with the receptors, provides a structural basis for the dual agonistic nature, as it shows how this compound engages conserved residues in both TLR7 and 8.

The solubility of 31 was found to be greater than 4 mg/mL in Fassif and Fessif buffers. Compound 31 showed good permeability and was found not to be a P-glycoprotein substrate. No CYP inhibition was observed across several isozymes, and no CYP3A4 induction was observed. At 1 µM concentration, 31 revealed no inhibition when tested in a panel of 230 kinases. Furthermore, 31 was found to be negative in an in vitro micronucleustest (MNT), in TK-6 cells. A moderate turnover was seen in liver microsomes in contrast to a higher turnover in whole cell hepatocytes, except for human hepatocytes. Plasma protein binding was low, ranging from 41 to



Page 12 of 43

66% bound (Table 4) across species. To validate these promising in vitro findings, 31 was further investigated in vivo. Table 4. In vitro Metabolic Stability and Plasma Protein Binding of 31.a Species

CLintb Liver Microsomes (µL/min/mg protein)

Mouse Rat Dog Cynomolgus monkey Human a

TOc Hepatocytes

100 24 21 16 8

Plasma Protein Binding (% bound)

99 96 77 98 35

60 53 41 54 66

Values were rounded to the nearest whole number, bIntrinsic clearance rate, cPercent turnover at 1µM after 60 min.

Pharmacokinetics. A pharmacokinetic study in male Sprague-Dawley rats investigated the presence of 31 in the portal vein as well as in the systemic circulation following p.o. administration of 10 mg/kg as a solution. Tmax in plasma and in liver occurred 0.5 h after dosing and the plasma Cmax in the systemic circulation was 16 ng/mL, and 285 ng/g in the liver. The liver to plasma ratio was found to be 16, and the AUC of 31 was 49 ng.h/mL. The extensive metabolism of the compound was confirmed by the hepatic extraction ratio of 0.74. Table 5. Mean Maximum Levels of Compound and Cytokines in the Plasma and Liver, Induced by 31 After Oral Dosing to C57Bl/6 Mice.a Dose (mg/kg) Cmax (ng/mL) c Cmax (ng/g) d AUC0-last (ng.h/mL) c AUC0-last (ng.h/g) d Conc. at 7h (ng/mL) c Conc. at 7h (ng/g) d IFNαmax (pg/mL) c IFNαmax (pg/g) d Max. IP10 mRNA expression b,c Max. IP10 mRNA expression b,d IL12p70 max (pg/mL) c IL12p70 max (pg/g) d IFNγ max (pg/mL) c

0 n.d. n.d. n.d. n.d. n.d. n.d. BQL 30 1 1 560 5 BQL

0.3 n.d. 32 n.d. 25 n.d. n.d. BQL 100 15 35 470 BQL BQL

1 9 188 n.d. 148 n.d. n.d. 180 215 845 775 380 BQL BQL

3 22 385 n.d. 362 n.d. n.d. 695 765 2485 845 330 BQL BQL

10 148 1417 141 2315 n.d. n.d. >3200 1365 2875 800 630 790 295

30 813 7613 679 7160 35.7 462 1930 955 1100 1180 820 2340 1010


Page 13 of 43 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


IFNγ max (pg/g) d BQL BQL BQL 150 430 700 IL1ß max (pg/mL) c 60 50 BQL 40 55 70 d IL1ß max (pg/g) BQL 9765 16445 22925 29525 56615 CXCL-1max (pg/mL) c 230 2180 2155 >10000 >10000 >10000 CXCL-1max (pg/g) d 25 7790 6755 35575 48485 >83333 IL6 max (pg/mL) c 505 480 1150 4970 >10000 >10000 d IL6 max (pg/g) 10 BQL 905 3225 3450 7555 TNFα max (pg/mL) c 70 BQL 335 1665 >10000 >10000 TNFα max (pg/g) d 15 1945 BQL 2865 8330 13600 IL10 max (pg/mL) c 120 345 485 1115 29450 5170 d IL10 max (pg/g) 2 730 535 1355 2265 5185 a Groups of 15 mice were orally dosed with 0.3, 1, 3, 10 or 30 mg/kg of 31 in 20% HP-ß-CD at 0.5 hours, blood was drawn to quantify compound. At 0.5, 1, 2, 4 or 7 hours, three mice of each group were sacrificed and blood was drawn to measure plasma and liver compound levels (by LC-MS), cytokines (IFNα by ELISA, IP10 by RT-qPCR and other cytokines by Multiplex). Values were rounded to the nearest whole number (PK parameters) or the nearest multiple of 5 (cytokines). BQL, below minimum quantifiable limit; bfold-change relative to vehicle. n.d.; not determined, cplasma, dliver.

Pharmacokinetic and Pharmacodynamic Studies in Mice and Cynomolgus Monkey. A mouse model was used to demonstrate the initial proof of concept to induce endogenous IFNα in vivo. Single oral administration of 0.3, 1, 3, 10 and 30 mg/kg doses of 31 were given to healthy C57BL/6J mice. The compound was found to be rapidly cleared. Both Cmax and AUC increased more than dose proportionally between 3 and 30 mg/kg (Table 5). Levels of IFNα peaked between one and two hours post administration in the plasma and at one hour in the liver. Increase in IFNα in plasma was seen from 1 to 3 mg/kg, but appeared to plateau with further increase in the dose to reach a possible saturation of endogenous IFNα production. Increase in IFNα in the liver was dose-proportional from 1 to 3 mg/kg, and less than dose-proportional from 3 to 30 mg/kg, suggesting a possible saturation of endogenous IFNα production as well (Table 5). Ip10 mRNA upregulation was also detectable from 0.3 mg/kg, in both plasma and liver, and seemed to reach a saturation level from 3 mg/kg. The production of other cytokines was also measured (Table 5). IL-12p70 was detectable in plasma from 30 mg/kg, in the liver at 10 mg/kg and IFNγ in plasma from 10 mg/kg, in the liver 3 mg/kg suggesting the induction of a TH1



Page 14 of 43

response at higher doses. Pro-inflammatory cytokines CXCL1, TNFα, IL-6 and IL-1β were measured from 0.3, 1 or 3 mg/kg, while the anti-inflammatory cytokine IL-10 was measured from 0.3 mg/kg. Overall, 31 could induce IFNα and pro-inflammatory cytokines at low dose, as well as TΗ1 and myeloid responses at higher dose. The compound was well tolerated without relevant change in body temperature. Only a mild drop in activity was observed at 1-2h postinfection at 30 mg/kg. Table 6. Mean Maximum Levels of Compound, Cytokines and ISGs in the Plasma Induced Following Oral Dosing of 31 in Cynomolgus Monkeya 0 Dose (mg/kg) 1 3 9 30 Cmax (ng/mL) n.d. 38 142 815 2070 AUC0-last (ng.h/mL) n.d. 75 383 4051 5202 Conc at 8h (ng/mL) n.d. 5.9 14 200 160 BQL >51200 3465 >51200 >51200 IFNαmax (pg/mL) IP10max (pg/mL) 18.5 1690 1490 2400 8195 IL12p70 max (pg/mL) 2.17 BQL BQL 3.8 10.1 IL-15 max (pg/mL) 20 20 60 75 85 TNFα max (pg/mL) 285 BQL 70 350 2990 IL6 max (pg/mL) 80 75 200 2265 5285 IL8 max (pg/mL) 430 n.d. 1505 1640 1840 IL1Ra max (pg/mL) 320 805 2020 2455 2970 MCP1 max (pg/mL) 385 2820 7720 8890 >10000 37 560 44 85 310 MIP-1β max (pg/mL) GCSF max (pg/mL) 85 BQL BQL 825 1135 IL10 max (pg/ml) 195 1275 1615 3785 9135 Isg15 mRNA max b 1 145 145 560 285 Mx1 mRNA max b,c 1 40 50 55 65 Oas1b mRNA max b 1 180 265 260 235 Tlr7 mRNA max b 1.8 4.8 5.1 3.5 4.2 a Three 2.5-year-old male treatment naive cynomolgus monkeys were orally dosed with 0, 1, 3, 9 and 30 mg/kg of 31 in 20% HP-ß-CD. A wash-out period of 2 weeks was planned in-between doses. Systemic compound concentrations were measured at 0.25, 0.5, 1, 2, 4, 8 and 24 hours post-dose (by LC-MS). Systemic cytokines levels were measured at 0.5, 1, 2, 4, 9 and 24 hours post-dose. Systemic ISG (ISG15, Mx1, OAS1), TLR7 and TLR8 mRNA expression levels were measured by Fluidigm RT-qPCR at 1, 2, 4, 8 and 24 hours post-dose. Blood was drawn to quantify compound. Values were rounded to the nearest whole number for PK parameters, to the nearest multiple of five for cytokines and ISG’s and to the first decimal for TLR’s. BQL, below minimum quantifiable limit; bfold-change relative to time 0 baseline. n.d.= not defined; cMx1 mRNA levels could only be measured reliably in 1 out of 3 animals.

Figure 4. Activation of Liver Kupffer Cells in Cynomolgus Monkey Demonstrated by CD68 IHC Staining 72h Post Oral Dose of 31.a


Page 15 of 43

A.

B.

C. 400 p = 0.04

CD68+ cells/mm²

300 200 100

C on tr ol

0

31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


a

A. Control animal; B. Dosed animal. C. Quantification of the CD68 IHC signal. Three 2.5-year-old male naive cynomolgus monkeys were dosed with 30 mg/kg of 31. After the 30 mg/kg dose animals were euthanized, and CD68 staining was performed on FFPE liver sections (B). In parallel, FFPE liver sections from unrelated control (washed out animals from another study) were processed the same way (A). (C) For each animal/treatment, the IHC slides were scanned and analyzed with the VisioPharm software. The positively labeled cells were counted in 30 mm2 area of the liver, then the total cell count/ per mm2 was calculated. Individual data, mean and standard deviation (SD) are shown. One-tailed student’s t-test with Welsh correction for non-identical SD was performed.

Healthy cynomolgus monkeys were chosen as the preclinical PK/PD model closest to humans in terms of immune system.23 Animals were given oral doses of 1, 3, 9, and 30 mg/kg of 31. No clinical observations and no change of body weight, body temperature nor clinical pathology 15 ACS Paragon Plus Environment


Page 16 of 43

changes observed until doses of 30 mg/kg where 2/3 animals exhibited emesis approximately 1h post-dosing. Absolute bioavailability of 31 was low (

8

Recommend Documents