108 escaping immune activation through the use of cpg depleted AAV vectors

108 Escaping Immune Activation through the Use of CpG Depleted AAV Vectors Molecular Therapy Volume 21, Supplement 1, May 2013 Copyright © The American Society of Gene & Cell Therapy S44 IMMUNOLOGIC &[.]

Molecular Therapy Volume 21, Supplement 1, May 2013 Copyright © The American Society of Gene & Cell Therapy

S44

IMMUNOLOGIC & HOST RESPONSES IN GENE & CELL THERAPY

serotype rh.10 adeno-associated vector expressing the human 1AT

cDNA, mediates persistent therapeutic levels of human 1AT to the

lung Based on these preclinical studies and in support of a clinical

trial, we evaluated AAVrh.10h1AT for toxicity and effi cacy in

C57Bl/6 mice and African green nonhuman primates (NHP) after a

one-time intrapleural administration Three groups of 40 mice were

intrapleurally administered phosphate-buffered saline (PBS control)

or AAVrh.10h1AT at 1010 and 1011 genomic copies (gc)/mouse Two

additional groups received PBS or 1011 gc AAVrh.10h1AT vector

directly into lung parenchyma, as worst case scenario for misplaced

vector administration The mice were evaluated for 6 months, and

serum chemistry, neutralizing AAVrh.10 antibodies, gross pathology

and microscopic histopathology on major organs assayed at 4, 29,

92 and 183 days post-dosing AAVrh.10h1AT-intrapleurally treated

mice showed no vector related adverse effects nor did the group

receiving vector to the lung parenchyma Neutralizing antibody titers

to AAVrh.10 increased in a dose-responsive manner Titers remained

high after day 29, except for 1011 gc dose group females, for which

the average titers decreased by 50% at day 92 and remained at that

level through day 183 For the NHP study, a total of 36 monkeys

pre-screened for <103 antibodies to AAVrh.10 were treated intrapleurally

with PBS or 1012 or 1013 vector gc/animal Each dose group was further

subdivided into three euthanasia subgroups: 29, 91, or 361 days

post-dosing No toxicity in any organ was observed at any dose or time

point AAVrh.10 neutralizing antibody titers peaked at day 29 and

decreased through day 91, leveling off by day 361 AAVrh.10h1AT

distribution was evaluated at 1 yr post vector administration by

h1AT mRNA specifi c RT-PCR and Taqman in chest wall, diaphragm

and lung tissues All (4 out of 4) NHP that received the high dose

and 3 out of 4 animals that received the low dose were positive for

ha1AT mRNA at 1 yr In summary, intrapleural administration of

AAVrh.10h1AT does not cause any systemic or organ/tissue toxicity

at the doses studied and demonstrated persistent mRNA expression,

proximal to the lung, at 1 year, meeting the criteria to support the

initiation of a clinical trial for a new therapeutic option treatment of

h1AT defi ciency

107 Allele-Preferred Targeted Correction of

CFTR Gene in Cystic Fibrosis Induced Pluripotent

Stem Cells

Jacquelin H Bui,1 Ana M Crane,1 Xuan S Li,1 Philipp Kramer,1

Manuel Gonzalez-Garay,2 Wei Liao,1 Jianbin Wang,3 Helena C

Sun,3 David E Paschon,3 Dmitry Y Guschin,3 Philip D Gregory,3

Darrell N Kotton,4 Michael C Holmes,3 Brian R Davis.1

1 Center for Stem Cell and Regenerative Medicine, University of

Texas Health Science Center, Houston, TX; 2 Center for Molecular

Imaging, University of Texas Health Science Center, Houston, TX;

3 Sangamo BioSciences, Inc., Richmond, CA; 4 Boston University

Pulmonary Center, Boston University Medical Center, Boston, MA.

Cellular transplantation of lung stem/progenitor cells represents a

potential therapeutic approach for a variety of inherited monogenic

diseases Crucial to the success of such a therapeutic strategy is that

the transplanted cells and their progeny are corrected for the

disease-causing mutation and that the transplanted cells do not elicit an

immune response in the recipient In order to satisfy these criteria, we

are pursuing a patient-specifi c approach in which, starting with skin

or blood cells from patients with inherited lung disorders, autologous

induced pluripotent stem cells (iPSCs) are fi rst derived Utilizing

site-specifi c homology-directed repair, the disease-causing mutation

is corrected in the endogenous, chromosomal DNA sequence Finally,

a directed differentiation approach is employed to obtain highly

purifi ed populations of the relevant stem/progenitor cells from the

corrected iPSCs for purposes of transplantation We have employed

this approach to generate corrected, autologous iPSCs for patients

with Cystic Fibrosis (CF) Starting with skin fi broblasts of patients

diagnosed with CF, we have derived and characterized iPSCs, confi rming their pluripotency We then utilized zinc fi nger nucleases (ZFNs), designed to target the endogenous CFTR gene, to mediate correction of the inherited genetic mutation in this locus via homology directed repair in these iPSCs The corrected CF iPSCs retain normal karyotype, pluripotency, and a human ES-like expression profi le Correction was achieved for reprogramming transgene-containing

as well as transgene-free CF iPSCs The corrected CF iPSCs, when induced to differentiate in vitro, express the corrected CFTR gene Importantly, we observed an exquisitely sensitive, homology-dependent preference for targeting one CFTR allele vs the other This allele-specifi c targeting offers the potential for preferential targeting

of ZFN-mediated correction to dominant mutant alleles

Immunologic & Host Responses in Gene & Cell Therapy

Use of CpG-Depleted AAV Vectors

Susan Faust,1 Peter Bell,1 Benjamin Cutler,2 Scott Ashley,1 Joseph Rabinowitz,3 James Wilson.1

1 Pathology and Laboratory Medicine, University of Pennsylvania, Philadelphia, PA; 2 Washington University, St Louis, MO;

3 Pharmacology, Temple University, Philadelphia, PA.

Muscle-directed gene delivery of structural proteins, such as sarcoglycans and dystrophin, offers a promising therapy for patients with muscular dystrophy Adeno-associated virus (AAV) vectors are leading candidates for gene therapy However, in clinical settings, activation of de novo and pre-existing cellular and humoral immunoreactivity toward viral capsid and encoded transgene products has compromised safety and long-term gene replacement success Toll like receptor 9 (TLR9) has the potential to recognize unmethylated CpG motifs in both single and double stranded therapeutic expression cassettes packaged in an AAV capsid The AAV serotype, AAVRh32.33, vigorously activates an immune response, mimicking

in mice the vibrant and destructive cytotoxic T lymphocyte (CTL) response toward AAV and consequent loss of transgene product exression observed in human clinical trials, thereby refl ecting the biology of other AAV capsids in relevant settings To evaluate the role of TLR9 signaling in the induction of AAV-associated immunity and its effect on transgene stability, WT and TLR9-/- mice were I.M injected with 1E11 GC of the immunogenic vector, AAVRh32.33, expressing a nuclear LacZ (nLacZ) transgene WT controls completely lost beta-gal positive cells at 60 days post-injection, while abrogation

of TLR9 signaling resulted in stable transgene expression Heavy I.M CD4+/CD8+ T cell infi ltrates, induction of MHC II on myofi bers, and elevated capsid- and transgene-reactive responses were also observed

in WT but not TLR9KO vector recipients To determine whether CpG-depletion of AAVRh32.33 vectors would abrogate the robust cellular immune response and transgene loss, we generated a LacZ CpG-depleted vector genome (RhCpg-) that retained 16 CpGs located

in the inverted terminal repeat sequence while the WT vector (CpG+) also has 308 CpGs in the LacZ transgene (total=324 CpGs) RhCpG+ transduced muscle exhibited a progressive loss of detectable beta-gal expression, while the muscle sections from RhCpG- transduced mice displayed robust and stable transgene expression Mice that received the RhCpG- vector exhibited a signifi cant reduction in the percentage of LacZ responsive CD8+ T cells compared to control mice Further, a signifi cant decrease of primed transgene and capsid reactive IFN-gamma ELISPOT responses were observed in mice that received the RhCpG-, but not CpG+, vector Muscle transduced with RhCpG- vector also revealed minimal cellular infi ltrates and MHC II expression compared to RhCpG+ transduced muscle In conclusion, the TLR9 signaling pathway leads to LacZ reactive T cell responses, heavy cellular infi ltrate, and transgene loss following AAVRh32.33

Molecular Therapy Volume 21, Supplement 1, May 2013

IMMUNOLOGIC & HOST RESPONSES IN GENE & CELL THERAPY

administration CpG-depletion of vector genomes circumvents

TLR9 activation, resulting in attenuated adaptive immune responses,

signifi cantly reduced cellular infi ltrate, and prolonged transgene

expression CpG-depleted AAV vectors provide a platform for

improvements in safety and effi cacy of AAV gene transfer in humans

Formation in Protein and Gene Therapies for

Hemophilia

Debalina Sarkar,1 David M Markusic,1 Todd M Brusko,1 Roland

W Herzog.1

1 Pediatrics and Pathology, University of Florida, Gainesville, FL;

2 Medicine, Harvard Medical Center, Boston, MA.

The X-linked bleeding disorders hemophilia A and B are caused by

defi ciency in coagulation factor VIII (FVIII) or IX (FIX), respectively

In a subset of patients, inhibitory antibodies form against the

clotting factor, representing a serious complication of treatment that

increases morbidity and mortality Such immune responses are also

a concern in emerging gene therapies Treg are a major component

of immune tolerance to coagulation factors Therefore, we sought

to study whether ex vivo expanded autologous Treg could suppress

inhibitor formation In order to achieve >98% purity of murine Treg,

splenic GFP+ cells were fl ow sorted after extraction from BALB/c

mice containing a GFP reporter linked to FoxP3 message Sorted

cells were stimulated in culture using anti-CD28/-CD3 beads and

high-level IL-2 (1000 U/ml), which was replenished every 2 days

Stimulation was repeated at 1 week After 2 weeks, 20- to 50-fold

expansion was repeatedly accomplished For our dose of 1x10^6

Treg/mouse, expansion was suffi cient to treat 10 recipient mice

per donor The frequency of peripheral Treg approximately doubled

at day 7 post-transplant but returned to the normal range wihin 1

month Treg isolated from BALB/c F9 -/- x FoxP3-GFP mice and

expanded in vitro were transferred to (BALB/c F9 -/-) hemophilia B

mice Two days later, these mice (n=4) were treated by IM injection

of AAV1-human FIX vector The resulting lack of inhibitor formation

was maintained in 3 of 4 mice for >5 months despite challenge with

hF.IX in adjuvant at week 16 Control animals (gene transfer but no

Treg) formed high-titer anti-hF.IX (20 g IgG1/ml,  10 BU at week

6), and maintained inhibitor titers of >5 BU While total anti-hFIX

IgG declined over time and circulating hF.IX antigen eventually

emerged in the controls, IgG titers increased again after challenge

Coagulation times were consistently prolonged compared to

Treg-treated mice In other experiments, ex vivo expanded Treg were

transferred to hemophilia A mice (BALB/c F8e16 -/-), which were

then treated IV with human F.VIII (1 IU, once per week) for 2 months

Treg transplant again effectively suppressed inhibitor formation

Inhibitor titers in control mice were 15-20 BU at 1 month and 30-40

BU at 2 months Treg treated mice (n=5) formed at most low-titer

inhibitors (0-3 BU) Secondary Treg transfer confi rmed induction

of a suppressive response against F.VIII Treg therapy was also able

to control inhibitor titers in hemophilia A mice with pre-existing

inhibitors (n=6, 25 BU due to F.VIII replacement therapy) Half of

these (n=3) received Treg, and all mice received 8 more weeks of

F.VIII treatment While inhibitor titers in control mice increased to

100 BU, Treg therapy effectively controlled the response (to 15-20

BU, P<0.001) In summary, polyclonal ex vivo expanded Treg, while

perhaps not as potent as antigen-specifi c Treg, show promise as an

adjunct therapy to enhance clinical ITI or prevent inhibitor formation

in gene therapy for hemophilia Experience from current clinical trials

with polyclonal Treg for graft vs host disease and type 1 diabetes

should facilitate translation of the approach

Immunity to AAV Using Capsid Decoys

Federico Mingozzi,1 Xavier M Anguela,1,2 Giulia Pavani,1 Yifeng Chen,1,2 Robert H Davidson,1 Daniel H Hui,1 Mustafa Yazicioglu,1

Liron Elkouby,1 Christian J Hinderer,3 Armida Faella,1 Carolann Howard,1 Alex Tai,1 Gregory M Podsakoff,1 Margaret C Ragni,4

Shangzhen Zhou,1 Etiena Basner-Tschakarjan,1 J Fraser Wright,1,3

Katherine A High.1,2,3

1 Children’s Hospital of Philadelphia, Philadelphia; 2 Howard Hughes Medical Institute, Philadelphia; 3 University of Pennsylvania, Philadelphia; 4 University of Pittsburgh, Pittsburgh.

Adeno-associated viral (AAV) vector-mediated gene transfer has shown great potential as a therapeutic platform for inherited and metabolic diseases Systemic delivery of AAV vectors through the bloodstream is a safe, non-invasive, and potentially effective strategy to target a variety of organs However, humoral immune responses directed against the viral capsid represent a major obstacle

to intravascular delivery of AAV vectors, since neutralizing antibodies (NAb) to AAV, highly prevalent in humans, can completely block vector transduction This was well exemplifi ed in a clinical study of liver gene transfer for hemophilia B, in which pre-existing anti-AAV NAb titers as low as 1:17 completely prevented liver transduction Here we show a novel strategy to overcome the limitation of pre-existing humoral immunity to AAV C57BL/6 mice (n=5 per treatment group) were passively immunized with increasing doses

of purifi ed human IgG injected intraperitoneally Twenty-four hours later, anti-AAV NAb titers ranged from 1:3 to 1:100, suffi cient to neutralize >90% of an AAV8 vector expressing coagulation factor

IX (AAV8-FIX) injected intravenously at a dose of 5x109 vg/mouse The same vector dose was then formulated in increasing amounts of AAV8 empty capsids, showing complete rescue of liver transduction

at pre-treatment titers up to 1:100 Rescue of transgene expression correlated with the detection of antibody:capsid immune complexes

in plasma of animals 24 hours after vector delivery, at levels up to 3.8X108 complexes/ml, which suggests that the empty capsids act

as decoys for anti-AAV NAb Safety and effi cacy of the approach was further validated in non-human primate studies, showing that formulation of AAV-FIX vectors in a 9-fold excess of empty capsids results in enhanced FIX levels (371±43 vs 129±82, p<0.05) with no evidence of enhanced immunogenicity (measured by ELISPOT and antibody ELISA) or of other immune complexes-induced toxicities

as illustrated by normal histopathology on spleen, kidney, and lungs Next, to decrease the potential for empty capsids to be taken up by target cells and be presented onto MHC I, we developed a mutant capsid decoy based on AAV2, which lacked infectivity Empty capsids derived from this mutant effectively rescued AAV-2, -5, -6, and -8 vectors from antibody neutralization Conversely, they were unable to enter target cells and to enter the MHC I antigen presentation pathway and trigger CD8+ T cell-mediated clearance of hepatocytes in vitro Application of these fi ndings to the development of personalized formulations of vector product for intravascular delivery, based on

a subject’s NAb titer, will facilitate safe, effective AAV-mediated gene transfer through the systemic circulation in the presence of anti-AAV antibodies