Paris NASH Meeting 2019 Poster – Selective ROCK2 Inhibitors for The

Selective ROCK2 inhibitors for the treatment of fibrosis Emily P. Offer; Stuart A. Best; Nicolas E.S. Guisot; Phillip MacFaul; Sara Ceccarelli; Matthew R. Box; Neil Hawkins; Rosie Knowles; Amy Cooke; Alison Hunter; Charles Crossland; Sam Smith; Rebecca Holland; Peter Bunyard; Clifford D. Jones; Richard Armer. Redx Pharma Plc, Alderley Park, Macclesfield, United Kingdom. INTRODUCTION RESULTS ROCK2 is central to disease processes driving fibrosis pathology Redx’s ROCK2 inhibitors are potent and highly selective • The Rho Associated Coiled-Coil Containing Protein Kinase (ROCK) serine/threonine kinases, ROCK1 and • Redx’s ROCK2 series of compounds are potent and highly selective against ROCK1 and a panel of kinases, ROCK2, are central signalling proteins that regulate a range of cellular responses. tested in biochemical and cellular mechanistic assays. • These processes are central to • Cellular potency of ROCK2 selective inhibitors determined by inhibition of pMYPT1, a substrate the aberrant wound healing downstream of ROCK; ROCK1 or ROCK2 selective cell lines were generated with shRNA. response that can progress to KD025 REDX10178 REDX10616 REDX10843 Assay ROCK cellular signalling chronic injury and organ IC50 (µM) IC50 (µM) IC50 (µM) IC50 (µM) fibrosis. ROCK2 activity 0.16 0.002 0.004 0.017 • ROCK pathway inhibition ROCK1 activity 9.8 0.2 3.0 2.5 decreases fibrosis in vivo and Cellular ROCK2 selective pMYPT1 1.0 0.9 1.0 2.4 fibrotic markers in vitro. Cellular parental MCF7 pMYPT1* 0.9 4.1 22 24 • ROCK2 is upregulated in Cellular ROCK1 selective pMYPT1* > 30 20 > 30 26 diseases associated with acute Table 1. Selective ROCK2 compounds activity in biochemical and cellular assays. Comparison with KD02522. *Selective ROCK2 compounds expected to be less active in parental & ROCK1 selective MCF7 assays. and chronic inflammation, for A: ROCK expression in MCF7 cell lines B: pMYPT1 mechanistic assay

example where damage is ) 120 Parental MCF7

% ( 100 ROCK1 KD caused by high glucose and n

o ROCK2 KD

i t

i 80

parental

ROCK1 KD ROCK1 ROCK2 KD ROCK2

high fat diets. h

n 60

ROCK1 e

g 40 a • ROCK2 increased in liver t n 20

GAPDH e

c r

fibrosis models and the e 0 ROCK2 P increased ROCK2 signalling 10-9 10-8 10-7 10-6 10-5 10-4 GAPDH [REDX10843] (M) drives HSC activation. Figure 1. ROCK2 is downstream of fibrosis mediators and drives pathways associated with fibrosis pathology. Figure 2. MCF7 KD lines generated with shRNA express only one isoform (A). Representative inhibition of pMYPT1 in parental and ROCK1 or ROCK2 knockdown cell lines. RESULTS

Assay REDX10843 REDX10843 has suitable in vitro ADME properties and is orally bioavailable 9 targets >90% inhibition; Kinase panel (468 kinases) at 10 µM 32 targets >50% inhibition A: Mouse PK of REDX10843 B: Rat PK of REDX10843 Selectivity • REDX10843 is highly selective across a panel of No target > 50% inhibition; 100 10 CEREP SafetyScreen44 at 10 µM

468 kinases, and in the CEREP safety panel 8 targets > 25% inhibition )

) 50 mg/kg REDX10843 100 mg/kg REDX10843

M M 

 10 mg/kg REDX10843 30 mg/kg REDX10843 hERG IC > 33 µM (

( 50

with no significant off target activities 10 n

n 1 o

o Cardiotoxicity hNav1.5 IC > 100 µM i

i 50

t t a

observed. a

r r t

t 1 hCav1.2 IC50 > 100 µM

n n

e e c

• No activity in cardiotoxicity assays. c 0.1 Mini-Ames 5 strains No significant increase in

n n o

o Genotoxicity C

C 0.1 numbers of revertant colonies

( S9 plate based) d

• d n

No genotoxicity or mutagenicity observed in in n u

u 0.01 Micronucleus test using TK6 cells No statistically significant o

o Mutagenicity b

vitro tests. b 0.01 n

n ( S9 metabolic activation) increase in micronuclei

U U

Mouse Rat Dog Human

• Low to medium in vitro clearance, conserved 0.001 0.001 0 4 8 12 16 20 24 0 4 8 12 16 20 24 Microsome stability CLint (µL/min/mg) 35 39 15 4.9 across species (mouse, rat, dog and human) Time (h) Time (h) Hepatocyte stability CLint (µL/min/106 cells) 18 15 19 2.3 and good oral exposure in mouse, rat and dog. Plasma protein binding Percentage free (Fu %) 13 14 23 13 Figure 3. Oral pharmacokinetic profiles of REDX10843 dosed in C57Bl/6 mice (A) and Han Wistar rats (B). Formulation: HPMC (0.5% w/v in water) / DMSO [90:10]. Table 2. In vitro properties for REDX10843. ROCK2 inhibitors reduce fibrosis markers in kidney models in vitro and in vivo ROCK2 inhibitors reduce fibrosis markers in liver models in vitro and in vivo A: ROCK2 selective compounds inhibit in vitro expression of fibrosis markers in kidney mesangial cells A: ROCK2 selective compounds reverse myofibroblast phenotype in HSC derived myofibroblasts in vitro • Mesangial cells cultured in high glucose for 9 days; model diabetic environment (compounds day 4-9). • Hepatic stellate cell line differentiated on stiff plastic for 2 weeks demonstrate myofibroblast • Protein expression of CTGF, fibronectin, PDGF-BB, TIMP-1 and MCP-1 detected in the culture media. morphology and phenotype by induction of αSMA protein expression. CTGF expression in culture supernatant • Reduction of αSMA with ROCK2 inhibitors indicates suppression of the myofibroblast phenotype. KD025 REDX10178 REDX10616 REDX10843

n 120 o

ASSAY i

IC (µM) IC (µM) IC (µM) IC (µM) s A: Expression of αSMA in LX2 myofibroblasts B: QuantificationRED ofX1 0α6SMA16 in LX2 myofibrolastsREDX10843

50 50 50 50 s 100

) l

r 120 100

o p

r 80 DMSO REDX10843: 1 µM x CTGF assay (WB) Inactive 0.4 1.5 t

0.4 )

n 100 n

o 80

)

i t

c 60

f s

80 o

t e

Secreted TIMP-1 e

0.9 0.2 0.9 2.8 c

n r

r 60

o O

40 p

c x

% 60

(

i o

t 40

A A Secreted PDGF-BB 2.9 0.2 0.4 1.4

20 a

40 f

(

 R

 20 Secreted MCP-1 10 0.3 0.3 2.2 0 20 ( O M M M M M M M S µ µ µ µ µ µ µ 0 0 M 0 5 5 .3 8 2 5 D 2 .2 0 .0 .0 0 10 3 1 0.4 0.1 0.04 0.01 10 3 1 0.4 0.1 0.04 0.01 Secreted MMP2 ND 1.2 2.2 1.1 1 0 0 .0 0 REDX10616 concentration (µM) REDX10843 concentration (µM) [REDX10616] C: Mean IC50 of percentage inhibition of αSMA Table 3. Summary of analysis of culture supernatant from cells cultured with Redx’s ROCK2 inhibitors and Figure 4. Representative data for REDX10616 inhibition of Assay KD025 (IC50) REDX10178 (IC50) REDX10616 (IC50) REDX10843 (IC50) KD025 as a comparison. Supernatant analysed by western blot (CTGF) or ELISA. Data are from n≥3. CTGF. αSMA expression 0.3Fib rµMosis area (Siriu0.7s R eµMd) 0.3 µM 0.5 µM

B: ROCK2 inhibition of inflammatory response in vivo, in a model of acute kidney injury Figure 6. Redx ROCK2) inhibitors reduces αSMA expression*** in LX-2 differentiated myofibroblasts. Myofibroblasts expressing αSMA are dosed with compound for 48 h and

Fibros2i.s0 area (Sirius Red) %

αSMA expression detected( by IHC. Representative images shown where green indicates αSMA and purple nuclei stain (DRAQ5) (A). Representative data of cells dosed

a ** ) • Cisplatin induces an acute inflammatory infiltrate and with REDX10616 or REDX10843e (n=4*** ) (B). Table of mean quantification of αSMA expression IC50 of inhibition response (C).

2.0 r

% a

( 1.5

Fibroe sis area (Sirius Red) *

C: Gene expression in the kidney a ** v

response in the kidney. e

i t

Nephrin r

* a

) 1.5 *** s B: REDX10843 1 .significantly0 reduces fibrosis parameters in the STAM NASH model

Nephrin * e

2.0 o

(

• i **

This inflammatory response induces injury in the kidney Bcl2

t a

Nephrin i *

d **

* s

Bcl2 • Selective e ROCK2 inhibitor REDX10843 dosed therapeutically in the murine STAM NASH model r

* 1.0

* *

R 0.5

a *

and leads to an increase in the expression of ROCK2. 1.5

Fibronectin p

* s

Bcl2 e

d u v significantly reduces fibrosis in the liver when dosed BID or QD at 50 mg/kg.

Fibronectin *

t *

† ‡ r

i i

Collagen 1 R 0.5

* • ROCK2 selective inhibitors REDX10178 and REDX10616 , s 0.0

Fibronectin 1.0 S

s o

Collagen 1 *

* Reticular Fibroblasts (ER-TR7)

A: Siriusi Red expression in the liver B: Reticular fibroblasts (ER-TR7) MMP-3 d reduce the expression of genes associated with Collag* en 1 Fibrosis area (Sirius Red)

e 0.0

S *

MMP-3 R 0.5

* ) inflammation and fibrosis in a dose dependent manner. MMP-2 s *** 6 MMP* -3 ***

2u .0 **

% *

i Control Telmisartan 10 mg/kg QD

( r **

MMP-2 *

a ) * **

* 0.0

e S

TIMP-1 e Vehicle QD REDX10843 50 mg/kg QD

r %

MMP* -2

(

a *

1.5

e Vehicle BID REDX10843 50 mg/kg BID

e Control Telmisartan 10 mg/kg QD

TIMP-1 a *

e 4

A: Schematic of dosing regime *

i r

e CTGF e Vehicle QD REDX10843 50 mg/kg QD

TIMP-1 t

G e CTGF * 1.0

Cisplatin 25 mg/kg IP e Vehicle BID REDX10843 50 mg/kg BID

TGFß t

CTGF * Control Telmisartan 10 mg/kg QD i

d 2

s e

Day TGFß *

0 1 2 3 4 o * Vehicle QD REDX10843 50 mg/kg QD

R 0.5

IL-6 P TGFß s

* Vehicle BID REDX10843 50 mg/kg BID

i r

IL-6 *

i * * 0.0 0

IL-6 ICAM-1 S

* *

ICAM-1 *

Compound treatment QD * *

ICAM-1 * IL-1ß

* *

IL-1ß * *

IL-1ß TNFa * Control Telmisartan 10 mg/kg QD *

B: Kidney protein expression of ROCK2 * Vehicle QD REDX10843 50 mg/kg QD *

TNFa * *

TNFa MCP-1 * Vehicle BID REDX10843 50 mg/kg BID *

0 .0 1 0 * *

MCP-1 * MCP-1 *

0 .0 0 8 -100 -80 -60 -40 -20 0 20

H D

P -100 -80 -60 -40 -20 0 20

A 0 .0 0 6 -100 -80 -60 -40 Cha-2n0ge from0 vehic2le0 (%)

G / 2 Change from vehicle (%) K 0 .0 0 4

C Change from vehicle (%) O

R 0 .0 0 2 REDX10616 50 mg/kg QD REDX10616 50 mg/kg QD REDX10616 10 mg/kg QD REDX10616 50 mg/kg QD 0 .0 0 0 REDX10616 10 mg/kg QD n a iv e c is p la tin REDX10616 10 mg/kg QD

Figure 5. Mice treated for 5 days with compound orally, QD. Single IP injection of cisplatin on day 3 (A). Protein expression of ROCK2 detected by western blot normalised Figure 7. In the STAM NASH model mice are treated with streptozotocin (STZ) at day 2, and fed HFD from week 4. Compounds dosed therapeutically from week 6-9. Sirius to GAPDH (B). RNA isolated from whole kidneys were analysed for changes in gene expression by qPCR, plotted as change from vehicle treated animals (C). Red expression in the liver shows collagen expression (A). Quantification of reticular fibroblasts (B). Representative images are at 200x original magnification. SUMMARY • Redx have developed a series of compounds that are potent ROCK2 inhibitors in biochemical & cellular in • Targeting ROCK2 selectively allows a safe cardiovascular profile, as previously demonstrated with vitro assays and highly selective against ROCK1 and a panel of kinases. REDX10178 in telemetered rats†. • Demonstration that physiologically relevant markers of fibrosis pathways can be modulated in vitro in • The encouraging profile of series compounds is representative of the potential of the chemical series disease relevant phenotypic assays. which are currently in lead optimisation. • No safety concerns highlighted from early in vitro assessment (hERG, CEREP, AMES, micronucleus). • ROCK2 inhibition modulates fibrotic parameters in vivo with REDX10843, selective ROCK2 inhibitor. References: 1. Soliman et al, 2016; 2. Xie et al, 2006; 3. Waddingham et al, 2015; 4. Cicek et al, 2013; 5. Shimizu et al, 2013; 6. Yao et al, 2013; 7. Okamoto et al, 2013; 8. Zhou et al 2012; 9. Hu et al. 2018; 10. Luo et al, 2012; 11 Zhang et al 2016; 12. Trebicka et al, 2007; 13. Wang et al, 2018; 14. Kolavennu et al. 2008; 15. Baba et al. 2014; 16. Sun et al. 2006; 17. Nozaki et al 2015; 18. Zhou et al 2013; 19. Ho et al. 2012; 20. Knipe et al 2015; 21. Kast et al, 2017; 22. Flynn et al, 2016. Note: †Data presented at ASN Oct 2018, ‡Data presented at Keystone NAFLD and NASH conference, Jan 2019.

5th Paris NASH meeting, July 2019 | 11th-12th July https://www.redxpharma.com/