Using Zinc Finger Nucleases for Subjects with Hemophilia B NIH - - PowerPoint PPT Presentation

Gene editing and Targeted Integration Using Zinc Finger Nucleases for Subjects with Hemophilia B

NIH Recombinant DNA Advisory Committee September 9, 2015 Nadia Ewing, MD, City of Hope John A. Zaia, MD, City of Hope Dale Ando, MD, Sangamo BioSciences, Inc. Michael Holmes, PhD, Sangamo BioSciences, Inc. Kathleen Meyer, MPH, PhD, DABT, Sangamo BioSciences, Inc.

Engineered ZFN

ZFP ZFP

• Contains two domains:

– Nuclease domain of FokI restriction enzyme – Designed zinc finger protein (ZFP)

• Cleaves as a dimer – cleavage requires the ZFNs to

be spaced by 5-6 bp

• May be engineered to cleave virtually any sequence
• Effectively a “designer restriction enzyme”

In Vivo Genome Editing – Modifying Liver Hepatocytes to Correct Monogenic Disease

Beumer et al PNAS 2008 (Dana Carroll)

ZFN-Driven Targeted Gene Addition ZFN-Driven Targeted Gene Addition / Gene Correction

Sangamo Confidential

Human albumin ZFN Pair and Corrective Gene Nucleic Acids Coding for ZFNs and Corrective Gene in AAV Vectors Peripheral IV Administration AAV Traffics to Liver And ZFNs Permanently Modify the Genome

AAV Vectors

Liver Cells Secrete Corrected Protein In Vivo Protein Replacement (e.g. Factor IX)

ZFN ZFN

Homology Corrective Gene

Systemic Delivery of ZFP Therapeutics via AAV Allows In Vivo Gene Correction of Monogenic Disease

Homology

9/16/2015 4

Albumin Provides an Ideal Safe Harbor Locus in Liver

• Targeted addition to a “Safe Harbor” locus

– Integration of a therapeutic transgene into a “safe harbor” – Albumin gene as proposed “safe harbor” locus: Safe, tissue-specific, and highly expressed – Targeting intronic sequence alleviates concerns of knocking out a large fraction of alleles by NHEJ – Single reagent set can be used for multiple indications

• Albumin—the most abundant blood plasma protein

 Safe to co-opt a small percentage of Albumin  Tissue specific and only produced in the liver  Very highly expressed relative to protein replacement therapies

Normal Albumin levels in blood - 40,000-50,000ug/ml; Normal F9 levels – 4-5ug/ml

• synthesis rate of albumin from liver – 105,000mg/week

ZFN-Based Approaches to Correct Hemophilia Using Albumin as a “Safe Harbor” Locus

hAlbumin: (Safe Harbor) Donor:

E1 E2

ZFN

hF9 Exons 2-8 Homology Homology E3 E4 E5 E6 E7 E8 E9 E10 E11 E12 E13 E14 E15 Stop ITR ITR

Therapeutic Transgene Cassette

Expression of hFIX Protein from the Albumin Locus

ZFN week 8 adult

Analyze clotting time

F9 ex2-8 donor 1x1011 AAV8 5x1011 AAV8 IV delivery

Analyze circulating hF.IX

Overview of Mouse Studies

8

E2 E3 E4 E5 E6 E7 E8 E9 E10 E11 E12 E13 E14 E15 Homology Homology w.t. Exons 2-8 E1

Albumin gene F9 donor:

ZFN ZFN

wild type hemophilia B model 2-60 weeks 2 weeks

Stable Circulating hF.IX Following mALB ZFN and Donor Delivery in WT Mice (Data through 60 weeks)

9/16/2015 9

Katherine High Lab Children's Hospital of Philadelphia

4 8 1230 4050 60

2000 4000 6000 Time (weeks after Injection) Plasma hF.IX (ng/ml-1)

4 8 1230 4050 60

20 40 60 80 100 Time (weeks after Injection) ALT IU/L

hF.IX Levels ALT Levels

Therapeutic hF.IX levels observed for 60 weeks (longest time point assayed) No change in ALT levels

10

Baseline Week 2 25 50 500 1000 1500 Plasma hF.IX (ng ml-1) Baseline Week 2 Wild Type 20 40 60 80 100

(s) n.s.

**

aPTT (s)

Antigen Activity 50 100

% of Normal

Correction of Prolonged Clotting Times in HB Mice

Katherine High Lab Children's Hospital of Philadelphia

Hemophilia B

• Inherited coagulation factor deficiency resulting from

reduced levels or absence of FIX

• Can routinely be distinguished from hemophilia A by

measuring FIX and FVIII activity

• Hemophilia A: 20.6 per 100,000 males

– Severe: 50-60%

• Hemophilia B: 5.3 per 100,000 males (1 in 30,000 males

world wide)

– Severe: 44%

Factor IX

• Factor IX gene (F9)

– On the long arm of the X chromosome – First cloned in 1982-1983, 34 kb long, contains 8 exons

• Serine protease (requires FVIII as cofactor to form the Xase

complex to cleave FX to generate Xa)

• Biosynthesis: liver
• Plasma concentration ~5 mg/mL (5,000 ng/mL)
• Normal FIX level: 50 - 150%; severity of disease correlates

with FIX levels (Severe - <1%, Moderate – 1-5%, Mild – 5-40%)

Control and Prevention of Bleeding in Hemophilia B Patients

–On demand (frequency depending on bleeding severity) vs prophylaxis (scheduled infusions twice/week to

• nce/1-2 weeks)
• Factor replacement products

– Fresh Frozen Plasma (when nothing else is available!) – Plasma derived FIX concentrates – Recombinant FIX concentrates – Extended half-life recombinant FIX

• For patients with inhibitory antibodies:

– Recombinant activated FVII, or activated prothrombin complex concentrates –

Economic Impact of Current Rx

–Direct cost

• Clotting factors account for ~98% of hemophilia care cost
• On demand: $200,000 to 300,000 for the average 60 kg

adult/year

• Prophylaxis: $500,000 to $900,000 /year
• Surgery or major trauma: ~$150,000 (FIX only)

–Indirect cost

• Disability
• Unemployment
• Absenteeism
• Early retirement

A Phase I, Open-Label, Ascending Dose Study to Assess the Safety and Tolerability of AAV Factor IX Gene Therapy via Zinc Finger Nuclease (ZFN) mediated targeted integration of SB-FIX in Adult Subjects with Hemophilia B

Investigational Products SB-FIX is composed of three AAV2/6 vectors that encode-- ZFN #1 that targets position 447-461 of the albumin locus ZFN #2 that targets position 468-485 of the albumin locus AAV 2/6 FIX Donor cDNA Study Design Phase 1, multi- site, open-label, dose escalation study Objectives Primary: Evaluate the safety and tolerability of SB- FIX Secondary: 1. Change from baseline in FIX levels 2. Change from baseline in coagulation parameters (aPTT) 3. Change from baseline in the frequency and severity of bleeding episodes 4. Immune response to AAV 2/6 and FIX inhibitors 5. Detection of SB-FIX in blood, saliva, urine, stool and semen

A Phase I, Open-Label, Ascending Dose Study to Assess the Safety and Tolerability of AAV Factor IX Gene Therapy via Zinc Finger Nuclease (ZFN) mediated targeted integration of SB-FIX in Adult Subjects with Hemophilia B

Main Inclusion Criteria

1. Signed informed consent 2. Male >18 years of age 3. Severe hemophilia B (native circulating FIX activity <1%) resulting from a missense mutation or a nonsense mutation that has not been associated with an inhibitor 4. Receiving prophylactic FIX replacement therapy as per current guidelines for the management of hemophilia, or on demand treatment of bleeding episodes 5. Received >20 doses of FIX replacement therapy with absent FIX inhibitor 6. No history of an allergic reaction or anaphylaxis to FIX products 7. No contraindication to the use of corticosteroids for immunosuppression 8. Agreement to use barrier contraceptive until at least 3 consecutive semen samples after SB-FIX are negative for AAV2/6

Study Schema

Screen Infusion and Safety Evaluations - 12 months

BL Day 0 -1 1 2 3 4 5 6 7 8 9 10 11 12 16 20 24 28 32 36 40 44 48 52 Week

Hospitalization for SB-FIX Infusion Long Term Follow Up 36 months Study Period ~ 15 months

M18, 24, 36, 48

3x/week ALT/AST

Mitigation of risk

• Exclusion of subjects with mutations associated with

inhibitor formation, or who have inhibitors at screening: monitoring for FIX inhibitors by (at least) monthly Bethesda assays

• Exclusion criteria for active hepatitis and liver disease
• Careful monitoring for transaminitis in weeks 2-12 and

treatment with corticosteroids if 2-fold rise above ULN

• Routine monitoring by liver elastography (Fibroscan)

Nonclinical Safety Evaluation Program for SB-FIX

• In Vitro Pharmacology Studies

– Proof-of-concept in mouse, nonhuman primate (NHP) and human hepatocytes

• In Vitro Safety Studies

– Soft agar transformation assay with human fibroblast cell line – Off-target analysis: SELEX-guided assessment (mouse, NHP and human genomes) – Off-target analysis (unbiased): Integration site assay

• In Vivo Pharmacology, Biodistribution and Toxicology Studies (discussed today)

– Mouse and NHP studies

In Vivo Pharmacology/Toxicology Studies

• Surrogate ZFNs and hF9 donor components required due to species-specific

differences in DNA sequences at the albumin intron 1 target locus

– Mouse – Nonhuman primate: cynomolgus and rhesus monkeys

• Study design

– Single-Dose IV administration – 1:1:8 ratio for ZFN1:ZFN2:hF9 donor – Co-administration of three AAV vectors on Day 1

• Pharmacology/Toxicology studies

– 3-month pilot pharmacodynamics (PD), biodistribution (BD) and toxicology study of AAV2/6 vectors in C57BL/6 mice – GLP 6-month PD, BD and toxicity study of AAV2/6 vectors in C57BL/6 mice – A sequential-dose PD, BD and toxicity study evaluating single- or co-administration

• f AAV2/8 vectors in rhesus monkeys; cynomolgus monkey bridging study

– A dose range-finding and dose-ratio study of AAV2/6 vectors in cynomolgus monkeys – A 60-day study to evaluate PD, BD and toxicity of AAV2/6 vectors in cynomolgus monkeys

21

GLP 6-Month Mouse Study 6-Month Cohort Biodistribution Data (vg/diploid genome)

Kidney Liver Lung Spleen Brain Heart Intestine Testes Semen 0.0001 0.001 0.01 0.1 1 1 10 100

196 197 198 201 202 203

AAV ZFN Copies/Diploid genome High Dose - 6 Month Cohort

Negative

Tissue AAV copies/dge

Kidney Liver Lung Spleen Brain Heart Intestine Testes Semen 0.0001 0.001 0.01 0.1 1 1 10 100

196 197 198 201 202 203

AAV hF9 Copies/Diploid Genome High Dose - 6 Month Cohort

Negative

Tissue AAV copies/dge

Highest AAV vector copy levels in liver No AAV vector detected in semen

Dose-response Relationship between Total NHP Surrogate AAV ZFNs and Gene Modification (% indels) in NHP Liver

Vehicle Control Low Dose Mid Dose High Dose 0.01 0.1 1 10

% Indels (Day 28)

Low dose: 1.5e12 vg/kg per ZFN and 1.2e13 vg/kg donor (total AAV dose=1.5e13 vg/kg) Mid dose: 5e12 vg/kg per ZFN and 4e13 vg/kg donor (total AAV dose=5e13 vg/kg) High dose: 1.5e13 vg/kg per ZFN and 1.2e14 vg/kg donor (total AAV dose=1.5e14 vg/kg)

Dose Response Relationship: Total AAV2/6 vectors and hFIX Levels

1.2e13 1.5e13 2.25e13 5e13 1.5e14 1 10 100 1000 10000

Total Dose (vg/kg) hFIX (ng/mL)

Therapeutic hFIX levels observed No change in circulating albumin levels throughout study

Non-responders not included in graph 1% of normal FIX levels

Summary of Liver ALT Levels and Terminal Histopathology

Strain/ Species AAV Dose (vg/kg) Number

• f

Animals Duration Immuno- suppression % Insertions and Deletions Peak ALT Elevations During In-Life Phase ALT Levels at Study End Liver Histopathology at Study End C57BL/6 Mouse 1e14 25 6 months No 30-48

Normal Normal No signs of hepatocellular hyperplasia, tumors or toxicity

Rhesus Monkey 1.5e14 2 > 1 year No 0.2 - 1.3

< 300 U/L; Day 21-49 Day 42-97 Normal No signs of hepatocellular hyperplasia, tumors or toxicity

Rhesus Monkey 1.5e14 6 8-9 months Yes 0.6-7.0

< 250 U/L; 3/6 animals following end of immuno- suppression Days 194-200 Normal No signs of hepatocellular hyperplasia, tumors or toxicity

Cynomolgus Monkey 1.5e14 2 3 months Yes 1.0-4.2

< 200 U/L; 1/3 animals Day 21-35 Normal No signs of hepatocellular hyperplasia, tumors or toxicity Species-specific ZFNs and hF9 Donor were used in these studies Normal ranges of ALT: Rhesus monkey 25-126 U/L; Cynomolgus monkey 14-127 U/L

Considerations for Clinical Dose Selection

• ZFN editing activity defined as % insertions/deletions (indels)

at the target albumin intron 1 locus in liver; measured by deep sequencing of the target locus

• hFIX expression dependent on sufficient levels of ZFN activity

and hF9 donor insertion into the target site

• Positive correlation between % indels (measure of gene

editing) at target locus and hFIX levels in NHPs

• Assumption that dose-response curve in similar for humans

and NHPs

• Selected clinical starting dose (5e12 vg/kg) is about half-log

less than lowest dose (1.2e13 vg/kg) evaluated in NHP study

Overview: Evaluating ZFN Specificity

• Evaluating ZFN specificity by monitoring off-target modification at

sites identified through:

I) SELEX-guided bioinformatic analysis  No off-target modification observed at the Top 40 predicted off-target sites in human transformed and primary hepatocytes II) Unbiased Off-target assessment via AAV integration site analysis  No off-targets found by mapping the sites of AAV integration III) Unbiased Off-target assessment via Oligo integration analysis 9/16/2015 26

Oligo capture analysis - overview

Oligo Capture Analysis Revealed Only One Off-Target Site

Rank Gene, (E)xon, Chromosome Capture events % indels low rAAV2/6 dose (2e4 vg/cell) % indels high rAAV2/6 dose (6e5 vg/cell)

On target Albumin 12540 6.1% 27.9%

1 MB

2091

N.S. N.S. 2 chr10

283

N.S. N.S. 3 USP25

73

N.S. N.S. 4 chr6

61

N.S. N.S. 5 MCC

48

N.S. N.S. 6 chr19

48

N.S. N.S. 7 chr17

48

N.S. N.S. 8 SMCHD1 (E) 42 N.S. 0.70 9 chr2

33

N.S. N.S. 10 DOCK1

30

N.S. N.S. 11 chr10

25

N.S. N.S.

* Locus not queried in follow up indel studies due to technical issues.

N.S. = not signifiant (Bonferroni-adjusted p-value is > 0.05)

Off-target identification by unbiased integration capture in K562 cells

Step 1 Human albumin SB-FIX ZFNs (SBS42906, SBS43043) Step 2

Off-target validation in Human hepatocytes

• Oligo capture assay was used to

identify off target sites: 60% of albumin alleles modified; 0.1% SMCHD1 alleles modified

• The 47 potential off-target sites

found by this method were evaluated in primary hepatocytes

• Only one of the 47 potential sites

showed any modification

• This site is exon 38 of the gene

Smchd1

Low Levels of Off-Target Activity Detected at Smchd1 in Three Different Human Hepatocyte Cells

On-target modification Off-target modification ZFN Dose Description MOI Albumin Smchd1 low On-target modification similar to levels expected in clinic 20K

6.1% N.S.

high On-target modification up to 6 fold higher than levels expected in clinic 600K

27.9% 0.70%

On-target modification Off-target modification ZFN Dose Description MOI Albumin Smchd1 medium On-target modification up to 2 fold higher than levels expected in clinic 3K

18.5% 0.33%

high On-target modification up to 7 fold higher than levels expected in clinic 30K

33.0% 0.40%

On-target modification Off-target modification ZFN Dose Description MOI Albumin Smchd1 low On-target modification similar to levels expected in clinic 0.1K

11.0% N.S.

high On-target modification up to 6 fold higher than levels expected in clinic 10K

29.0% 0.20%

N.S. = not significant (Bonferroni-adjusted p-value is >0.05)

Human Primary Hepatocytes Human Hepato R cells Human HepG2 Hepatoma cells

• In the three tested hepatocyte

lines, there is a clear dose relationship between on-target and Smchd1 off-target activity

• In 2/3 cases there is no detectable
• ff-target activity below 15-20%
• n-target activity
• Levels of on-target modification in

the clinical setting are predicted to be <10% at albumin

Smchd1 Was Found to be Weakly Targeted in Non-Human Primate Cells In Vitro and In Vivo

A) Smchd1 off-target site is completely conserved between human and NHP B) NHP surrogate reagents can cut Smchd1 in vitro in a dose-dependent manner: C) In Day 92 samples of two NHP studies (NHP04 and NHP05) there is no detectable off-target activity at Smchd1 despite on-target activities of up to 7.4% (NHP04) or 8.7% (NHP05).

On-target modification Off-target modification ZFN Dose Description DNA Albumin Smchd1 low

On-target modification up to 2 fold higher than levels expected in clinic

50ng

14.3% N.S.

high

On-target modification up to 3 fold higher than levels expected in clinic

400ng

37.1% 0.30%

N.S. = not significant (Bonferroni-adjusted p-value is >0.05)

Rhesus Rf/6A cell line

Smchd1 gene function

• Smchd1 (structural maintenance of chromosomes flexible hinge domain

containing 1 (SMCHD1)

• Homozygous k.o. female mice are embryonic lethal > presumably due to defects in

X-inactivation and misregulation of CpG island methylation. Homozygous male k.o. are initially ok, about half die early (unknown), the other half has normal lifespan/health > No phenotypes associated with liver were reported; > ZFN expression is restricted to liver through use of a hepatocyte-specific promoter; > Little to no off-target activity seen at on-target levels that model therapeutic setting; > Low levels of SMCHD1 modification should not affect liver cells given they are polyploid

• Smchd1 promotes NHEJ and suppresses HDR in response to

DNA damage (knock-down causes decreased cell survival after IR) > We have not observed any increase in DNA Damage Response markers in vitro

• Smchd1 mutations are associated with Facioscapulohumeral muscular

dystrophy type 2 (FSHD2) > This disease is caused by muscle cell-specific misregulation of DUX4 transcript; > ZFN expression is restricted to liver through use of a hepatocyte-specific promoter; > No ZFN activity has been observed in tissues other than the liver