SAFETY FEATURES IN THE DESIGN,

1
SAFETY FEATURES IN THE DESIGN, MANUFACTURE AND
CLINICAL MONITORING OF LENTIVECTORS FOR THE
TREATMENT OF PARKINSONS DISEASE, PROSTATE
CANCER AND AIDS.
FDA/BRMA 25-26 Oct 2001 Susan M Kingsman Senior
Vice President Research
2
The Retroviridae
Mammalian C-type
Lentivirus
Avian C-type
BLV/ HTLV
Spumavirus
D-type
B-type
Most Vectors
All share RT(pol) but differ in morphology,
pathogenicity and replication and gene expression
characteristics
3
Lentiviruses Analysis of Pol Sequences
Primate Lentiviruses
Non-Primate Lentiviruses
Equine Infectious Anaemia Virus, EIAV A
non-primate lentivirus Human Immunodeficiency
Virus, HIV A primate lentivirus
4
Why Develop Lentivirus-Based Vectors? Unique
Advantages for Gene Therapy
1. Simplicity 2. Up to 11kb capacity 3. Defined
integration of genes 4. Long term expression 5.
Transduction of non-dividing cells 6.
Transduction of slowly dividing cells

• Vectors for long term, stable therapy of chronic
  diseases
• Vectors for unmet medical needs

5
EIAV Vectors Gene Transfer to the Nervous System
Hippocampus/septal cholinergic system Site of
learning memory Major damage in stroke Primary
target in Alzheimers
Rat
Substantia nigra Primary target in Parkinsons
Muscle
Striatum Primary target in Huntingtons
Spinal motoneurons Primary target in ALS, SMN
Nucleus acumbens site of addictions
A powerful approach to the management of
neurological disease
6
EIAV Vectors Long Term Gene Expression in Vivo
Rat thalamus
8 days
24 weeks
Sustained expression minimises the need for
repeat invasive delivery
7
EIAV Vectors No Systemic Toxicity at Doses
Relevant to the Clinic

• Tail vein (107)
• 12 rats 21d
• No liver tox or damage by histology
• No vector in liver or lung

• Intra-tumoral (107)
• 21d No abnormal tox

• Intra-cranial (107)
• Over 100 rats
• Up to 8mths
• No clinical signs
• No histological abnormality
• Mild acute inflammation

• Intramyocardial (107)
• 80 rats
• 14d, no clinical signs of toxicity

Following MLV toxicity and biodistribution
protocols
8
An EIAV-Based Vector for The Treatment of
Parkinsons Disease
SIN LTR
IRES
IRES
TH
CH1
AADC
CMVp
6-OHDA lesion
Control
Aromatic AA Dopa Decarboxylase (AADC)
Tonic expression of dopamine, augmentation of
L-DOPA therapy
9
An EIAV Based Vector for the Treatment of Late
Stage Prostate Cancer
PURO
CMVp
TKm
CMVp
Wst-1 assay/PC3
120
100
80
Abs
PC-3 xenograft
60
Control
LNCaP
40
20

• Slow growing tumour
• 100 fold improved TK
• TK clinical experience
• Prodrug strategy terminates therapy easily
• Therapy selective for dividing cells
• Recurrent intractable malignant disease
• Local delivery with transrectal ultrasound
  guidance

0
GFP
TKm
-
Efficacy in progress

• MCA Preliminary discussions on LentiVectors
• GTAC and MCA protocol outline early 2002

10
LentiVector Specification
Basic Design and Production
Enhanced Design and Production
Gene specific considerations
Growth factors
Prodrug activating enzymes
Metabolic replacement
Cell death regulators
Disease and patient specific considerations
Terminal disease
Chronic disease
Adults
Children
Specification is an on-going process. Clinical
evaluation should occur in parallel with vector
and process development for certain indications
11
LentiVectors Key Vector-Specific Safety Issues

• Replication Competent Lentiviruses (RCL)
• in the product
• Minimise the generation and impact of RCLs
• Mobilisation of the transfer vector in target
  cells
• Minimise the inappropriate dissemination of the
  transgene

12
Minimising the Generation and Impact of RCLs

• If possible use a non-pathogenic virus
• Split the vector production system into at least
  three components
• If possible use a stable producer cell line
• Eliminate all non-essential coding and cis-active
  sequences
• Minimise the potential for homologous
  recombination
• Reduce packaging of vector helper components
• Use sensitive calibrated assays for detecting RCLs

13
Minimising the Generation and Impact of RCL
If possible use a non-pathogenic virus
Consequences of any RCL may be minimised
14
Features of HIV and EIAV
HIV
EIAV

• Extensive analysis
• Complex (6 accessory genes)
• Human pathogen
• Replicates in human cells
• Fatal immunodeficiency
• Global pandemic
• 30M AIDS/HIVve (worldwide)

• Moderate analysis
• Simple (3 accessory genes)
• Equine pathogen
• No replication in human cells
• Self limiting anaemia
• Endemic in horses in the Tropics
• Rare disease in stables (worldwide)

• Different lentiviruses have different safety
  profiles
• EIAV is rarely fatal in horses, does not cause
  immunodeficiency
• No a priori expectation that EIAV would be a
  human pathogen
• Low probability of a patient encountering EIAV

15
Minimising the Generation and Impact of RCL
Split the vector production system into at least
three components
Increases the number of recombination events
required to generate an RCL
16
LentiVectors The Basic System
1
RNA
2
3
gag-pol
env
3 Component Split System
1
Therapeutic Gene(s)
Vector genome
2
3
Low probability of generating RCLs requiring four
cross overs, two non-homologous
17
Minimising the Generation and Impact of RCLs
If possible use stable producer cell lines

• DNA recombination unlikely
• Genetically stable
• Conventional manufacturing parameters
  established
• Characterised starting materials

18
Minimising the Generation and Impact of RCL
Eliminate all non-essential coding and
cis-active sequences
Reduces recombination and pathogenic potential
19
Lentiviral ProViral Genomes
EIAV Non-Primate, 3 accessory genes
tat
rev
rev
LTR
tat
LTR
gag
env
pol
S2
RT RH DU IN
SD
TAR
Y
RRE
Pun
cPPT
HIV Primate, 6 accessory genes
vpr
pol
vpu
nef
env
RT RH IN
TAR
tat
vif
SD
rev
Y
cPPT
RRE
Pun
Transfer vector construction is complicated due
to multiple introns and accessory genes
20
The Additional Genes of Lentiviruses
Tat, Nef, Vif, Vpu, Vpr, potential
pathogenicity factors in man
21
1st Generation EIAV Vector Producer Cell8Z-20
Y
John Olsens BiG-45 Packaging Line
Issues Recombination homology, S2 and Rev
expressed in production
22
VIRUS YIELD FROM 8Z-20 PRODUCER


6
1x10


5
1x10


Viral Titer
(TU/ml)

4

1x10

3

1x10

0
-
24
24
-
48
48
-
72
96
-
120
72
-
96





Time (hours)


-
-
-
-
Sodium butyrate











Doxycycline

• Stable producer line constructed
• Scaled in roller bottles gt106 t.d.u./ml for 5
  days
• Standard vector for benchmarking improvements

23
EIAV Minimal Transfer Vector
pONY8.1 series GFP, LacZ, (optional cPPT, WPRE,
RRE)
pEV53B
Rev
Rev
pA
SD
Pol
Gag
CMVp
S2

(Y)
RRE
cPPT

• Minimal transfer vectors do not express any EIAV
  proteins.
• gt8kb insert capacity.
• Homology at Y
• cPPT and RRE optimise yields or potency but
  create regions of homology

24
Minimising the Generation and Impact of RCL
Minimise the potential for homologous
recombination
Rev
Rev
pA
pEV53B
SD
Pol
Gag
CMVp
S2

(Y)
cPPT
RRE

• Remove all homologous sequences from the Gag-Pol
  packaging plasmid
• Do not compromise expression of Gag-Pol
• Examine requirement for RRE/Rev

25
Lentivirus Codon Usage
Lentiviruses maintain a suboptimal codon bias
26
Codon Optimised Gag-Pol
pEV53B
Rev
Rev
pA
Pol
SD
Gag
CMVp
S2

(Y)
cPPT
RRE
pESYNGP

• Codons changed across gag-pol except in the
  critical frame-shifting overlap region
• Removes all blocks of sequence homology
• Rev-independent
• Achieved for EIAV and HIV gag-pol

27
Minimal EIAV Vector System
pONY8.1
pESYNGP
cPPT
pTOG

• No obligate requirement for any accessory genes
  but Rev/RRE can improve yields
• No functional viral proteins or significant
  coding regions in the transfer vector
• Transfer vector contains only 1100 nt of
  original EIAV nucleic acid
• No significant homology between components
• Approved in UK by HSE for CL1 use with non-toxic
  genes i.e. lowest containment
• Similar HIV-1 system has been constructed at OBM

28
2nd Generation Packaging Cell ESYN-29-iG Series
pESYNGP
pTOG pcDNA6/TR
pA
pA
TetO2
CMVp
VSV-G
TetR
CMVp
pTO-Rev
pA
TetO2
Optional codon optimised Rev
CMVp
Rev
Work in progress in collaboration with John
Olsen The system aims to meet the 8Z-20
production characteristics
29
Minimising the Generation and Impact of RCL
Reduce packaging of vector helper components
Lessons from retroviral vectors Eliminate
packaging and dimerisation signals
30
Self Packaging Gag-Pol Transcripts
Gag-Pol expression plasmids
Y
PBS
U3
R
U5
Gag-Pol
WT Y
CMV
YD (various)
Gag-Pol
CMV
Y0 (Synthetic)
Gag-Pol
XX
RNA only control
CMV
Gag
STOP

• Defining sequences that allow Gag-Pol proteins
  to package the cognate mRNA
• Defining sequences that allow co-packaging of
  Gag-Pol and transfer vector RNA

31
Self-packaging of EIAV Gag-Pol mRNA
Partial or complete removal of Y reduces
packaging to background levels No evidence for
piggy backing via dimerisation with the
transfer vector
32
Minimising the Generation and Impact of RCLs
Use sensitive, calibrated assays to detect RCLs
The assays must relate to the properties of any
RCL
33
Theoretical RCL structures A Generic View

• All RCLs have Gag-Pol
• The most likely Gag-Pol is vector derived
• Any other Gag-Pol is only rendered transmissible
  with
• vector derived Env

• Screen for transmissible reverse transcriptase
  (Pol).
• Qualify ambiguous results by a second screen for
  trasmissible Gag and Pol and by screening for
  transmissible Env

34
Proposed Testing for RCL

• Test vector preparations and post-production
  cells
• following current CBER guidelines for MLV
  derived vectors
• Use F-PERT assay as the primary assay tool
• Fluorescence-based Product Enhanced Reverse
  Transcriptase Assay
• Use PCR assay to resolve ambiguous PERT assay
  results
• Specific assay for hypothetical recombinant
  molecular structures

We have developed these assays and strategies in
collaboration and discussion with relevant UK
Government agencies (LGC, NIBSC) and with a
contract manufacturer, Q-One Biotech
35
F-PERT Product Enhanced Reverse Transcriptase
Assay (MLV, AMV, HIV-1/2, EIAV, SRV-1, HTLV-1/2,
FLV, FIV, BIV, CAEV, MVV, PoERV, SFV, SMRV)
Reverse transcription of MS2 RNA
NP-40
Disrupted VLP
MS2 cDNA
ABI Prism? 7700 (TaqMan)
pol (10-100/virion)

• Independent of the nature of events leading to
  RCLs
• Broadly applicable and high sensitivity (10-100
  particles)
• Modified to control against false positives

36
The F-PERT Assay Is Highly Sensitive
42
40
38
36
34
32
30
Ct
28
26
24
22
20
18
16
14
12
10
EIAV-1
EIAV-2
EIAV-3
EIAV-4
EIAV-5
Media
MLV-1
MLV-2
MLV-3
MLV-4
MLV-5
water
Media/DB
water/DB
MS2cDNA
EIAV neat
MLV neat
EIAV undisrupted
MLV undisrupted
Both MLV (Mn2-dependent) and EIAV
(Mg2-dependent) RT activities were detected by
PERT assay over a wide range of dilutions.
37
PCR Assay Nested Primers For The Detection
Of gag, pol and VSV-G

• Nested primers detect gag, pol and env,
  separately or as a linkage group
• No cross reaction with other retroviruses or
  lentiviruses
• Developed in collaboration with the Laboratory
  of the Government Chemist

38
(No Transcript)
39
RCL Amplification and Detection
Transduction by test article
Split cells at appropriate ratio over appropriate
time

• Positive Standard
• FeLV B/C
• Efficient amplification
• in 293 lymphoid cells
• Regulatory standard
• No LentiVector
• interference

• HEK293
• Tranduced by
• wide range of
• Vectors and
• pseudotypes
• Lymphoid (Raji, H9, Sup1)
• Potential in-patient
• target for RCLs

PERT assay on cell-free supernatants
PCR assay for EIAV gag, pol and env on genomic
DNA

• Amplification process is essential
• Readily adapted for screening producer cells

40
Minimising the Generation and Impact of RCL

• If possible use a non-pathogenic virus
• Split the vector production system into at least
  three components
• Preferably use a stable producer cell line
• Eliminate all non-essential coding and
  cis-active sequences
• Minimise the potential for homologous
  recombination
• Reduce packaging of vector helper components
• Use sensitive, calibrated assays for RCLs after
  amplification
• in human cells

Oxford BioMedica proposes that this is a
specification for discussion
41
LentiVectors Key Vector-Specific Safety Issues

• Replication Competent Lentiviruses in the product
• Minimise the generation and impact of RCLs
• Mobilisation of the transfer vector in target
  cells

Minimise the inappropriate dissemination of the
transgene
42
Minimising Mobilisation
Use a transfer vector that is poorly
mobilised Reduce the level of mobilisable RNA
in the target cell
43
Minimising Mobilisation
Use a transfer vector that is poorly mobilised
HIV-1 is the agent for mobilisation that is of
most concern
44
LentiVectors Cross - Packaging Assays
HEK293T
Biological titre on D17 cells by FACS for GFP
after 5d And after 2 serial passages
GFP
Gag-Pol
Env
1/4
1
HIV
1
1/4
1
MLV
Virion associated GFP RNA(TaqMan)
Test pairwise combinations for packaging and
transduction
45
LentiVector and Retroviral Vector Cross
Transmission Initial Titres
EIAV genome
HIV genome
MLV genome
Titre (TU per ml)
Gag-Pol
EIAV
HIV
MLV

• Cross mobilisation between EIAV and HIV occurs
  at less than 1000 fold
• the efficiency of homologous mobilisation
• EIAV and MLV are similarly poorly mobilised by
  HIV
• NB MLV is already in the clinic in HIVve
  patients

46
Minimising Mobilisation
Reduce the level of mobilisable RNA in the
target cell

• Achieved by modifying the transfer vector e.g.
  SIN vectors
• Should not be required for LentiVectors that are
  already poorly mobilised

47
Self Inactivating (SIN) EIAV Vectors
pONY8.1 SIN Vector Configuration in the
transduced target cell
SIN LTR

CMVp
LacZ
mRNA in target cell
Very low level expression of genomic RNA

• Residual viral genome is 900nt
• Deletes the LTR no evidence that this deletion
  is beneficial per se and it reduces
• vector options for second transcription units.
• SIN transfer vectors cannot be introduced into
  packaging cells by transduction unless a
• conditional SIN is used

48
Performance of EIAV LTR and SIN Vectors
In progress SIN and EIAV LTR vectors reduce
transferable RNA levels in the target cell by at
least 3 logs.
49
Mobilisation of HIV Vectors by HIV
A special case for the treatment of AIDS?
Turns virus escape into an attribute
50
ImmStat A Mobilisable Therapeutic Rz Vector for
the treatment of AIDS
Anti-HIV and chemokine receptor Ribozymes
CMV
R-U5
LTR
Y
Multi-Rz
CMVp
Tat inducible genomic and therapeutic RNA
Y
ImmStat Vector Particle
Constitutive therapeutic RNA

• pSYNGP allows Rz directed against the conserved
  gag-pol
• No immune responses anticipated against
  therapeutic RNA
• Mobilisation amplifies the therapy if virus
  escapes the first wave of Rzs

51
HIV-1 Production from Infected ImmStat Positive
and Negative cells
30000
25000
? Genetic variant ?
HIVGFP
20000
RT Units
15000
10000
HIVGFP- ImmStat
5000
0
0
2
4
6
8
10
12
14
16
18
20
22
24
Days
ImmStat delays HIV replication
52
LentiVectors Key Vector-Specific Safety Issues

• Replication Competent Lentiviruses in the product
• Minimise the generation and impact of RCLs
• Mobilisation of the transfer vector in target
  cells
• Minimise the inappropriate dissemination of the
  transgene

53
A LentiVector System With a Good Safety Profile
cPPT
Lac Z
LTR
pONY8.1 Z or G
CMVp
Y
WPRE
pA
pESYNGP
pTOG

• No pathogenic proteins
• Very poorly mobilised by HIV (sim. MLV)
• No potential for homologous recombination
• RCLs extremely unlikely

54
LentiVector Specification A Proposal

• A LentiVector system must show efficacy in a
  relevant animal model
• It must be capable of scale production at GMP
• It should be configured with the following
  points in mind

• 1. Eliminate non-essential proteins and
  sequences.
• 2. Ensure extremely low, preferably zero,
  homologous recombination potential between vector
  components.
• Show significant differential between self
  mobilisation and mobilisation with HIV or, use a
  vector that after integration generates a
  significant reduction in mobilisable RNA. Use
  MLV as a benchmark for mobilisation.
• Use a minimum of three split components in a
  stable packaging cell line
• or in a very low-homologous-recombination
  transient system.
• Use sensitive, calibrated assays for RCLs (follow
  MLV guide) with an amplification step in an
  appropriate human cell line.
• 6. Test for no significant acute toxicity in the
  animal efficacy model.

55
Supporting References
V.N. Kim et al (1998) Minimal requirement for a
lentivirus vector based on human
immunodeficiency virus type 1. J. Virol. 72,
811 K.A. Mitrophanous et al (1999) Stable gene
transfer to the nervous system using a
non-primate lentiviral vector E. Kotsopoulou et
al (2000) A Rev-independent human
immunodeficiency virus type 1 based vector that
exploits a codon-optimised gag-pol gene. J.
Virol. 74, 4839 N.D. Mazarakis et al (2001)
Rabies virus pseudotyping of lentiviral
vectors enables retrograde axonal transport and
acccess to the nervous system after peripheral
delivery. Human Molecular Genetics, 10,
2109. E.Martin-Rendon, et al (2001) Lentiviral
vectors for the treatment of Neurodegenerative
diseases. Curr. Opinion Molecular Therapeutics,
3, 476.
56
The LentiVector Team
Virology Kyri Mitrophanous Jonathan Rohll Fiona
Ellard Fraser Wilkes Anna Olsen Pippa
Radcliffe James Miskin Danny Chipchase Margaret
Esapa
Neurobiology Nick Mazarakis Enca
Martin-Rendon Mimoun Azzouz Karen Vincent Rob
Barber
Production Andrew Slade Wendy Blakemore Jay
Brown Linda Dove
BioMedica Inc Doug Jolly Holger Roehl
Collaborators John Olsen Karen O Malley James
Uney LGC NIBSC Q-One Biotech
Consultants Jonathon Weber Mark Saunders Robin
Weiss Mary Collins Patrick Aebischer Krys
Bankiewicz

Biological Systems Stuart Naylor Katie
Binley On Kan Leigh Griffiths Emma Carter Zoe
Askham