Therapeutic Levels of FVIII Generated by CRISPR/Cas9-Mediated In Vivo Genome Editing in Hemophilia A Mice

Alan Brooks, Karen Vo, Dariusz Wodziak, Rangoli Aeran, Keith Abe, Cornell Mallari, Valerie Guerrero, Christopher Cheng#, Andrew ScharenbergCasebia Therapeutics, San Francisco, CA and Cambridge, MA #

ASGCT 22nd Annual Meeting, 2019

Session: “Gene Therapy for Metabolic Disorders: Proof of Concept and Beyond”

April 30th, 3:30-5:15, Heights Courtyard 1Abstract #428

Disclosures

| Confidential 2

All authors are employees of Casebia Therapeutics LLC

Hemophilia A - Background

|Confidential 3

• Genetic disorder caused by non-functional FVIII gene resulting in deficiency of active FVIII protein

• Protein replacement is the SOC:• 1 to 3 x weekly i.v. infusions• Peaks and troughs of FVIII resulting in less than ideal protection

• Ideal therapy is a constant level of FVIII in the normal range (50-150%) to provide continuous hemostasis

• AAV based gene therapy shows promise in clinical trials but may have limitations:• Life-long persistence of episomal based vectors un-proven• Only 1 treatment possible (due to nAB to AAV) with variable and unpredictable FVIII levels• Concern about utility in pediatrics due to vector dilution

• Approach presented: • Expression from FVIII gene integrated at a defined site in the genome• Potential for titration to the normal range which may represent the ideal curative therapy• Absence of liver specific promoter/enhancer minimizes risks of insertional gene activation

Targeted Gene Insertion Mediated by CRISPR-Cas9

| Confidential

CAS9

Exon 2Exon 1Albumin

Promoter

Exon 2Exon 1 GOI (secretable protein, no SP)SA pA

AAV packaged donor

GOI (secretable protein, no SP)SA pA

Exon 2Exon 1

GO secretableprotein, no SP) SApA or

orExon 2Exon 1

NHEJ

Forward Integration:

Reverse Integration:

INDELS (no impact on Alb expression):

Hybrid mRNA

SecretedProtein

In Vivo LNP Delivery of spCas9 mRNA /gRNA Targeting Albumin

Confidential 5

• Cationic lipid, neutral lipids, PEG-lipid• Single gRNA (chemically modified)• spCas9 mRNA (with NLS)

Self assembly(PrecisionNanoSystemsflow cell)

LNP• 50 – 90 nm• RNA concentration determined by Ribogreen• Dose mice by total RNA concentration (sgRNA + mRNA)

sgRNA 1 targets intron 1 of mouse albumin

Dynamic Light Scattering

I.V. LNPOn-target INDELS at day 7

HemAMice

# Genomic DNA from liver or spleen analyzed by TIDE

#

Visit Poster Abstract #494 for more details

0 . 5 1 2 P B S0

5

1 0

1 5

2 0

2 5

D o s e ( m g / k g )

To

tal

IND

EL

%

L iv e r l e f t l o b e

L i v e r r i g h t l o b e

S p le e n

Repeat LNP Dosing

Confidential 6

Murine SEAP pASAAAV8

2e12 vg/kg injected in HemA mice

Repeated LNP dosing, gRNA 1 (2mg/kg) T1= 0 days T2= 28 days T3= 49 days

SEAP activity in plasma measured weekly after each LNP dose Mean SEAP from 3 to 4 weekly assays plotted

Stuffer

0

2000

4000

6000

8000

10000

12000

14000

16000

1 2 3

SEAP

act

ivity

(mic

roU

/ml)

T1 T2 T3# Background SEAP activity in mice that received AAV only was 0

#

• Sequential increases in gene expression after each LNP dose

• This modality has the potential to tune the desired expression level after a single dose of AAV

4.7 Kb

Therapeutic Levels of FVIII After Genome Editing in HemA mice

Confidential 7

Human FVIII cDNA pASAAAV8

AAV-FVIII + LNP (2 mg/kg) –sgRNA1AAV-FVIII only

HemAMice

• Therapeutic FVIII activity achieved, approaching the normal range

• No FVIII when only the AAV-FVIII donor was injected demonstrating that cleavage of the genome required

• Antibodies against the foreign human FVIII protein develop in some mice

AAV + LNPAAV only

1 2 3 4

0

2 0

4 0

6 0

8 0

1 0 0

FV

III

Ac

tiv

ity

(%

of

no

rma

l)

2 - 1

2 - 2

2 - 3

2 - 4

2 - 5

W e e k s p o s t L N P

*

* Antibodies to human FVIII present

Targeted Integration Correlates to Cellular Distribution of Hybrid mRNA

Confidential 8

Day 10: FVIII activity in bloodDay 25: BaseScope™ and Targeted Integration by DD-PCR

BaseScope™ for hybrid mRNA in liver

Naï

ve m

ice

AAV8

-FVI

II

+LN

P

2% to 4% ofcells positive

Z ZAlb Exon1 FVIII

Human FVIII cDNA pASAAAV8

Group(n=5)

FVIII activity on day 10(% of normal)

Targeted Integration(% per haploid genome)

AAV8-FVIII +LNP 26 +/- 9 2.5 +/- 0.6

AAV8-HLP-FVIII #

28 +/- 5(peak on d30)

-

Naïve mice 0 0

# HLP promoter driving same FVIII CDS (but with signal peptide)

AAV8-FVIII (2e12 vg/kg) + LNP (2mg/kg)AAV8-HLP-FVIII (2e12 vg/kg)

HemAMice

Human FVIII cDNA pASAIntron 1For

RevP∗Intron 1DD-PCR:

Human FVIII cDNA (with SP)HLP pAAAV8

Long Term Expression of FVIII

Confidential

9

• Immune deficient NSG mice to prevent antibodies to human FVIII• FVIII detected with human FVIII specific capture-activity assay

Weeks post LNP

FVIII

act

ivity

(% o

f nor

mal

)

2e12 vg/kg AAV8 donor + 2mg/kg LNP2e13 vg/kg AAV8 donor + 2mg/kg LNPNaïve Mice

Stable expression to 7 months and continuing to monitor

FVIII levels dependent on dose of AAV donor

AAV and LNP Dose Response in HemA Mice

Confidential 10

Human FVIII cDNA pASAAAV8 +/- LNP

• FVIII level can be modulated by LNP dose

• Higher doses of AAV enable FVIII expression at lower LNP doses

• FVIII levels correlate to targeted integration frequency

% Mouse had antibodies to human FVIII on d14# Mouse had no INDELS indicative of failed LNP delivery

G 1 G 2 G 3 G 4 G 8 G 9 G 1 0

0

2 0

4 0

6 0

FV

III

Ac

tiv

ity

on

da

y 1

4

(%

of

no

rma

l)

A A V D o s e ( v g / k g ) : 2 e 1 2 2 e 1 2 2 e 1 2 2 e 1 2 1 e 1 2 5 e 1 2 2 . 5 e 1 3

L N P D o s e ( m g / k g ) : 0 . 5 1 2 0 1 1 1

2e12 vg/kg AAV

LNP LNP 1 mg/kg

AAV

0.010.11

0.62

0.01 0.080.19

0.53

# %

No

LNP

cont

rol

(Mean targeted integration %)

(0.01)(0.11)

(0.62)

(0.01) (0.08)(0.19)

(0.53)

2e12AAV Dose (vg/kg): 2e12 2e12 2e12 1e12 5e12 2.5e130.5 1 2 0 1 1 1LNP Dose (mg/kg):

Selection of gRNA can Impact Expression

Confidential 11

Human FVIII cDNA pASAAAV8

Optimized LNP, two gRNA targeting Albumin intron 1 gRNA1 (used for results in previous slides) gRNA2

FVIII Day 8 post LNP

• 5-fold improved expression using gRNA2 vs gRNA1• Improvement with gRNA 2 was not due to higher INDELS• 100% FVIII with low variability achieved with 2e12 vg/kg AAV8 donor• Measurement of targeted integration is planned

LNP only (1.5 mg/kg)AAV-FVIII (2e12 vg/kg) + LNP (1.5 mg/kg)

HemAMice

Mice dosed with LNP only

g R N A 1 g R N A 20

2 0

4 0

6 0

8 0

1 0 0

IND

EL

% (

d7

)

g R N A 1 g R N A 20

2 0

4 0

6 0

8 0

1 0 0

1 2 0

FV

III

Ac

tiv

ity

(%

of

no

rma

l)

Summary and Conclusions

Confidential 12

• CRISPR-Cas9 mediated targeted integration of FVIII at the albumin locus generates therapeutic levels of FVIII in mice, up to 100% of normal

• Low levels of targeted integration are sufficient for curative levels of FVIII

• Expression was stable (7 months and continuing)

• Targeted integration correlated to FVIII levels and was dependent on dose of both AAV and LNP

• FVIII level could be modulated by LNP dose

• Transient delivery of the nuclease as a LNP packaged mRNA has the potential for repeat dosing to “dial-in” FVIII levels while minimizing off-target risk

• Further optimization and evaluation in higher species in progress

Acknowledgments

Confidential 13

Chandra Patel

Luis Gamboa

Greg Cost

Gene Uenishi

Christopher Cheng

Faye Wu

Samantha Chin

Scott Munzer

Karolina Kosakowska

Kui Wang

Patrick Au

Mihail Rokas

Jim Burns

Co-Authors Additional ContributorsKaren Vo

Dariusz Wodziak

Rangoli Aeran

Keith Abe

Cornell Mallari

Valerie Guerrero

Christopher Cheng

Andrew Scharenberg