PIPE-505 is a gamma secretase inhibitor in developmentfor the treatment of SNHL associated with cochlearsynaptopathy. A series of in vitro and in vivo studies inanimal models of auditory loss have demonstrated twodistinct mechanisms of action (MOA) leading torestoration of hearing function. Specifically, PIPE-505 1)facilitates SGN neurite growth via the Netrin/DCC pathwayleading to regeneration of inner hair cell ribbon synapsesand 2) increases Atoh1 expression via reduced Notchsignaling leading to outer hair formation. These cellularregenerative effects together, restore auditory function.

In summary, PIPE-505 restores SGN connections to inner haircells and regenerates outer hair cells. These effects aremediated via two distinct γ-secretase substrates, DCC andNotch, respectively. A first-in-human study is planned toevaluate PIPE-505 in patients with SNHL associated withcochlear synaptopathy. Measures of audibility as well asspeech intelligibility will be assessed.

Isolated SGN

PIPE-505 causes dose dependent type I spiral

ganglion neurite outgrowth

PIPE-505 (μM)

NXT1505 neurite outgrowth DRC

[NXT1505](µM)

neu

rite

len

gth

vehic

leNT3

0.00

01

0.00

10.

01 0.1 1 10

0.000

0.002

0.004

0.006

0.008

Neurite outgrowth mediated by Netrin/DCC pathway

PIPE-505 restores synapses in vivo, mouse model

Introduction PIPE-505 increases hair cells via Notch inhibition

PIPE-505 improves auditory measures

F r e q u e n c y ( k H z )

Th

re

sh

old

(d

B)

4 8 1 6 2 4 3 2

6 0

7 0

8 0

9 0

C o n t r o l e a r

P IP E - 5 0 5 , 2 %

* * * ** * * *

* * * *

DMSO

PIPE-505

P IP E - 5 0 5 M

He

s5

(n

or

m t

oR

pl1

9)

v e h ic le .0 3 .3 3 1 0

0 .0

0 .5

1 .0

PIPE-505 restores auditory nerve synapses and hair cells

.001 .01 .1 140

45

50

55

PIPE-505 M

MY

O7

A+ O

HC

s p

er

10

0µ

m

ABR Wave 1PIPE-505 restores ABR

wave 1 amplitude, a measure

of AN function

ABR ThresholdPIPE-505 improves ABR

threshold, a measure of OHC

function

Hes5

Fo

ld

Incre

ase/v

eh

icle

PF2 0.2% 2%0.0

0.5

1.0

# O

HC

/re

gio

n

untreated Veh 0.02% 0.2% 2% 0

1

2

3

4

PIPE-505

***

***

↓ Atoh1

Ato

h 1

(fo

ld in

cre

ase/v

eh

icle

)

vehicle 0.2% 2%0

1

2

3

**↑Ex vivo explant In vivo drug treatment

PIPE-505 restores synapses in vivo, guinea pig model

[3H

] G

SI

bin

din

g

Vehicle 0.02% 0.2% 2%0

75

150

225

***

*

IT injection (30l)

Human PK prediction5% dose level>30d sustained exposure

autoradiography

Notch

Hes5

↑Atoh1

PIPE-505

↓

γ-secretase

↓Hes5

↑OHC

OHC

d B S P L

AB

R W

av

e I

1 0 2 0 3 0 4 0 . 5 0 6 0 7 0 . 8 0 .

0 .0

0 .5

1 .0

1 .5

V e h ic le

P IP E - 5 0 5 , 2 %

N a iv e a n im a l

*

Ne

uri

te l

en

gth

Veh

icle

1µM

150

5NT3

10pM

AZD

3839

100p

M A

ZD38

39

1nM

AZD

3839

10nM

AZD

3839

100n

M A

ZD38

39

1µM

AZD

3839

10pM

BM

S-7

0816

3

100p

M B

MS-7

0816

3

1nM

BM

S-7

0816

3

10nM

BM

S-7

0816

3

100n

M B

MS-7

0816

3

1µM

BM

S-7

0816

3

0.000

0.002

0.004

0.006

0.008

0.010

BACEi

Notch-

sparing

Brain Research Bulletin 2016

*** P<0.001, 1-way ANOVA, Tukeys post-hoc analysis

Nai

ve

98dB

0.2%

PIP

E-505

2% P

IPE-5

05

8

10

12

14

16

CtB

P2

pu

nc

ta/I

HC

P<0.0001

P=0.0013

98dB

# IHC synapses DCC α-fragment

CtBP2/GluR2

vehicle 2% PIPE-505

Hes5 Atoh1

↑

Hes

5

(fo

ld d

ecre

ase

/ve

hic

le)

Ne

uri

te le

ng

th

G041

Syn

ap

ses/IH

C

naive vehicle 0.02% 0.2% 2%

0

5

10

15

20

% NXT1505

******

P <0.05 1-way ANOVA, Tukey's post-hoc

# IHC synapses

PIPE-505

304.01Discovery of PIPE-505, a small molecule therapeutic for the treatment of sensorineural hearing loss (SNHL) associated with cochlear synaptopathy

*Daniel S Lorrain, Michael Poon, Karin J Stebbins, Geraldine Cabrera, Alex Broadhead, Jon Seiders, and Jeff RoppePipeline Therapeutics, San Diego, CA

Introduction

Small molecule inhibition of the muscarinic M1 acetylcholine receptor by potent, selective antagonists facilitate

OPC differentiation

*Michael M Poon1, Kym I Lorrain1, Ariana O Lorenzana1, Karin J Stebbins1, Alexander Broadhead1, Thomas O Schrader2, Yifeng Xiong2, Jill Baccei2, Ari J

Green4, Jonah R Chan4, Daniel S Lorrain3; 1Biol., 2Chem., 3Pipeline Therapeutics, San Diego, CA; 4Neurol., UCSF, San Francisco, CA

Inhibition of the muscarinic acetylcholinergic receptors by

non-selective muscarinic antagonists (e.g., clemastine,

benztropine) accelerates the differentiation of

oligodendrocyte precursor cells (OPCs) into oligodendrocytes

(OLs). Subsequent work has implicated the M1 isoform as

being a key driver of this phenomenon. In-house chemistry

efforts have identified a number of potent, selective M1

antagonists. Using these, we have characterized the effects of

inhibiting M1 in a diverse set of in vitro assays, including OPC

differentiation, cortical myelination, and organotypic brain

slice. Our data show that a selective, small molecule inhibitor

of M1 is sufficient to drive OPCs towards differentiation and

that the resulting oligodendrocytes express myelin basic

protein. Moreover, these OLs are functional, i.e., capable of

axonal wrapping and induction of nodes of Ranvier. Of note,

an M3 selective antagonist (Sagara et al., 2006) was not active

in a rat OL differentiation assay. In concert with our in vivo

data (also presented at this meeting), a strong case can be

made that the development of an M1 selective small

molecule antagonist is a promising approach for treating

demyelinating diseases such as multiple sclerosis.

[3H]NMS membrane binding

Rat OPC differentiation

Lysolecithin mouse brain slice Human brain slice

Fold selectivity against M1

Compound M1 IC50 (nM) M2/M1 M3/M1 M4/M1

Benztropine 3.19 16.9 11.2 4.78

Compound 57 0.716 343 763 430

PIPE-359 1.69 102 43 26

Compound 77 2.1 98.8 1270 212

PIPE-307 2.35 555 64.2 54.2

Compound 29 6.69 57.4 347 91.9

PIPE-683 7.45 698 175 292

Compound 107 8.91 178 117 313

Compound 51 13.5 417 1590 24.5

Compound 25 19.6 128 199 241

Compound 14 51.5 124 217 58.4

Fold selectivity against M1

Compound M1 Avg Ki (nM) M2/M1 M3/M1 M4/M1 M5/M1

Benztropine 1.14 16 2.67 8.21 2.7

PIPE-359 0.144 130 14.4 45.1 17.4

PIPE-307 0.349 73 18.5 38 259

Compound 57 1.13 22 7.11 29.9 5.37

Compound 25 1.41 160 8.81 189 736

Compound 77 1.48 8.8 41.1 13.5 54.5

Compound 51 2.34 390 113 148 538

Compound 29 2.55 90 17.4 1.95 6.07

Compound 14 3.6 >7692 59.5 174 583

PIPE-683 4.04 87 13.3 121 167

Compound 107 7.55 120 38.4 93.1 n.d.

Calcium mobilization

-9 -8 -7 -6 -5

-50

0

50

100

log[OPC-0001307], M

% O

PC

Dif

fere

nti

ati

on

(no

rma

lize

d t

o T

3)

IC50 25.4

Table 1 Pipeline compounds are potent and selective for human M1 in an mAChR recombinant membrane binding assay.

Vehicle T3 1µM PIPE-307

MBPHoechst

Table 2 Pipeline compounds are potent and selective in a cellularsetting. Compounds were evaluated in CHO-K1 cells overexpressing one ofM1-4 receptors for inhibition of ACh-induced calcium release at EC80

concentrations.

Compound EC50 (nM)

PIPE-683 18.9

PIPE-359 20.3

PIPE-307 25.4

Compound 107 78.5

Compound 77 175

M3 selective* > 3.00E+03

Figure 2 Pipeline compounds induced Mbp in cultured cortical mousebrain slice demyelinated with lysolecithin. Slices were cultured atpostnatal day 17, demyelinated and treated with compound. Mbp wasmeasured by quantitative PCR. The highly M1 selective peptide MT7 wasused as a positive control.

Veh 683 Veh 683 Veh 683 Veh 6830.0

0.5

1.0

1.5

2.0

OPC 683 pulse dose (rOPC diff; well average)

fold

MB

P (

no

rm t

o v

eh

icle

)

2h 6h 24h 72h

n.s. n.s. P=0.002 P=0.048

** *

Students t-testTime in compound

Dunnett's multiple

comparisons test Significant? Summary Adjusted P Value

Vehicle vs. MT7 Yes * 0.0136

Vehicle vs. PIPE-359 No ns 0.1802

Vehicle vs. Compound 77 Yes * 0.0444

-10 -9 -8 -7 -6 -5

0

10000

20000

30000

40000

[ACh] in OPC conditioned media

log[ACh], M

RL

U

60,000 12nM

40,000 5nM

30,000 2nM

cells/well nM ACh

Vehicle MT7 OPC359 OPC797

0

1

2

3

4

5

MBP M1

Fo

ld c

ha

ng

e v

s v

eh

icle

(2

-ddC

t )

300nM 1µM 1µM

* *

Figure 1 Pipeline compounds induce OL differentiation in rat OPCs at nMpotencies. Compounds were evaluated by immunocytochemistry in rat OPCs(Mei et al 2016). ACh levels in OPC conditioned media measured by calciumflux in hM1-CHO. Pulse dosing using PIPE-683, a structural analog of PIPE-307,shows 6h exposure is sufficient to initiate OPC differentiation.

Compound EC50 (nM)

PIPE-359 7.9

PIPE-307 10

Compound 107 30

PIPE-683 36.2

Compound 77 44.3

EC50 25.4 nM

Figure 4 Pipeline M1antagonists induced Mbpin a naïve human corticalbrain slice assay. Sliceswere incubated in MT7 orcompound for 9 days priorto RNA isolation andQPCR.

*Small molecule M3 selective antagonist from Banyu (Sagara et al 2006).Ki (nM) for M1: 4670, M2: 6730, M3: 25.5, M4: 3600 in 3HNMS membrane binding

ConclusionRat cortical myelination

-11 -10 -9 -8 -7 -6 -5

10

20

30

40

Cortical coculture (myelination)

[OPC1307](logM)

Myelin

ati

on

in

dex

(avg

MB

P s

eg

men

t le

ng

th/O

L)

MT7

vehicle

EC50 16.2nM

Figure 3 DifferentiatedOLs are myelinationcompetent. Myelinationwas evaluated in a ratcortical myelination assayas described previously(Lariosa-Willingham et al2016). Myelin segmentswere identified by MBPcolocalization with Tuj1(axonal marker) andaveraged per OL.

MBPTuj1Hoechst

Vehicle

PIPE-307

MT7

vehicle

Selective inhibition of M1 results in the differentiation ofOPCs into mature oligodendrocytes. Here, we described theidentification of potent, selective small molecule M1antagonists as evaluated by [3H]NMS binding and calciummobilization assays and further showed that these moleculesinduce myelination-competent oligodendrocytes. Thesemolecules also induced Mbp in mouse and humanorganotypic slice models. Together, this provides compellingevidence that inhibition of M1 with small moleculeantagonists developed at Pipeline have a positive impact intreating demyelinating disorders such as multiple sclerosis. Atthis point, a clinical development candidate has beenidentified and IND-enabling studies have been initiated.

Sagara, Y. et al. Identification of a novel 4-aminomethylpiperidine class of M3 muscarinicreceptor antagonists and structural insight into their M3 selectivity. J MedChem, 2006;49(19), 5653–5663.

Lariosa-Willingham, K.D., et al. Development of a central nervous system axonal myelinationassay for high throughput screening. BMC Neuro, 2016;17(6).

Mei, F. et al. Accelerated remyelination during inflammatory demyelination prevents axonalloss and improves functional recovery. eLife, 2016; 5: e18246.

References

PIPE-683 pulse dose

(rOPC diff; well average)

[PIPE-307](logM)

PIPE-307

[PIPE-307](logM)

1µM

Cmpd 77

1µM

PIPE-359

[PIPE-307](logM)

206.19

• These data highlight the therapeutic potential of a selective M1Rantagonist to benefit conditions such as MS in which demyelinationplays a role.

• A clinical development candidate has been identified and IND-enablingstudies have been initiated.

PIPE-359 binds to M1 with high affinity and demonstrates

selectivity over other muscarinic receptors

PIPE-359 promotes OPC differentiation in vitro and

increases remyelination ex vivo

Introduction

PIPE-359 enhances remyelination in a murine cuprizone model

PIPE-359 improves behavioral score, VEP N1 latency and

g-ratios in a murine MOG-EAE model

Orally administered PIPE-359 occupies M1 receptors and inhibits

M1 function in mouse forebrain

PIPE-359, a novel, potent and selective M1 muscarinic receptor antagonist as a therapeutic approach for remyelination in multiple sclerosis

* Karin J Stebbins1, Michael M Poon1, Alexander Broadhead1, Geraldine Edu1, Ariana Lorenzana1, Kym I Lorrain1, Thomas Schrader1, Yifeng Xiong1, Jill Baccei1, Christopher Baccei1, Ari J Green2, Jonah R Chan2 and Daniel S Lorrain1 ., 1 Pipeline Therapeutics, San Diego, CA; 2UCSF, Department of Neurology, San Francisco, CA

Potency (nM)

Membrane binding, Ki 0.144

Calcium flux, IC50 1.69

Fold-selectivity

M2/M1 M3/M1 M4/M1 M5/M1

Membrane binding, Ki 130 14 45 17

Calcium flux, IC50 102 43 26 315

-9 -8 -7 -6 -5

-25

0

25

50

75

100

log[PIPE-359], M

Veh

icle

Su

b %

of

T3 IC50: 20.3 nM

Novel small molecule approaches aimed at stimulating remyelination

would greatly complement immunotherapies and provide significant

neural protection in demyelinating conditions such as multiple sclerosis

(MS). Recently, we described the muscarinic M1 receptor (M1R) as an

important regulator of oligodendrocyte precursor cell (OPC)

differentiation and a promising target for drug discovery. We developed

PIPE-359, a novel, potent and selective M1R antagonist and highlight its

potential for remyelination.

0 4 8 12 16 20 240

50

100

Time (h)

M1

Occ

up

an

cy

(%

)

vehicle T3 300 nM PIPE-359

-10 -9 -8 -7 -6 -50.0

0.5

1.0

1.5

2.0

[PIPE-359](logM)

Mb

p i

nd

ucti

on

(o

ver

veh

icle

)

EC50 7.9 nM

MT7

naive

LPC

Vehicle LPC MT7 PIPE-3590

20

40

60

80

100

120

140

MB

P a

rea /

Ho

ech

st

co

un

t

300 nM 1 mM

*

Vehicle LPC MT7 PIPE-3590.000

0.004

0.008

0.012

Ca

sp

r co

un

t/T

uj1

are

a

300 nM 1 mM

MBP Caspr Hoechst

0 4 8 12 16 20 240.000001

0.00001

0.0001

0.001

0.01

0.1

1

10

Time (h)

PIP

E-3

59

(m

M)

Plasma Total Brain Total

M1 Ki

Brain Unbound

Slice EC50

Plasma Unbound

0 2 4 6 80.4

0.6

0.8

1.0

axon diameter (mm)

g-r

ati

o

Naive

remyelinated axons

unmyelinated axons

0 2 4 6 80.4

0.6

0.8

1.0

axon diameter (mm)

g-r

ati

o

Vehicle

remyelinated axons

unmyelinated axons

0 2 4 6 80.4

0.6

0.8

1.0

axon diameter (mm)

g-r

ati

o

PIPE-359

remyelinated axons

unmyelinated axons

0 2 4 6 80.4

0.6

0.8

1.0

axon diameter (mm)

g-r

ati

o

Naive

PIPE-359

Vehicle

remyelinated axons

unmyelinated axons

****

**** P<0.0001 vs. Vehicle and vs. Naive, ANOVA,

Tukey's multiple comparisons test

Naive Vehicle PIPE-3590

10

20

30

% r

em

yelin

ate

d a

xo

ns

******

Naive Vehicle PIPE-3590

20

40

60

80

% u

nm

yelin

ate

d a

xo

ns ********

mean SEM, 2 1500x TEM images of the distal lumbar spinal cord region were quantified per animal, n=5-6 animals per group

** p < 0.01, **** p < 0.0001 vs. vehicle, ANOVA with Dunnett's multiple comparisons test

0 4 8 12 16 20 240.001

0.01

0.1

1

10

Time (h)

Pla

sm

a P

IPE

-359 (m

M) 2 mg/kg IV

10 mg/kg PO

BLQ for IV is 0.005 mM (PO 0.009 mM)

Vehicle BQCA PIPE-359 MT7

0

100

200

300

Fo

reb

rain

[IP

1]

(pm

ol/m

g p

rote

in)

**

**** ***

100 mg/kg LiCl

** p<0.01, *** p<0.001, **** p<0.0001 vs. BQCA,

ANOVA with Dunnett's multiple comparison test

veh MT7 0.1 1 10 30

0

50

100

% M

1 O

ccu

pa

ncy

(2h

)

PIPE-359 (mg/kg)

Conclusion

0 4 8 12 16 20 24

Time (h)

PIP

E-3

59

(m

M)

Brain UnboundPlasma Unbound

M1

M3

M5M4

M2

M4

M2

M5

M3

M1

M1 occupancy ED50 ~ 2 mg/kg 30 mg/kg PIPE-359 provides ≥ 50% M1 occupancy for ~20h

30 mg/kg PIPE-359 fully inhibits M1-mediated IP1 accumulation

demyelination remyelination Free floating brain sections Confocal microscopy

0.2% cuprizone diet x 5 weeks

Normal chowVehicle or M1 antagonist

x 2 weeks

Normal 5 wks Veh 1 3 10 30

0

50

100

% r

em

yelin

ati

on

in

co

rtex

(MB

P)

PIPE-359 (mg/kg)

2 weeks remyelination

**

**

***

* p<0.05, ** p<0.01, **** p<0.00 vs. vehicle, ANOVA with Dunnett's multiple comparisons test

** p<0.05 Normal chow vs. 5 weeks CPZ diet, Student's t-test

Vehicle or PIPE-359 QD daily behavioral score

(d0-d21)

TEM on spinal cord and optic

nerves

Image analysis

(g-ratios)

Visual Evoked Potential (baseline, d7, d14, d21)

Induce MOG (d0)

0 2 4 6 8 10 12 14 16 18 20 22

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

Days post MOG induction

Clin

ical D

isab

ilit

y S

co

re PIPE-359

Veh

****

**** p<0.0001, mixed-effects model with Sidak's multiple comparisons test

pre-MOG 7d 14d 21d45

50

55

60

65

70

75

Days post MOG induction

VE

P N

1 L

ate

ncy

(m

s) Vehicle

PIPE-359

*

****

* P <0.05, **** P < 0.0001 vs. vehicle, 2way ANOVA, Tukey's

multiple comparison test

baseline N1 latency

PIPE-359 also promotes remyelination in optic nerves from this study (see Edu et al. poster)

30 mg/kg improves behavioral score 30 mg/kg improves VEP N1 latency

30 mg/kg PIPE-359 remyelinates axons in the spinal cord

PIPE-359 Exposure in rat

AUCP.O.

(hr*mg/mL)0.14

t1/2, P.O. (hr) 8.6

Cmax (mM) 0.09 @ 1h

F (%) 4.5

AUCI.V. (hr*mg/mL) 0.5

Cl (mL/min/kg), ER 61, 0.87

Vd (L/kg) 2.2

t1/2, I.V. (hr) 0.9

msome Clint

(mL/min/mg)23

PPB (%) 20

BTB (%) 13

[Brain]2h (mM) 0.026

[Plasma]2h (mM) 0.080

[CSF]2h (mM) <0.004

Kp, Kp,uu 0.32, 0.21

PPB (%) 20

BTB (%) 13

30 mg/kg PIPE-359 provides full M1 coverage for 8h in mouse brain

Normal chow

5 weeks CPZ diet

Vehicle

30 mg/kg PIPE-359

Quantify ROI in cortex

PIPE-359 dose-dependently enhances remyelination in a non-immune-mediated cuprizone model

PIPE-359 dose-dependently differentiates rat OPCs to oligodendrocytes in vitro

PIPE-359 dose-dependently induces Mbpin cultured mouse cortical slices

PIPE-359 increases MBP and Caspr protein in mouse cortical slices

MBP Hoechst

MBP

MBP

0.00001 0.0001 0.001 0.01 0.1

0

50

100

150

Free brain PIPE-359 (mM)

% M

1 O

ccu

pa

ncy

extrapolated

M1 occupancy EC50 ~ 3 nM

PIPE-359 increases the percentage of remyelinated axons and decreases the percentage of unmyelinated axons

206.18

LPC PIPE-359Vehicle

Multiple sclerosis is characterized by immune mediated myelin injuryand progressive axonal loss. Visual evoked potential (VEP) is a clinicallytranslatable model used in patients with multiple sclerosis due to itsability to measure myelin damage of the visual pathway through thelatency of VEP1 - which reflects the velocity of signal conduction alongthe visual pathway; while the amplitude of VEP is believed to beclosely correlated with axonal damage of the retinal ganglion cells(RGC)3. PIPE-359 is a novel, potent and selective M1 antagonist withgood oral exposure and brain penetration which is efficacious inrodent models of demyelination such as cuprizone and experimentalautoimmune encephalitis (EAE). Flash VEPs were recorded from EAEmice to determine if a selective M1 antagonist can demonstratefunctional remyelination. Spinal cords and optic nerves were collectedfor electron microscopy (EM) imaging and g-ratios were calculated toconfirm remyelination.

Introduction

VEP amplitude symmetry is preserved by M1 antagonists

The muscarinic M1 antagonist PIPE-359 demonstrates remyelination in vivo through visual evoked potential (VEP) and electron microscopy (EM) of mice with experimental autoimmune encephalitis (EAE)

Geraldine C Edu*1, Karin J Stebbins1, Alexander R Broadhead1, Michael M Poon1, Ariana O Lorenzana1, Thomas Schrader1, Yifeng Xiong1, Jill Baccei1, Ari J Green2, Jonah R Chan2 and Daniel S Lorrain1. 1Pipeline Therapeutics, San Diego, CA; 2Neurol., UCSF, San Francisco, CA

0 2 4 6 8 10 12 14 16 18 20 22

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

Day

Clin

ical S

co

re

PIPE-359

Vehicle

PIPE-359 reduces N1 latency shifts and N1P2 amplitude degradation

PIPE-359 is efficacious in VEP of EAE mice

0 2 4 6 8

0.4

0.6

0.8

1.0

axon diameter (um)

g-r

ati

o

Vehicle

PIPE-359

1 image quantified per ON (n=4 independent animals)

**** P < 0.0001 vs. vehicle, unpaired t-test

unmyelinated axons

remyelinated axons

****

Vehicle PIPE-359

0

10

20

30

40

% r

em

yelin

ate

d a

xo

ns

per

1500x E

M im

ag

e

✱✱

1 image quantified per ON from 4 independent animals

** P < 0.01 vs. Vehicle, unpaired t-test

Vehicle PIPE-359

0

10

20

30

40

50

% u

nm

yelin

ate

d a

xo

ns

per

1500x E

M im

ag

e

✱

1 image quantified per ON from 4 independent animals

* P < 0.05 vs. Vehicle, unpaired t-test

naive vehicle PIPE-3590

10

20

30

% r

em

yelin

ate

d a

xo

ns

******

naive vehicle PIPE-3590

20

40

60

80

% u

nm

yelin

ate

d a

xo

ns ********

mean SEM, 2 1500x TEM images of the distal lumbar spinal cord region were quantified per animal, n=5-6 animals per group

** p < 0.01, **** p < 0.0001 vs. vehicle, ANOVA with Dunnett's multiple comparisons test

Profiling M1 antagonists through VEP

Conclusions

40 60 80 100 1200

10

20

30

40

N1 Latency (ms)

N1-P

2 A

mp

litu

de (

V)

VEP at 21 daysvehicle (C1)

PIPE-307 30mg/kg

PIPE-307 3mg/kg

vehicle (C2)PIPE-683 30mg/kg

vehicle (C3)PIPE-359 30mg/kg

Control

40 60 80 100 1200.0

0.5

1.0

1.5

2.0

N1 Latency (ms)

Am

plitu

de R

ati

o

VEP at 21 days

PIPE-359 30mg/kg

PIPE-307 3mg/kg

PIPE-307 30mg/kg

PIPE-683 30mg/kg

Control

vehicle (merge)

40 60 80 100 1200

10

20

30

40

N1 Latency (ms)

N1-P

2 A

mp

litu

de (

V) vehicle

PIPE-359

VEP @ 21d

** p= 0.009181

*** p=0.0003392

r2= 0.3211

r2=0.3728

-20 0 20 40 60 80-20

0

20

40

60

80

100

N1 Latency Shift (ms)

N1-P

2 A

mp

litu

de D

ecre

ase (

V)

vehicle

PIPE-359r2=0.4640

**** p<0.0001

* p=0.04936

r2=0.2085

VEP @ 21d

0 7 14 2145

50

55

60

65

70

75

N1 L

ate

nc

y (

ms)

Vehicle

PIPE-359

*

****

* p=0.0196, **** p<0.0001 vs vehicle via ANOVA with Tukey's

days post MOG induction

➢ VEP is a sensitive measure of remyelination due to its ability to detectimpairment in the visual pathway before the onset of clinical disability inEAE mice.

➢ M1 antagonists demonstrate robust remyelination and axonal protection asseen by reduced N1 latency shifts and preserved VEP amplitude waveformsymmetry .

➢ Multiple compounds screened through this in vivo discovery paradigm havedemonstrated remyelination thus confirming a small molecule selective M1antagonist is a promising approach to treat multiple sclerosis.

➢ A clinical development candidate has been identified and IND-enablingstudies have been initiated

0 7 14 210

5

10

15

20

25

N1 L

ate

nc

y S

hif

t (m

s) Vehicle

PIPE-359

***

****

*** p=0.0008, **** p<0.0001 vs vehicle via ANOVA with Tukey

days post MOG induction

0 7 14 210

10

20

30

40

50

N1-P

2 A

mp

litu

de D

ecre

ase (

V)

Vehicle

PIPE-359 *#

* p=0.0304, # p=0.0260 vs vehicle via ANOVA with Tukey's

days post MOG induction

control

ve

hic

le

PIP

E-3

07

3m

pk

PIP

E-3

07

PIP

E-6

83

PIP

E-3

59

Co

ntr

ol

0.0

0.5

1.0

1.5

2.0

Am

plitu

de R

ati

o

30mg/kg

**p=0.0026, ## p=0.0032, *p=0.0356 vs

vehicle via ANOVA with Dunnett

VEP at 21days

✱✱

##

✱

Naïve Controls PIPE-307 3mg/kg

Vehicle PIPE-307 30mg/kg

Right

Left

206.19

EM of spinal cord and optic nerve show remyelination by PIPE-359

Fig 2A Differences in N1 latency shift can be detected at 7 days post MOG and becomes more significant overtime suggesting functional remyelination. 2B. N1-P2 amplitude decreases aredue to axonal damage at the retinal ganglion cells (You, Y. et.al 2011) and PIPE-359 prevents N1P2 amplitude degradation over time. 2C. At 21days in a scatter plot of each eye N1 latencyshifts and N1-P2 amplitude decreases are significantly correlated measures. PIPE-359 (Y = 1.365*X + 19.88) had a significantly more positive slope than vehicle (Y = 0.4056*X + 29.71).

ve

hic

le

PIP

E-3

07

3m

pk

PIP

E-3

07

PIP

E-6

83

PIP

E-3

59

Co

ntr

ol

40

60

80

100

N1

La

ten

cy (

ms)

@ 2

1 d

ays

30mg/kg

✱✱✱

✱✱✱✱

✱✱

####

***p=0.0003, **** p<0.0001, **p=0.0016, #### p<0.0001

vs vehicle via ANOVA with Dunnett

ve

hic

le

PIP

E-3

07

3m

pk

PIP

E-3

07

PIP

E-6

83

PIP

E-3

59

Co

ntr

ol

0

10

20

30

40

N1-P

2 A

mp

litu

de (

V)

@ 2

1d

ays

30mg/kg

**p=0.0037, * p=0.0111 vs vehicle ANOVA with Dunnett

✱✱

✱

A

B D

CA B C

Fig 3A. EM of the spinal cord and g-ratio calculation shows that PIPE-359 shows significant remyelination.3B . PIPE-359 shows %remyelination and % unmyelinated axons show values close to naïve mice. 3C. EMof the optic nerve and g-ratio calculation shows that PIPE-359 shows significant remyelination. 3D. PIPE-359 treated mice have a greater percentage of remyelinated axons than vehicle treated mice.

A B

Fig 5A. M1 antagonists all showed significant N1 latency difference from vehicle treated EAE mice by 21 days post MOG induction. B. Not all M1 antagonists but PIPE-307 atboth 3 and 30mg/kg showed a significant difference in N1P2 amplitude from vehicle at 21days. C. N1 latency vs N1P2 amplitude linear regression (xy scatter data points notshown) at 21 days of each M1 antagonist where the desired profile is low N1 latency and high N1-P2 amplitude. D. N1 latency vs amplitude ratio linear regression (xy scatterdata points not shown) at 21 days the desired profile is a positive slope where demyelination is seen with a negative slope (Y = -0.01194*X + 1.446). M1 antagonist PIPE-307at both 3mg/kg (Y = 0.03111*X - 0.5940) and 30mg/kg (Y = 0.05091*X - 1.282) achieves a positive slope very close to control mice (Y = 0.02540*X - 0.08203) .

Fig 4A-D. Sample VEP waveform examples of each treatment group showing 3 VEP traces and the average trace. Top panel is the right eye,bottom panel is the left eye. 4E. Standard VEP waveform components. Amplitude ratio description and calculation. 4F. M1 antagonists do notshow a degradation of N1P2 amplitude values and thus preserve of the symmetry of P1N1 and N1P2 amplitude waveforms

Fig 1A. Methods study timeline of PIPE-359 in the EAE model withflash VEP recordings performed in parallel. 1B. EAE mice are darkadapted in a room with red lights for 1hr prior to recording. Mice areanesthetized with ketamine/xylazine (75/10mg/kg) and eyes aredilated with 1% Tropicamide for 2 minutes. Genteal eyedrops wereapplied for lubrication and mice are placed on the Diagnosys CelerisModel D430 (Lowell, MA) 37°F heated platform. Electrodes arecarefully placed and flash VEP execution is performed 10 minutespost anesthesia wherein each exam consists of at least 3 runs, withpulse intensity 3 cd.s/m2, frequency 1 Hz, on time 4 ms, pulse color:white-6500K, 100 sweeps per result. Flash VEPs are recorded fromeach eye independently and simultaneously. N1 latency isdetermined by the average N1 from 3 VEP traces. 1C. Clinical signs ofEAE are first detected at day 11 and by the end of the study PIPE-359significantly alleviates EAE clinical disability. 1D. In same animals inparallel, flash VEP detects a difference in N1 latency with PIPE-359 asearly as 7days post MOG induction, over time the effect becomesmore significant out to 21days. 1E. Each individual eye scatter plotshow that N1-P2 amplitude and N1 latency measures aresignificantly correlated measures at 21 days post MOG induction andPIPE-359 (Y = -0.03580*X + 2.971) has a significantly more positiveslope than vehicle (Y = -0.01338*X + 1.656).

C D

A B C D E

F

A B

References

1. Green AJ, Gelfand JM, Cree BA, Bevan C, Boscardin WJ, Mei F, Inman J, Arnow S, Devereux M, Abounasr A, Nobuta H, Zhu A, Friessen M, Gerona R, von Büdingen HC, Henry RG, Hauser SL, Chan JR. Clemastine fumarate as a remyelinating therapy for multiple sclerosis (ReBUILD): a randomised, controlled, double-blind, crossover trial. Lancet. 2017 Dec 2;390(10111):2481-2489.

2. You Y, Klistorner A, Thie J, Graham SL. Latency delay of visual evoked potential is a real measurement of demyelination in a rat model of optic neuritis.Invest Ophthalmol Vis Sci. 2011;52(9):6911–6918.

3. You Y, Klistorner A, Thie J, Gupta VK, Graham SL. Axonal loss in a rat model of optic neuritis is closely correlated with visual evoked potential amplitudes using electroencephalogram-based scaling. Invest Ophthalmol Vis Sci. 2012;53:3662.

C D

E

0 2 4 6 80.4

0.6

0.8

1.0

axon diameter (m)

g-r

ati

o

Naive

PIPE-359

Vehicle

remyelinated axons

unmyelinated axons

****

**** P<0.0001 PIPE-359 and naive vs. Vehicle, ANOVA, Tukey's