762. The Impact of Architecture and Dynamics on the Outcome of Tumor Virotherapy

RNA Virus Vectors II induced pluripotent stem (iPS) cell, which required reprogramming of differentiated tissue cells by expressing multiple exogenous genes in a single cell. These vectors will also become an effective tool for large-scale protein production: the gene expression induced by the vector is quite strong as comparable as to that observed in CHO/ dhfr gene amplification system, reached up to 50 pg protein/cell/ day without the need of time-consuming amplification procedure (Nishimura, K. et al., ibid.). Altogether, these persistent SeV vectors have a number of advantageous characteristics as a tool for gene therapy, for regeneration medicine and for manufacturing protein pharmaceutics.

760. Expression of microRNAs from Lentiviral Vectors Beate K. Syttkus, Andreas Hofmann, Bodo Haas, Alexander Pfeifer. 1 Institute of Pharmacology and Toxicology, University of Bonn, Bonn, Germany.

MicroRNAs (miRNAs) are recently discovered, 20-25 nt long, noncoding RNAs targeting messenger RNAs (mRNAs) in a mechanism leading to regulation on either transcriptional or translational level. In many cases, miRNAs bind to the 3´UTR of the target mRNA at multiple sites, and regulates expression at the translational level. In contrast to small interfering RNAs (siRNAs), miRNAs only require complementarity in a small seed region (7 nt) to in induce gene silencing. Therefore, a single miRNA can regulate multiple targets and miRNAs are predicted to target one third of the protein coding genes. The number of genes regulated by these small RNAs indicates the potential for development of novel drugs as well as for the diagnosis of several human diseases including cancers. To analyse the role of miRNAs in cellular function and dysfunction systems for overexpression will be necessary. As lentiviral vectors can be used for long term expression and generation of transgenic animals, we analysed different vector configurations for optimal miRNA overexpression. Here, we analysed different lentiviral vectormiRNA expression cassettes for optimal configuration of miRNA expression. We designed cytomegalovirus (CMV) promoter-driven lentiviral vectors in three configurations carrying the pri-miRNA coding region plus flanking sequences either 5´ of a GFP coding region (CMV-miRNA-GFP) or 3´ of a GFP coding region (CMVGFP-miRNA) or without GFP (CMV-miRNA). We used these vectors to transduce cells and measured the amount of miRNAs expressed via quantitative Real-Time PCR. We chose miRNA195 and miRNA 143, which have been shown to have diverse expression patterns. Cells transduced with CMV-GFP-miRNA143 showed 15 fold increased microRNA levels compared to wildtype cells and 3 fold higher expression than the CMV-miRNA construct. Similar expression patterns were observed after transductions with the three analogous constructs carrying miRNA195 . Next, we used the best expressing vector configuration, to generate vectors carrying multiple miRNAs. Presently, we are analysing these multi-miRNA vectors as compared to transduction with lentiviral constructs carrying single miRNAs. Our data indicate that lentiviral vectors are a powerful tool for overexpression of microRNAs and that the configuration in which the miRNA are cloned strongly influences the expression levels.

761. Transduction of Primary Human Monocytes Using HIV-1-Derived Vector Particles Is Enhanced by the Viral Protein Vpx Provided by SIVsmPBjDerived Virus-Like Particles Silke Schüle,1 Björn-Philipp Kloke,1 Sabine Heidmeier,1 Julia Kaiser,1 Klaus Cichutek,1 Matthias Schweizer.1 1 Division Medical Biotechnology, Paul-Ehrlich-Institut, Langen, Germany.

The viral protein Vpx is essential for replication of SIVsmPBj in monocyte-derived macrophages (Fletcher et al., 1996). Recently, we described efficient transduction of primary human monocytes using SIVsmPBj-derived vector particles, which could not be achieved by HIV-1-derived vectors (which do not encode Vpx) (Mühlebach et al., Mol. Ther.12, 2005). We demonstrated that the Vpx protein is the sole accessory protein in PBj-derived vectors needed for efficient transduction of monocytes (Wolfrum et al., Virology 364, 2007). To check the possibility of providing the Vpx function to HIV vectors, we generated here Vpr/Vpx fusion proteins. Vpr was hypothesized to be required for fusion protein incorporation into HIV-1 vector particles, Vpx was supposed to provide the gene transfer function for monocytes. Incorporation of the Vpr/Vpx fusion proteins by SIVsmPBj vectors lacking all accesssory genes enabled transduction of monocytes, as expected. Also, all Vpr/Vpx fusion proteins were found to be incorporated into HIV-1 vector particles. However, HIV-1 vector particles incorporating the Vpr/Vpx proteins did not lead to detectable transduction of primary human monocytes. This indicated that for monocyte transduction the Vpx protein had to be delivered in the background of SIVsmPBj vector particles. To confirm this hypothesis, Vpx proteins were provided by pre-incubation of monocytes with non-transducing virus-like particles (VLPs) derived from SIVsmPBj or HIV-1. When VLPs derived from SIVsmPBj were used, gene transfer into monocytes pre-treated with Vpx- or Vpr/Vpx-containing PBj-VLPs was achieved using HIV-1 vector particles. In contrast, delivery of the Vpr/Vpx via HIV-1-VLPs did not mediate transduction of primary monocytes by HIV-1 vector particles. This confirmed the hypothesis that monocyte transduction by SIVsmPBj-derived vectors requires specific Gag or Pol functions in addition to PBj-Vpx. To determine Vpx functions required for monocyte transduction, we generated various Vpx constructs containing mutations in highly conserved amino acid residues. The localisation pattern of the new constructs was analysed by Confocal Laser-Scanning-Microscopy. VLPs were generated containing the mutant Vpx variants to identify amino acid residues that are essential for monocyte transduction by SIVsmPBj or HIV-1 vectors. The proper function of Vpx was impaired by all mutations, which was reflected by a decreased capability of monocyte transduction. We identified the tyrosine at position 69 as an essential amino acid for Vpx function, since transduction of Vpx- ala 69-VLP pretreated monocytes by SIVsmmPBj or HIV-1 vectors was completely abolished.

762. The Impact of Architecture and Dynamics on the Outcome of Tumor Virotherapy

Carlos Reis,1 Jorge M. Pacheco,1 David Dingli.2 1 CFTC and Department of Physics, University of Lisbon, Lisbon, Portugal; 2Department of Molecular Medicine, Mayo Clinic, Rochester, MN. Introduction: Replication competent viruses based on the Edmonston vaccine strain of measles virus (MV-Edm) have potent and selective activity against various types of tumor in vitro but the responses in vivo are more variable. Some tumors are eliminated consistently while others persist despite evidence of ongoing viral propagation. In order to understand these disparate results, we have developed models for the spatial growth of a tumor population followed by infection with a replicating virus that can spread by cell

Molecular Therapy Volume 16, Supplement 1, May 2008 Copyright © The American Society of Gene Therapy

S285

RNA Virus Vectors II to cell fusion ultimately leading to cell death. We utilize the model to explore the impact of tumor architecture and the dynamics of tumor cell-virus interactions on the outcome of therapy. Methods: We simulated two patterns of tumor growth: a standard spatial growth model leading to compact three-dimensional tumor shapes, and a fractal-like tumor. In its compact form, whenever a tumor cell reproduced, it sampled the surrounding lattice space to fill a compartment if this was empty or ‘push’ a cell away. In the case of fractal tumor growth, each cell division is followed by a biased random walk and the cell occupies a space close to the parent cell. In both models, tumor growth is stochastic although the dynamics of tumor growth and resulting tumor architecture are different. The virus is introduced at localized regions of the tumor, where virus infection occurs with a given probability. Infected cells can fuse with surrounding cells with a variable probability and infected cells die with a probability that is higher than that of the untreated tumor. The simulation stopped when the population of uninfected cells reached a pre-defined carrying capacity. We performed 106 simulations for each set of parameters and summarized the results. Results: The starting tumor size at which infection takes place ranged from 10 to 90% of the tumor carrying capacity; subsequently, 1% of the tumor cell population was infected through virotherapy. The three major variables that influenced outcome were: initial tumor size, the rate with which infected cells died and the tumor architecture. Smaller tumors are more likely to be eliminated by the virus, irrespective of the architecture. Tumors with fractal architecture are significantly more difficult to eradicate and in such cases, viruses that slowly kill tumor cells are more effective. For a compact tumor, the sensitivity to the location of initial infection is less important compared to fractal-like tumors, for which a central infection is associated with a higher probability of eradication. Conclusions: The outcome of tumor virotherapy is highly variable. Success or failure depends on the size of the tumor at time of therapy, tumor architecture, and the rate with which infected cells die. Smaller tumors are more likely to be eradicated. Viruses that kill cells slowly are associated with a higher probability of successful therapy, a feature which proves crucial for success in fractal-like tumors. The optimal sites of infection of the tumor also depend on the architecture with opposite results for compact versus fractal tumors.

763. Toward Gene Therapy of Cancer: A Model for Selective Expression of Anti-Life and Pro-Death Transgenes

Agnes Holczbauer,1 Adriana Zingone,2 Kritika Kachapati,3 Suresh K. Arya.3,4 1 Laboraotry of Experimental Carcinigenesis, National Cancer Institute, National Institutes of Health, Bethesda, MD; 2Genetics Branch, National Cancer Institute, National Institutes of Health, Bethesda, MD; 3Basic Research Laboratory, National Cnacer Institute, National Institutes of Health, Bethesda, MD; 4 Developmental Therapeutics Program, National Cnacer Institute, National Institutes of Health, Bethesda, MD.

To enhance effectiveness of gene therapy of cancer, we think it makes sense to target both cell survival and apoptosis pathways. As prototypes, we chose Bcl2 shRNA to down regulate cell survival and Bax transgene to up regulate apoptosis. To achieve cancer cell selectivity and yet maintain high level expression, we designed a transregulation model combining ideas from molecular virology and cancer biology. The model consists of driving the expression of Bcl2 shRNA and Bax transgene with HIV viral LTR, activate viral LTR with viral transactivator Tat, but drive viral Tat expression with cancer cell selective cellular hTERT promoter. The model requires that the LTR-gene X not be expressed without transactivation by Tat in target cells, that the hTERT promoter be able to drive Tat expression in these cells, and that Tat be able to activate LTR-gene X expression S286

in trans. We are now screening NCI 60 cell panel to test this model, with GFP as the indicator gene or gene X. The panel consists of cell lines derived from several of human cancer types, thus providing a spectrum of phenotypes. Of the cell lines tested so far, all but one met the above criteria – they are negative for GFP when transduced with the lentiviral GFP vector without an internal promoter (i.e. driven by the lentiviral LTR itself) and they display readily detectable GFP when co-transduced with the lentiviral Tat vector. One exception was that it did not support the expression of GFP from the GFP vector either with or without co-transduction with the Tat vector. This cell line could turn out to be valuable in shedding additional light on the mechanism and specificity of Tat tranactivation. Interestingly, we yet have not encountered a cell line form the panel where lentiviral GFP vector is constitutively expressed, that is cell line contains factor(s) that can complement viral Tat. While the picture might change when more cell lines and primary tumor cells are screened, so far the model holds true.

764. Effective and Stable Gene Transfer into Human NK Cells Using an HIV-1-Based Lentiviral Vector System

Su Su, Ramanathan Muthalagu, Duc M. Nguyen, Aleah Smith, Keyvan Keyvanfar, Sheila Rao, Andreas Lundqvist, Maria Berg, Richard Childs. 1 Hematology Branch, NHLBI/NIH, Bethesda, MD. Natural killer cells (NK cells) play an important role in innate immunity against tumors and viral infection. The development of an efficient method to genetically modify NK cells could be useful to genetically manipulate NK cells for therapeutic purposes and for studies characterizing NK cell tumor trafficking in vivo. Although HIV-1 based lentiviral vectors (LVs) have been used to efficiently transfer genes into human T-cells, little data exist on the use of LV vectors to transduce NK cells. In this study, we designed a variety of HIV-based LV vectors expressing enhanced green fluorescence protein (EGFP) controlled by different promoters (MSCV-LTR, sEF1a or human ubiquitin C) to transduce CD3- and CD56+ and/or CD16+ primary human NK cells. EGFP expression was measured by flow cytometry 3-4 days following LV transduction to assess transduction efficiencies (TE). Flow cytometric analysis of NK cells stained with propidium iodide revealed polybrene, an agent used to enhance TE, was highly toxic to NK cells, inducing up to 25% NK cell death after 8 hours of treatment at 8µg/mL. In contrast, exposure of NK cells to protamine sulfate at 8µg/ml for up to 16 hours enhanced LV TE without inducing obvious NK cell death. LVs with the MSCV-LTR promoter were found to mediate the most effective gene transfer to primary NK cells compared LVs with the sEF1a promoter or human ubiquitin C promoter. Using an EGFP LV driven by a MSCV-LTR promoter in the presence of protamine sulfate, we successfully transduced the natural killer cell line NKL, with long-term EGFP expression detected in up to 98% NKL cells. Using the same conditions, the following observations were made: 1) Freshly isolated NK cells were difficult to transduce, with TE in the range of only 2-17% 2) Culturing freshly isolated NK cells for 24 hours in the presence of IL-2 (500 U/mL) significantly enhanced LV TE, with 20-30% of NK cells obtained from 3 different donors expressing EGFP 3) For freshly isolated NK cells cultured in IL-2 with irradiated EBV-LCL feeder cells x 9 days, up to 55% TEs were obtained after a single or double round of transductions respectively 4) Retronectin significantly augmented LV transduction of human NK cells; NK cells cultured in retronectin-coated plates were transduced efficiently with a low MOI 5) and maintained EGFP expression and >90% viability 10 days following LV transduction 6) In contrast to un-transduced NK cells, the phenotype of LV-transduced NK cells did not change significantly . Conclusion: Retronectin significantly augments LV transduction of human NK cells with a low MOI. Using Molecular Therapy Volume 16, Supplement 1, May 2008 Copyright © The American Society of Gene Therapy