85 modeling the genomic integration site distribution of genotoxic and non genotoxic retroviral vectors in hematopoietic stemprogenitor cells in vivo

85 Modeling the Genomic Integration Site Distribution of Genotoxic and Non Genotoxic Retroviral Vectors in Hematopoietic Stem/Progenitor Cells In Vivo Molecular Therapy Volume 17, Supplement 1, May 20[.]

Molecular Therapy Volume 17, Supplement 1, May 2009 Copyright © The American Society of Gene Therapy

S34

RNA VIRUS VECTORS I

83 Optimization of Lentiviral Transduction of

Human Peripheral Blood Lymphocytes for TCR

Adoptive Cell Therapy

Neel K Karne,1 Scicheng Yang,1 Stephanie G Downey,1 Steven A

Rosenberg,1 Richard A Morgan,1 Steven A Feldman.1

1 Department of Surgery, National Cancer Institute, National

Institutes of Health, Bethesda, MD.

Adoptive cell therapy using tumor infi ltrating lymphocytes (TIL)

for the treatment of patients with metastatic melanoma has shown

dramatic tumor regressions As an alternative to TIL, peripheral blood

lymphocytes (PBL) have been genetically modifi ed by γ-retroviral

transduction to introduce a tumor-specifi c T cell receptor (TCR)

followed by a rapid expansion Lentiviral vectors can effectively

transduce PBL at much higher cell densities, theoretically obviating

the need for rapid expansion and may enable the adoptive transfer

of a less differentiated cell population The goal of this study was

to determine optimal parameters for transduction and expansion

PBL using lentiviral vectors After development of an optimized

lentiviral vector encoding a MART-1-specifi c TCR, we explored cell

density at transduction, time of transduction, stimulation method,

and various interleukin 2 (IL-2) concentrations post-transduction

Increasing cell density did not affect transduction effi ciency, but

did result in decreased cell expansion Anti-CD3/anti-CD28 bead

stimulation of PBL resulted in a higher percentage of transduced T

cells which expanded approximately 2-fold better when compared to

OKT3-stimulated cells Interestingly, transduced cells stimulated with

OKT3 preferentially expanded the CD8+ T cell subset Functionally,

there was no difference between cells stimulated with beads or

OKT3 Increased concentrations of IL-2 or a combination of soluble

OKT3 and anti-CD28 did not affect transduction or subsequent cell

expansion; however, inclusion of soluble OKT3 and/or anti-CD28

post-transduction resulted in higher numbers of CD8+ T cells as

compared to bead or OKT3 stimulation alone These fi ndings provide

a strategy for the optimization of lentiviral transduction of PBL with

the goal of developing a clinical protocol for the treatment of patients

with metastatic melanoma

84 Histone H3.1 Enhances the Retroviral

Transduction of a Hematopoietic Cell Line Model

and Primary Hematopoietic Stem Cells

Pascal R Beauchesne,1,2 Katherine J Bruce,1,2 Michelle Miller,3

Sanja Sekulovic,3 Bruce D Bowen,2 R Keith Humphries,3 James

M Piret.1,2

1 Michael Smith Laboratories, University of British Columbia,

Vancouver, BC, Canada; 2 Chemical and Biological Engineering,

University of British Columbia, Vancouver, BC, Canada; 3 Terry

Fox Laboratory, British Columbia Cancer Research Center,

Vancouver, BC, Canada; 4 Medicine, University of British

Columbia, Vancouver, BC, Canada.

The effectiveness of recombinant retrovirus-mediated gene transfer

can be severely limited due to mass transport limitations, thereby

requiring vector concentration steps and repeated exposures of the

target cells to achieve desirable scientifi c and therapeutic outcomes

The low diffusivity (∼6 x 10-8 cm2/s) and short half-life (∼6 h) of

retroviruses combined with various physico-chemical forces limit

the encounter frequency of vectors and target cells We previously

reported that whole cell or nuclei lysate fractions increased the

transduction of TF-1 cells by over 40-fold relative to a no additive

control Despite their effectiveness, the complexity of cell lysates

may limit their use, especially at the clinical level Defi ned cationic

reagents such as protamine sulfate, an arginine-rich peptide, have

been shown to improve retrovirus-mediated gene transfer by 5- to

8-fold in TF-1 cells Histones, a group of cationic nuclear proteins

rich in lysine and arginine residues, provide comparable enhancement

However, their use with retroviral vectors has been limited to mixed tissue-derived preparations, thus confounding the properties of the individual histone types To further defi ne the active components within the nuclear lysate fraction, we screened the effect of individual histone types on the transduction effi ciency of TF-1 cells A

arginine-rich fraction (f3) mainly composed of histone H3 extracted from calf thymus were added to TF-1 cells and retroviral vectors with

a GFP reporter gene produced from PG-13 packaging cells Unlike protamine sulfate that requires a tissue culture-treated surface, both histone fractions enhanced retroviral transduction independently of surface type While the f1 histone fraction yielded a similar increase

as protamine sulfate, the f3 fraction enhanced transduction by over 35-fold, a 4-fold improvement over protamine sulfate Similar results were obtained using unmodifi ed human recombinant histones as H1, H2A, H2B and H4 matched the effect of protamine sulfate, while the H3.1 variant provided a 35-fold increase in transduction This recombinant H3.1 histone protocol was successfully applied

to the retroviral transduction of mouse bone marrow enriched in hematopoietic stem cells (HSC) by 5-FU pre-treatment and compared

to a co-culture method The addition of histone H3.1 was as effective for transducing the HSC population where 50% of donor-derived cells were GFP+ compared to 43% using the co-culture method and 5.6% with no additives as verifi ed by a long term reconstitution assay This novel protocol based on the recombinant human histone H3.1 provides

a fully-defi ned alternative with higher transduction effi ciency and is compatible with sensitive primary cell populations such as HSCs

85 Modeling the Genomic Integration Site Distribution of Genotoxic and Non Genotoxic Retroviral Vectors in Hematopoietic Stem/

Progenitor Cells In Vivo

Eugenio Montini*,1 Daniela Cesana*,1,2 Jacopo Sgualdino,1,2

Manfred Schmidt,3 Alessia Capotondo,1,2 Francesca Sanvito,4

Cynthia Bartholomä,3 Marco Ranzani,1,2 Fabrizio Benedicenti,1

Lucia Sergi Sergi,1 Claudio Doglioni,4 Alessandra Biffi ,1 Christof VonKalle,3 Luigi Naldini.1,2

1 HSR-TIGET, Milan, Italy; 2 San Raffaele University, Milan, Italy;

3 NTC, Heidelberg, Germany; 4 HSR, Milan, Italy.

Gamma-Retroviral vectors (gRV), which are commonly used in

gene therapy, can trigger oncogenesis by insertional mutagenesis

We previously dissected the contribution of vector design and viral integration site selection (ISS) to oncogenesis using an in vivo genotoxicity assay based on transplantation of vector-treated tumor prone mouse hematopoietic stem/progenitor cells (HSPC)

By swapping genetic elements between gRV and lentiviral vectors (LV), we demonstrated that transcriptionally active long terminal repeats (LTRs) are major determinants of genotoxicity even when reconstituted in LVs, and that self-inactivating (SIN) LTRs enhanced the safety of gRVs By comparing the genotoxicity of vectors with matched active LTRs, we were able to determine that substantially greater LV integration loads are required to approach the same oncogenic risk as gRV This difference in facilitating oncogenesis is likely to be explained by the observed preferential targeting of cancer genes by gRVs This integration-site bias was intrinsic to gRVs, as

it was also observed for SIN-gRVs that lacked genotoxicity in our model These studies validate the biosafety of the SIN design and the lower impact of LV ISS on genotoxicity The use of a modifi ed

LV with active LTR was instrumental to dissect the role of ISS in modulating vector genotoxicity in mice Here we took advantage

of this genotoxic LV to mutagenize human CD34+ HSPC and

study its integration pattern in vitro and in vivo in human/mouse

hematochimeras Cord blood CD34+ cells were transduced with

-/- gchain -/- mice to allow expansion and differentiation in vivo High

Molecular Therapy Volume 17, Supplement 1, May 2009

BIOLOGY OF AAV AND VECTOR DEVELOPMENT I

levels of transduction were achieved in vitro (>90%) and substantial

engraftment of marked human cells was found in the bone marrow,

spleen and thymus of mice at 13 weeks after transplant The

genomic integration profi le of the these vectors is currently being

determined after 2, 4, 6 and 8 weeks of in vitro culture and in blood

and organs of transplanted mice at 13 and 20 weeks after transplant

Comparison of the genomic integration pattern of the genotoxic and

SIN LV at different time points will indicate whether signifi cant

differences can be observed between these vectors and during time

We expect that the genotoxic LV induces clonal dominance driven by

insertional mutagenesis and that its integration profi le preferentially

target specifi c gene classes in vivo and uncover new genes whose

deregulation may lead to clonal outgrowth of human hematopoietic

cells Moreover, the genotoxic LV integration profi le could be used as

reference against which to assess integration data obtained from the

growing number of clinical trials using LV for HSPC gene therapy

*equal contribution

86 Functional Analysis of Retroviral Vector

Integration Sites in the Treatment of Murine

X-Linked Chronic Granulomatous Disease Using

Gene Ontology

Troy B Hawkins,1 Mohammed A Sadat,2 Jessica Dantzer,1,3 Ken

Cornetta,1,2,4 Sean D Mooney,1,3 Mary C Dinauer.1,2,5

1 Medical and Molecular Genetics, Indiana University School of

Medicine, Indianapolis, IN; 2 Pediatrics, Wells Center for Pediatric

Research, Riley Hospital for Children, Indianapolis, IN; 3 Center

for Computational Biology and Bioinformatics, Indiana University

School of Medicine, Indianapolis, IN; 4 Medicine, Indiana

University School of Medicine, Indianapolis, IN; 5 Microbiology

and Immunology, Indiana University School of Medicine,

Indianapolis, IN.

Retroviral vectors have become a common and effective tool in the

treatment of the genetic defects characteristic of inherited immune

disorders such as X-linked Severe Combined Immunodefi ciency

(SCID) and Chronic granulomatous disease (CGD) This treatment

involves the genetic modification of hematopoietic stem cells

(HSCs) by random integration of retroviral DNA into the host cell

genome followed by competitive clonal selection and engraftment

Gene therapy by transduction of retroviral vectors has been shown

to effect unintended consequences on host cells, disrupting cellular

gene function through activation by potent viral enhancers located

in vector long-terminal repeats (LTRs) Here we describe analysis of

Gene Ontology (GO) functional annotations of genes neighboring

236 retroviral insertion sites (RISs) in a study of gene therapy

treatment of murine X-CGD Genomic regions downstream of vector

integrations were amplifi ed and identifi ed by ligation-mediated

PCR (LM-PCR), sequenced, and aligned to corresponding regions

in the mouse genome Genes within ± 300kb were identifi ed by the

SeqMap tool and annotated with GO biological process and molecular

function terms as provided by MGI Gene annotation subgraphs were

completed using the full set of ancestral terms and analyzed for

over-representation against the MGI background using the hypergeometric

distribution Terms identifi ed as over-represented (phg ≤ 0.05) were

compared between test cohorts in the study: lethally vs sub-lethally

irradiated mice and primary vs secondary and tertiary transplant

recipients Integration-associated genes in both the lethally and

sub-lethally irradiated cohorts were enriched for biological process and

molecular function terms related to transcription Some differences

between the two groups were identifi ed, with genes near RISs in

lethally irradiated mice being signifi cantly more enriched (pt ≤ 0.05)

for cell proliferation, GTPase signal transduction, and hematopoiesis,

and those in sub-lethally irradiated mice being more enriched for

phosphorylation and protein modifi cation Among specifi c tissue

subsets, genes near integrations in spleen samples were enriched for

phosphate metabolism and kinase activity in sub-lethally irradiated mice and those in bone marrow (BM) or both BM and spleen were enriched for regulatory processes and transcription factor activity in lethally irradiated mice Secondary and tertiary transplant recipients were enriched for genes related to transcription, similar to those found

in primary transplant recipients

Biology of AAV and Vector Development I

87 The Viral Capsid Signifi cantly Infl uences the Pathway and Rate of Adeno-Associated Virus Intracellular Traffi cking in HeLa Cells

Nicholas W Keiser,1 Paul M Kaminsky,1 Ziying Yan,1 John F Engelhardt.1

1 Department of Anatomy and Cell Biology, University of Iowa, Iowa City, IA.

Intracellular traffi cking has been identifi ed as a rate-limiting step

in adeno-associated virus (AAV) transduction Following receptor binding, AAV enters the cell via clathrin-mediated endocytosis, and traffi cs through various endosomal compartments before it escapes into the cytoplasm and enters the nucleus Previous studies by our laboratory have identifi ed the late and recycling endosomes

as vesicular compartments through which AAV type 2 (AAV2) traffi cs prior to nuclear entry in HeLa cells However, the precise endosomal traffi cking pathway of other AAV serotypes is not as well established We hypothesized that differences in capsid structure would result in endocytic traffi cking pathways for AAV type 1 and

5 that were fundamentally distinct from that of AAV2 In order to compare endocytosis and traffi cking pathways of AAV1 and AAV5 to AAV2, we labeled these AAV serotypes with different fl uorochromes, allowing for simultaneous imaging of two different serotypes in cells HeLa cells were co-infected with Alexa 488-labeled AAV1

or AAV5 along with Alexa 568-labeled AAV2 Cells were fi xed at various time points up to 60 minutes post-infection and analyzed

by confocal fl uorescence microscopy Co-localization analysis showed that the most overlap between AAV2 and AAV5 fl uorescence occurred during the fi rst 20 minutes of infection, while the traffi cking pathways diverged at later time points In contrast, there was minimal overlap between AAV1 and AAV2 during both early and late stages

of infection In addition, we noted stark differences in the rate of AAV traffi cking from the plasma membrane to the nucleus AAV1 entered cells within the fi rst 5-10 minutes of infection and rapidly traffi cked to the perinuclear space, whereas AAV2 and AAV5 entered more gradually, and did not accumulate in the perinuclear space until

60 minutes post-infection Using quantitative PCR, we compared the number of viral genomes in the nucleus and cytoplasm following infection Consistent with the results from confocal imaging suggesting rapid traffi cking to the nucleus for AAV1, we found that

a greater percentage of AAV1 genomes were present in the nucleus compared to the other two serotypes However, we found that at the same multiplicity of infection (MOI), the number of AAV1 genomes endocytosed into Hela cells was signifi cantly less than that for AAV2

or AAV5 We also noted signifi cantly lower levels of transduction with AAV1 compared to AAV2 and AAV5 These fi ndings show that barriers to viral entry, intracellular traffi cking, and transgene expression may be different for each AAV serotype In conclusion, the results of this study suggest that distinct intracellular traffi cking pathways for AAV1, 2, and 5 exist in HeLa cells We propose a model

by which the utilization of different receptors determines the specifi c pathway and rate of AAV traffi cking to the nucleus Understanding the determinants in the viral capsid that direct viral movement through various steps in the infectious process may enable construction of virions that more effectively express their encoded transgenes