112 Assessing Potency in Lentiviral Vectors Pseudotyped with VSV G, RD114 or FV Envelope Glycoproteins Molecular Therapy Volume 19, Supplement 1, May 2011 Copyright © The American Society of Gene & Ce[.]
RNA VIRUS VECTORS I we present a bioinformatic analysis of an altered integration pattern achieved with IN-endonuclease fusion protein The fusion protein was constructed by fusing an intron-encoded endonuclease I-PpoI C-terminal to HIV-1 IN According to our data, the fusion binds and cuts at its cellular target sequences For integration site studies we tested a non-cutting version, harboring a point mutation N119A, which abolishes the DNA-cutting properties while retaining the DNA-binding activity Third generation self-inactivating LVVs, containing fusion proteins either with wild-type IN or its catalysis deprived mutant D64V, were then produced in 293T cells by standard cotransfection of four plasmids, resulting in ∼10^9 and ∼10^8 TU/ ml titers, respectively, with GFP-transgenes With these and control vectors, HeLa, HEK 293 and MRC-5 cells were transduced, and genomic DNA-areas flanking the integration sites were sequenced by both traditional Sanger sequencing and high-throughput methods Comparing with wtIN vectors and published integration site data, our IN-fusion protein containing LVVs were found to have a reduced tendency to integrate into gene-dense chromosomes and disparities with regard to genotoxicity-related as well as certain other genetic elements, suggesting safer target site profile in terms of integration derived genotoxicity 110 Development and Characterization of Targeted Lentiviral Vectors Pseudotyped with the Tupaia paramyxovirus Glycoproteins Theresa Enkirch,1 Sabrina Kneissl,2 Birgit Hoyler,1 Wolfgang Stremmel,1 Christian J Buchholz,2 Christoph Springfeld.1 Department of Internal Medicine IV, University of Heidelberg, Heidelberg, Germany; 2Division of Medical Biotechnology, PaulEhrlich-Institute, Langen, Germany Background Lentiviral vectors are vectors of choice for many gene therapy applications Recently, efficiently targeted lentiviral vectors pseudotyped with the glycoproteins of the paramyxovirus measles virus (MV) have been reported However, MV antibodies in the general population might limit the clinical use of these vectors As an alternative, we propose to generate lentiviral vectors pseudotyped with the glycoproteins of Tupaia paramyxovirus (TPMV), an animal paramyxovirus with no known close relatives in humans These glycoproteins can be targeted by single-chain antibodies similar to the MV glycoproteins, but have so far not been incorporated in lentiviral vectors The Tupaia paramyxovirus does not infect human cells, so that there are probably no pre-existing antibodies against TPMV in humans Methods For optimal incorporation into lentiviral vectors, the TPMV glycoproteins were modified by truncation of their cytoplasmic tails and targeted by displaying a single-chain antibody against CD20 on the H protein These cytoplasmic tail variants were characterized by Western Blot analysis and tested for their ability to induce syncytia in CD20-positive cells Lentiviral vectors were generated by transfection of 293T cells with a transfer vector, a packaging plasmid and expression plasmids encoding the TPMV glycoproteins Transduction of target cells by these vectors was analyzed by FACS Results The modified glycoproteins were expressed to similar levels as the unmodified proteins and were able to mediate specific fusion of CD20-positive cells Cytoplasmic tail variants that were most efficiently incorporated into lentiviral vectors were identified and with these glycoproteins, titers similar to those obtained with MV-pseudotyped vectors were reached (106 t.u./ml) The resulting vectors selectively transduced CD-20-positive cells and are able to mediate specific gene transfer to primary human B-cells Conclusions We conclude that targeted lentiviral vectors pseudotyped with TPMV glycoproteins can be generated to high titers and targeted to CD20-positive cells We are currently investigating if these vectors are neutralized by human antibodies and will test their ability to mediate gene transfer in vivo S44 111 Reliability Metrics for LM/LAM/nrLAM-PCR Coupled with High Throughput Sequencing Troy B Hawkins,1 Aparna Jasti,1 Aaron Ernstberger,1 Keithanne Mockaitis,2 Kenneth Cornetta.1 Indiana University School of Medicine, Indianapolis, IN; 2Indiana University, Bloomington, IN The first step in understanding the potential impact of insertional mutagenesis in the context of retroviral gene therapy is identifying a comprehensive array of integration sites (ISs) from transduced cells Advances in sequencing technology have spurred the development of techniques to assess vector integration in high throughput Vectorgenome junction fragments produced by ligase-mediated (LM)PCR and derivative protocols (linear-amplification-mediated (LAM)PCR and non-restrictive (nr)LAM-PCR) are now routinely sequenced in parallel using DNA bar codes and 454 pyrosequencing Although this methodology is widely used, there is a lack of a comprehensive benchmark to validate the reliability of the resulting data We set out to define a series of metrics for assessing large integration datasets produced by coupling next generation sequencing to traditional capture and amplification protocols, designed to answer fundamental questions of quality and utility of these high throughput data Specificity: Utilizing filters on a large sequence dataset can identify high quality sequences that represent genomic integrations, but these represent only a fraction of the data Specificity metrics take into account the remaining sequence data: theoretical yield (TY) is a measure of expected true positives in the data; vector specificity (SV), genomic specificity (SG), and integration specificity (SI) are measures of sequence product mapping to vector LTR, any genomic sequence, and definable ISs in the host genome, respectively Stepped specificity metrics help to identify the impact of variability in sample preparation and IS distribution Sensitivity: Increasing sequencing capabilities several orders of magnitude above traditional Sanger methods also increases the probability that fragments captured in the pre-sequencing phase of integration analysis will be sequenced This assumption allows us to assess the lower limit of sensitivity of the traditional protocols for capturing and identifying ISs and utilize parallel or multiplexed internal controls in integration analysis methodologies Sensitivity metrics are imperative to understanding identification of integration events relative to the controls Results: To evaluate these metrics, we developed lentiviral-transduced K562 clones with verified ISs Clones were mixed pairwise in serial dilutions to simulate polyclonal samples We sequenced LM, LAM, and nrLAM-PCR products by parallel 454 sequencing in three separate runs and assessed specificity and sensitivity using the metrics described here We found that SV was consistently high (99.82±0.17%, n=216), but SG and SI were highly variable between individual clones (21.42±16.80% and 14.91±14.94%, respectively, n=216) In addition, we were able to directly compare performance and lower-limit sensitivity of the three protocols and identify steps in each that may introduce errors that lead to decreased reliability We believe that these standardized metrics will be key to fully understanding large sequence datasets that are now the norm, and represent a significant step towards validated controls in IS analysis protocols applied to clinical samples 112 Assessing Potency in Lentiviral Vectors Pseudotyped with VSV-G, RD114 or FV Envelope Glycoproteins Anna C Pulliam-Leath,1 Aparna Jasti,1 Helmut Hanenberg,2 Dirk Lindemann,3 Kenneth Cornetta.1 Medical & Molecular Genetics, Indiana University School of Medicine, Indianapolis, IN; 2Herman B Wells Center for Pediatric Research, Indiana University School of Medicine, Indianapolis, IN; 3Medizinische Fakultät Carl Gustav Carus, Technische Universität Dresden, Institut für Virologie, Dresden, Germany Molecular Therapy Volume 19, Supplement 1, May 2011 Copyright © The American Society of Gene & Cell Therapy RNA VIRUS VECTORS I Viral vectors are frequently pseudotyped in order to alter viral tropism The change in potency of the resulting pseudotypes requires that infectious titer is estimated very carefully Many promising pseudotypes are reported in the literature; however, direct comparison between pseudotypes has been difficult due to the lack of standardized studies This work focuses on standardizing and comparing production methods, titer methodology and virus yield for three promising envelope glycoproteins: VSV-G, feline endogenous retrovirus glycoprotein (RD114) and a modified foamy virus glycoprotein, pcoPE01 (FV), envelopes For this study, the CSCGW SIN-lentiviral vector containing CMV-GFP (6ug/T25 flask) was packaged using the pMDLg containing GAG/POL (2.2ug/T25 flask) and pRSV/ REV (1.1ug/T25 flask) plasmids Physical titer was assessed using a commercial ELISA for p24 Infectious titer was assessed by dilutional titer on HT1080 cells with the number of vector expressing cells determined by flow cytometric analysis for GFP The optimal envelope plasmid concentration for VSV-G was previously determined to be (1.5ug/T25 flask) Using the concentration of CSCGW and packaging plasmids determined to be optimal for VSV-G, infectious titer of the other envelope pseudotypes was optimized by varying the envelope plasmid concentration For RD114 a range of 0.1-5.0ug/T25 flask was evaluated and the optimal titer was produced at 1.8-2.8ug/T25 flask For FV, a range of 0.1-5.0ug/T25 flask was evaluated and the optimal titer was produced at 0.1-1.0ug/T25 flask In this initial evaluation, titers were performed with polybrene (8ug/mL) for VSV-G and RD114 and without for the foamy envelope, since investigators using the foamy envelope have titered without this polycation previously However, as the use of polycations varies among laboratories, this variable was evaluated in titer assays After a 24 hour incubation with vector (either with or without polybrene), the titer for VSV-G was determined to be 4.6X10^7 with versus 2.4X10^6 without polybrene and for RD114 5.2X10^6 with versus 2.2X10^6 without polybrene Interestingly, for FV polybrene inhibited vector transduction with the estimated titer without polybrene being significantly higher than with this agent (4.1X10^7 versus 7.9X10^6 respectively) As laboratories also vary the time of vector exposure, versus 24 hour transduction times were compared For transduction on HT1080 cells we found significant decreases in calculated titer for hour versus 24 hour vector exposure for all envelopes Our results consistently show that when producing maximal titer levels using the optimized conditions only the FV envelope is capable of providing a yield comparable to VSV-G, with RD114 yielding only 22% of VSV-G post-filtration and pre-concentration This finding has important implications when looking to scale up vector production for clinical applications 113 Tracking Retroviral-Integrated Clones with Modified Non-Restriction Enzyme LAM-PCR Technology Chuanfeng Wu,1 Alexander Jares,1 Thomas Winkler,1 Jianjun Xie,1 Andre Larochelle,1 Cynthia E Dunbar.1 Hematology Branch, National Heart, Lung, and Blood Institute, NIH, Bethesda, MD Retroviral vectors are an efficient and widely employed means of introducing an expression cassette into target cells These vectors have been shown to integrate semi-randomly into the genome, often with associated genotoxicity Consequently, the ability to characterize flanking genomic DNA of individual integration sites is critical for assessing the risk of insertional mutagenesis and for tracking individual clones to answer biologic questions Linear amplification-mediated PCR (LAM-PCR) is the best established technique for isolating proviral/host genomic DNA junctions However, LAM-PCR is subject to selection bias inherent to the assay’s use of restriction enzymes, and is further limited by an inability to discriminate between flanking genomic DNA of interest and uninformative internal vector DNA We have designed and tested for efficiency and reliability a modified Molecular Therapy Volume 19, Supplement 1, May 2011 Copyright © The American Society of Gene & Cell Therapy LAM-PCR (RE-free LAM-PCR) approach, based on a similar published technique (FLEA-PCR), which does not use restriction enzymes and that effectively selects against internal sequences Linear PCR priming does not discriminate between the transgene’s 5’ and 3’ long terminal repeats (LTRs), leading to external and internal linear PCR products In RE-free LAM-PCR, the use of restriction enzymes is circumvented by employing a mixture of oligonucleotides with a known 5’ sequence and then six random bases at the 3’ end These random hexamer-containing primers can anneal to the 3’ end of the linear PCR product, allowing for subsequent complementary strand synthesis up to and including the LTR at the 5’ end of the linear PCR product Blocking oligonucleotides anneal specifically to internal vector linear PCR products, and prevent complementary strand synthesis from progressing fully to the LTR Therefore, in the subsequent exponential PCR, vector sequences lack the LTR and not undergo amplification Ten different HIV-GFP transduced K562 clones were isolated by FACS Southern Blot analysis revealed that each clone contained a single integration site (single-copy); the site was confirmed by sequencing All 10 single-copy clones were mixed in equal cell ratios Using 100ng/reaction of starting DNA from the cell mixture, we found that one reaction of RE-free LAM-PCR retrieved vector integration sites out of 10, corresponding to 90% efficiency No internal and only 15% uninformative sequences were found using RE-free LAM-PCR, whereas internal and uninformative sequences accounted for respectively 14% and 23% using LAM-PCR The absence of internal vector sequences in RE-free LAM-PCR allows for subsequent high-throughput sequencing; detection of integration sites independently of restriction sites allows for more uniform detection with increased reaction replicates Current efforts are focused on detecting integration sites from different types of vectors, including HIV, SIV, and γ-retrovirus, as well as determining parameters for RE-free LAM-PCR sensitivity in both low and high-marking Rhesus macaque samples, followed by 454 high-throughput sequencing REfree LAM-PCR constitutes a significant improvement towards more reliable and quantitative detection of retroviral integration sites 114 GP64 Pseudotyped Feline Immunodeficiency Virus-Based Vector Transduces Murine Primary Nasal Epithelial Cells with Greater Efficacy Than Primary Tracheal Epithelial Cells Mayumi Oakland,1 Paul B McCray,1,2 Patrick L Sinn.2 Department of Microbiology, The University of Iowa, Iowa City, IA; 2Department of Pediatrics, The University of Iowa, Iowa City, IA Persistent viral vector mediated transgene expression in the airways requires delivery to cells with progenitor capacity and avoidance of immune responses to the vector or vector encoded proteins In previous studies in mouse nasal epithelia we demonstrated that repeated daily or weekly applications of GP64 pseudotyped feline immunodeficiency virus (FIV)-based vector did not elicit inhibitory adaptive immune responses Further, we observed that in vivo GP64-FIV mediated gene transfer was more efficient in the nasal airways than the large airways of the mouse lung We hypothesized that in vivo gene transfer was limited, in part, by anatomical and immunological barriers in the mouse airways Here we show that a GP64-FIV vector transduced primary cultures of well-differentiated mouse nasal epithelia with greater efficacy than primary cultures of mouse tracheal epithelia Both types of well differentiated epithelia were morphologically similar and were confirmed to contain ciliated, non-ciliated, and basal cells We further demonstrate that vector delivered to the in vivo mouse airways recruited macrophages and neutrophils; however the levels of recruitment were similar to the vehicle control Depletion of macrophages modestly improved transduction efficiency; however, the lower airway transduction efficiency was not improved in mice deficient for T and B cells or S45 ... evaluated in titer assays After a 24 hour incubation with vector (either with or without polybrene), the titer for VSV- G was determined to be 4.6X10^7 with versus 2.4X10^6 without polybrene and for RD114. .. comparable to VSV- G, with RD114 yielding only 22% of VSV- G post-filtration and pre-concentration This finding has important implications when looking to scale up vector production for clinical applications... foamy virus glycoprotein, pcoPE01 (FV) , envelopes For this study, the CSCGW SIN -lentiviral vector containing CMV-GFP (6ug/T25 flask) was packaged using the pMDLg containing GAG/POL (2.2ug/T25