RESEARCH ARTICLE Gene transduction efficiency in cells of different species by HIV and EIAV vectors. Introduction. Results

Transcription

1 (2002) 9, Nature Publishing Group All rights reserved /02 $ RESEARCH ARTICLE Gene transduction efficiency in cells of different species by HIV and EIAV vectors Y Ikeda 1, MKL Collins 1, PA Radcliffe 2, KA Mitrophanous 2 and Y Takeuchi 1 1 Department of Immunology and Molecular Pathology, Windeyer Institute of Medical Sciences, University College London, London, UK; and 2 Oxford BioMedica (UK) Ltd, Oxford, UK The ability of human immunodeficiency virus (HIV)- and equine infectious anaemia virus (EIAV)-based vectors to transduce cell lines from a range of species was compared. Both vectors carried the vesicular stomatitis virus G (VSV- G) envelope protein and encoded an enhanced green fluorescent protein (egfp) gene driven by a human cytomegalovirus (CMV) early promoter. Immunostaining for viral core proteins and VSV-G was used to demonstrate that the HIV and EIAV vector preparations contained similar numbers of virus particles. Various cell lines were transduced with these vectors and the transduction efficiency was estimated by measuring egfp expression. Efficient transduction by both vectors was observed in human, hamster, pig, horse, cat and dog cell lines, although EIAV vector was about 10-fold less efficient in human, hamster and pig cells normalised to the total number of viral particles. This could be partly explained by the lower RNA genome levels per particle for EIAV as measured by real-time RT-PCR. Rodent cells appeared to be transduced inefficiently with both vectors, but when the CMV promoter was substituted with the EF1 promoter in the HIV vectors, the expression level increased leading to an increase in the measurable level of transduction. (2002) 9, doi: /sj.gt Keywords: lentivirus; HIV vector; EIAV vector; host range Introduction Lentivirus-based vectors are currently under development in the field of gene therapy, since they, unlike oncoretrovirus vectors, offer opportunities for delivering genes to many types of target cells which do not or only slowly divide. Human immunodeficiency virus (HIV)- based vectors have been used in vitro and in vivo in a number of situations where terminally differentiated cells are the targets or where the physiological conditions of a target cell population leads to low mitotic indices. 1 5 Since the advanced HIV vectors contain heterologous envelope proteins, such as vesicular stomatitis virus G (VSV-G), and have been minimized by removing the accessory genes such as tat, nef, vpu, vpr and vif, replication competent virus has not been detected in these systems. Despite this disablement of HIV-based vectors, safety concerns still exist, especially if they are to be used in HIV-infected patients. As an alternative lentivirus-based approach, vectors derived from non-primate lentiviruses not known to be infectious or pathogenic for humans, such as feline immunodeficiency virus (FIV) and equine infectious anaemia virus (EIAV), have been developed and their ability to transduce non-dividing cells established Since FIVor EIAV-based vectors have minimal sequence homology to HIV and the viruses remain largely host range restric- Correspondence: Y Takeuchi, Wohl Virion Centre, Department of Immunology and Molecular Pathology, Windeyer Institute of Medical Sciences, University College London, 46 Cleveland Street, London W1T 4JF, UK Received 10 January 2002; accepted 18 February 2002 ted to their original host due to the weak transcriptional activity in human cells, 8,13 these non-primate lentivirusbased vectors may offer solutions to some of the safety concerns. However, there are few studies directly comparing infectivity of HIV- and non-primate lentivirusbased vectors. In the case of rat neural cells, EIAV vector transduction efficiencies have been shown to be comparable with HIV-based vectors. 10,12 Animal model experiments are often necessary for preclinical testing of gene therapy protocols. It is therefore useful to examine host ranges of vectors, as it is preferable if the same vector can be used in pre-clinical and clinical trials. To date the detailed host range of lentiviral vectors, especially non-primate lentiviral vectors, remains poorly characterised. In one study, HIV and simian immunodeficiency virus (SIV) vectors, which were pseudotyped with VSV-G and contained a green fluorescent protein gene driven by the vector LTR, were reported to be restricted in cells of some species. 14 In this study, we have compared efficiency of transduction of cell lines from various species by HIV and EIAV vectors containing a reporter gene driven either by an internal CMV promoter or by an internal EF1 promoter. Results Production and characterisation of HIV and EIAV vector particles HIV (HRCMVG) and EIAV (pony8.0cmvg) vectors, which contain an enhanced green fluorescent protein (egfp) gene driven by a CMV early promoter as a

2 reporter and bear VSV-G envelope proteins were prepared by three plasmid transfection in 293T cells. 3,10,15 The packaging construct, pcmvr8.9 for HRCMVG lacks accessory genes, nef, vif, vpr and vpu, while pony3.1 for pony8.0cmvg contains all EIAV accessory genes including tat, rev and S2. In order to quantitate the number of vector particles, vector harvests were spotted on glass, fixed and immunostained for virus core and VSV- G proteins as described previously. 16 The specificity of staining was demonstrated using control preparations lacking either core proteins (Figure 1a, VSV-G) or VSV- G (Figure 1a, Gag-Pol) and by the fact that -HIV CA and -EIAV antibodies did not stain EIAV and HIV vector particles respectively (data not shown). Spots in the VSV- G preparations are likely to be vesicles formed by VSV- G. 17,18 Most fluorescent spots in vector preparations were double-stained for core proteins and VSV-G (Figure 1a). The size of these fluorescent spots was estimated by filtration to be between 100 and 200 nm, similar to that of 110 nm diameter fluorescent beads (Figure 1b, beads added before filtration), suggesting these spots represent single vector particles. 16 The number of vector particles was estimated by adding fluorescent beads of known concentration to the vector preparation (Figure 1b, beads added after filtration). Several independent preparations of both HIV and EIAV vectors gave counts of particles/ml. In order to compare infectivity, two preparations each of HIV and EIAV vectors were titrated for egfp transduction on HeLa cells and the results are shown together with particle counts in Figure 1c. The titer of EIAV vector was about 10-fold lower than that of HIV vector, indicating the ability of the same number of vector particles to transduce HeLa cells was about 10-fold lower for the EIAV vector than that for the HIV vector (Figure 1c, titer/particle). These results suggest either that EIAV preparations contain a smaller number of functional particles or that the EIAV vector is intrinsically less effective in transducing HeLa cells. In the latter case, the EIAV vector might infect some other cells better than, or as well as the HIV vectors. As another comparison of HIV and EIAV vector preparations, the level of RNA genomes packaged in vector particles was examined. Cell supernatants were harvested from 293T cells transfected with egfp vector and VSV-G plasmids, as well as either a Gag-Pol plasmid (full vector) or an empty control plasmid (Gag-Pol ve). These harvests were subjected to particle counting and analysed for RNA level by real-time RT-PCR using a primer set for the egfp gene shared by the both vectors, as described previously 19 (Figure 2). Ct values for Gag-Pol-ve harvests were about 10 cycles higher than those of the corresponding full vector harvests, suggesting RNA secretion without Gag-Pol was about one thousandth (1/2 10 ) of packaged RNA. The difference in the Ct value between HIV and EIAV vector preparations was 1.4, suggesting that this HIV vector preparation had about a three-fold higher level of RNA genome than the EIAV vector preparation. The particle number of HIV and EIAV vector in this experiment was and , respectively (Figure 2b). The RNA level per particle of this particular HIV vector preparation was therefore estimated about 3.5-fold higher than that of the EIAV vector preparation. The RT-PCR assay was carried out on two separately prepared batches of vector. On both occasions the HIV gen- Figure 1 (a) Visualization of HIV and EIAV vector particles. Virus supernatants of HIV and EIAV vectors were fixed with 4% paraformaldehyde on glass slides. After permeabilisation, immunostaining was performed as described in Materials and methods. Wild-type vectors were produced by transfection of packaging, vector and VSV-G plasmids. VSV- G vectors lack packaging proteins and Gag-Pol vectors lack VSV-G. Images for green and red fluorescence were acquired separately and overlaid (merged). (b) Estimation of size and number of EIAV and HIV particles. Vector supernatants were filtered through 450-, 200-, 100-, and 20- nm pore-size filters in succession. Red fluorescent microspheres were added to the supernatant before filtration (top and middle panels) or after filtration (bottom panels) at /ml and processed for immunostaining for HIV CA or EIAV-specific proteins. (c) Stained viral particles and fluorescent microspheres from the samples filtered through 200-nm poresize filter before microsphere addition were counted in five random fields and the microsphere/viral particle ratio was measured. Results are expressed as mean values of the estimated ratios ± standard errors of the mean. egfp titers were measured on HeLa cells. 933

3 934 Figure 2 (a) Real-time RT-PCR measurement of viral RNA genome in HIV and EIAV vector preparations. Genome levels were measured by using a primer set to egfp as described in Materials and methods. Representative data of duplicated experiments are shown. Mean Ct values of duplicated experiments are shown for a threshold setting of Rn. (b) The RNA level per particle. The number of viral particles in the vector preparations used in real-time RT-PCR experiments were counted as described in Figure 1. ome levels were found to be slightly higher (1.5- to threefold). When normalised to particle numbers, these results show that the RNA levels per particle of the HIV vector preparations were approximately six-fold higher than that of the EIAV vector preparations (data not shown). The level of RNA packaging depends on the vector production protocol, as well as the vector genome structure. Since the production protocols for the vectors used in this study were only roughly optimised, there is a possibility that RNA level per particle could be improved for either vector. Differential transduction efficiency by HIV and EIAV on cell lines of different species origin Infectivity of EIAV and HIV vectors was compared on various cell lines. Transduction efficiency on human, cat and rodent cell lines is depicted in Figure 3. These cells were challenged by both vectors at a single dose of MOI 1 (MOI, multiplicity of infection based on the titer on HeLa cells), which resulted in about 60% transduction of HeLa cells in accordance with a statistical theory. 20 Viral vector transduction as measured by GFP expression in mouse and rat cells was extremely poor by both vectors, while the EIAV vector transduced cat CRFK cells more efficiently than HeLa cells and almost all CRFK cells expressed egfp. An extended panel of cell lines of different species origin was then tested for susceptibility to HIV and EIAV vectors. Cells were infected with serial dilutions of vector preparations and titres for each cell line were calculated from the data points where 5 20% of the cell population was egfp-positive. Titres standardised by physical particle number of vectors are shown in Figure 4. Some cat, dog and horse cell lines were efficiently transduced by both vectors, while the HIV vec- Figure 3 Infection of cells from different species with HIV and EIAV vectors. Equivalent amounts of the vectors ( IU on HeLa cells) were added to the indicated cells ( ). Four days later, the infected human (HeLa, HT1080 and 293), cat (CRFK), mouse (NIH/3T3 and Balb/c/3T3), rat (HSN) and Chinese hamster (CHO) cells were analysed by fluorescence microscopy. tor was more effective on human, hamster and pig cell lines. These results do not exclude the possibility that the vector preparations contain differing numbers of transduction competent particles. In particular, the RT-PCR measurement of RNA genome indicates that EIAV vector preparations used in these experiments may contain a greater number of empty particles which lack genome. However, it is clear that the intrinsic transduction efficiencies of the two lentiviral vector systems differ depending on target cell type. Improvement of expression in rodent cells by promoter substitution in HIV vectors To examine the possibility that inefficient function of the CMV promoter after successful virus entry and integration caused apparent low transduction efficiency in some cell lines, we constructed and tested an HIV vector genome, HREFG containing the human EF-1 promoter in place of the CMV promoter. All rodent cells tested showed a bigger shift of the egfp-positive population when this promoter was used, and apparent titres for HSN, NRK and NIH3T3 cells significantly increased compared with HRCMVG vector (Figure 5). These results indicated that the apparent inefficiency of HRCMVG vector to transduce some rodent cells was due to low activity of the CMV promoter. In contrast, HRCMVG vector showed stronger expression than HREFG vector in human cells (Figure 5). We also tested if promoter substitution could improve EIAV titer, by making an EIAV vector containing the human EF-1 promoter, pony8.0efg. Titers on HeLa cells and relative titers for each cell line standardised by titers on HeLa cells are shown for HIV and EIAV vectors in Table 1. As depicted in Figure 5, the rodent cells have higher GFP expression when transduced with the HREFG vector compared with the HRCMVG vector, while this effect was not detected for rabbit, fat-tailed dunnart (marsupials) or quail cells (Table 1). Similar results were observed when the CMV promoter was replaced with human -actin promoter or spleen focus forming virus LTR (data not shown). In contrast to HIV

4 935 Figure 4 Transduction efficiencies of HIV and EIAV vectors in cells from different host species. The average titer on each cell line was derived from at least two independent experiments. Figure 5 Flow cytometric analysis of the human and rodent cells infected with HIV vectors with CMV or EF1 internal promoter. The indicated cells were infected with HIVCMVG or HIVEFG and characterised 4 days after infection.

5 936 Table 1 Relative titres (%) a of HIV and EIAV vectors on various cell lines Titre on HeLa cells (IU/ml) HRCMVG HREFG pony8.0cmvg pony8.0efg Species Cells Human HeLa Cat CRFK Mouse NIH/3T L-2, M. dunni Rat HSN NRK Hamster BHK ND 43.6 ND Rabbit SIRC Cow MDBK Dunnart SC ND Quail QT % Relative titre (titre on cells shown 100)/(titre on HeLa cells). vectors, replacement of CMV promoter by EF-1 promoter in EIAV vectors did not improve titres on rodent cells (Table 1, compare pony8.0cmvg and pony8.0efg columns). Discussion HIV and EIAV vectors containing a similar egfp expression cassette using the early CMV promoter were prepared by transient, three-plasmid transfection. First, immunostaining specific for HIV and EIAV core proteins and VSV-G envelope protein was established. HIV and EIAV vector preparations had similar numbers of particles that were doubly stained for core and VSV-G. This technique will be useful in characterisation and quality control of vector preparations. Indeed, our HIV and EIAV vector preparations in this study were shown to have only a limited number of virus core particles without VSV-G or VSV-G vesicles without virus cores. It is interesting that VSV-G vesicles were not detected in the vector preparations, suggesting that VSV-G is incorporated into retrovirus particles more efficiently than into vesicles. The amount of RNA packaged in the HIV and EIAV preparations was shown to be similar by real-time PCR. The ratio between titre and physical particle number was estimated for both HIV and EIAV vectors to be up to 1/1000 depending on the target cells. It is noteworthy that this is similar to that of an MLV vector produced by a stable packaging cell line. 16 The transduction efficiency of the HIV and EIAV vectors were tested on various cell lines from different species. Both vectors transduced some non-human cell lines, including mouse and rat cells relatively poorly (Figures 3 and 4). This might pose a problem in pre-clinical studies using rodent model systems with these vectors in that gene transfer may be lower in these systems than in humans. It is noteworthy that this particular HIV vector containing the internal CMV promoter is one of the most widely used lentiviral vectors in the gene therapy field. We found that one reason for the apparently low transduction efficiency of rodent cells with HIV vector was due to the low expression from the CMV promoter, as HIV vectors with an EF1 promoter showed higher transduction levels of murine and rat cells at an efficiency close to that on human cells with a similar level of transgene expression. These results suggest that transduction efficiency as measured by GFP expression is likely to be underestimating the level of integration. This makes the choice of the internal promoter critical for animal model experiments using lentiviral vectors. As the main aim of gene transfer in gene therapy is to produce transgene products, we adopted a method to measure a gene product to assess vector efficacy in this study. In contrast, accurate measurements of vector integration by Southern blot analysis or quantitative PCR would be useful to assess the vector efficacy in gene delivery. Naturally, GFP transduction scores by FACS analysis tend to be lower than those by real-time PCR assay for provirus numbers on cells which have smaller FACS shift upon GFP transduction (PAR and KAM, unpublished data). The function of the internal promoter seems to be affected by the vector construct as a whole because CMV-eGFP expression cassettes in HIV, EIAV and MLV vector contexts express different levels of egfp (data not shown). Promoter interference between viral LTR and internal promoters may be possible and an important factor. However, it is true for many vectors that CMV internal promoter express poorly in many rodent cells compared with human cells. The reason for inefficient HIV transduction in other species, including rabbit and cow is less clear. Like other studies, 14 our results on rabbit cells suggested that there is a block to HIV infection at a post-entry, pre-integration step, as the vectors with EF1 promoter were still low titer on rabbit SIRC cells, though the few egfp positive cells were brightly fluorescent (data not shown). Replacement of the CMV promoter in EIAV vectors did not significantly improve their performance and further investigation will be required to establish whether the inefficient transduction of rodent and some other poorly transduced cell types is due to low levels of transgene expression or integration. Both EIAV and HIV vectors appeared to be able to efficiently transduce cells of many species including humans. EIAV vectors transduced human cells in general only 10-fold less efficiently than HIV, while humans are natural hosts for HIV, but not for EIAV. Further improvement of EIAV vectors could therefore make them a safer alternative to HIV vectors in human gene therapy. Fur-

6 thermore their potential for veterinary applications should stimulate additional interest. Materials and methods Cells and viruses Human HeLa, 293, 293T, HT1080, A375, TE671 and ShSy- 5y cells and mouse NIH/3T3, Bab/c/3T3, Mus dunni and L-2,3 cells, mouse-rat hybrid ND7 cells, hamster CHO, A23 and BHK cells, rat HSN and NRK cells, porcine PAE and ST Iowa cells, rabbit SIRC cells, bovine MDBK cells, equine NBL-6 cells, feline CRFK, AH927 and QN10 cells, canine A72, D17 and DK cells, dunnart SC300 cells, quail QT6 and QT35 cells were grown in Dulbecco s modified Eagle s medium supplemented with 10% (vol/vol) heat-inactivated foetal calf serum (FCS) and antibiotics. Human DG75, BC3, Jurkat and MOLT-4 cells, mouse EL4 cells, and feline FL4 cells were cultured in RPMI 1640 medium supplemented with 10% FCS and antibiotics. IL-2-dependent feline lymphoblastoid MYA- 1 cells were cultured in RPMI 1640 medium supplemented with 10% FCS, antibiotics and human recombinant IL-2 (100 U/ml). Plasmids HIV-1-based plasmids and VSV-G expressing plasmid (pmdg) were kindly provided by D Trono (University of Geneva, Geneva, Switzerland). The HIV packaging plasmid pcmvr8.9 carries gag, pol, tat, and rev genes. The HIV vector genome plasmid phr -CMV-eGFP (HRCMVG) contains a cytomegalovirus (CMV)-driven enhanced green fluorescent protein (egfp) expression cassette. phrefg contains human elongation factor 1 (EF1 ) promoter in place of CMV. The EIAV-based vector system has been published. 10 EIAV-based packaging plasmid pony3.1 carries gag, pol, tat, rev and the accessory gene S2. The EIAV vector genome plasmid pony8.0g (pony8.0cmvg) contains a CMV-driven egfp expression cassette and the pony8.0efg carries EF1 promoter in place of CMV. Virus production Viruses were produced essentially as described previously, 21 by transient transfection of 293T cells with a weight ratio of 3:2:1 of vector to packaging to envelope plasmids using Lipofectamine (Gibco, Paisley, UK) per the manufacturer s instructions. To obtain control vectors lacking any of the three plasmids, equivalent amounts of psp72 plasmid was added to the transfection mixture. Transfected cells were then washed and grown for 24 h in serum-free OptiMEM (Gibco) at 37 C and for another 24 h at 32 C. Supernatants were harvested, passed through a m filter. The viruses were then concentrated by ultracentrifugation, g for 90 min. Viruses were aliquoted and stored at 80 C before use. Virus assays To determine transduction efficiencies, a total of cells (adherent cells) or cells (suspension cells) per well in 24-well plates were inoculated with serial dilutions of the virus supernatants overnight. Numbers of infected cells were determined by measurement of egfp expression by fluorescence-activated cell sorting (FACS) using a FACScan and CELL QUEST software (Becton Dickinson, Franklin Lake, NJ, USA). Antibodies Anti-EIAV sera from experimentally infected horses were kindly provided by H Sentsui (National Institute of Animal Health, Tsukuba, Japan) and were used to stain EIAV virions. A mouse monoclonal anti-hiv-1 p24 IgG antibody EVA365 (AIDS reagents, MRC Programme) was used to stain HIV-1 virions. Rabbit polyclonal antibodies were used to detect VSV-G glycoprotein. Fluorescein isothiocyanate (FITC)-conjugated goat IgG against mouse IgG (Jackson, West Grove, PA, USA), FITC-conjugated goat IgG against horse IgG (Jackson), and Texas red-conjugated donkey IgG against rabbit IgG (Jackson) were used as secondary antibodies. Immunofluorescent staining of viral particles Immunofluorescent staining of viral particles was performed as described previously. 16 Briefly, 50 l of unconcentrated freshly harvested viral supernatants was incubated on glass slides for 1 h at 37 C in the presence of 8 g of Polybrene (Sigma, Poole, UK) per ml. Fixed virus particles were permeabilised with 0.2% Triton X-100 for 15 min at room temperature and washed three times with phosphate-buffered saline (PBS). Virus samples were incubated with anti-vsv-g antibodies and either anti- EIAV (EIAV samples) or anti-hiv-1 p24 (HIV samples) primary antibodies for 45 min at room temperature, washed three times with PBS, and incubated with the appropriate secondary antibodies for 45 min at room temperature. After being washed three times with PBS, the slides were mounted with immunofluorescence mounting medium (Dako, Carpinteria, CA, USA) and observed by confocal microscopy (MRC 1024 (BioRad, Hercules, CA, USA) equipped with a krypton-argon laser). All pictures were acquired by using Kalman filtration and analysed with Lasersharp software (BioRad). Estimation of particle size, infectious titer and physical number of HIV and EIAV vectors Viral supernatants were sequentially passed through syringe filters with the following pore sizes: 450 nm (Sartorius, Goettingen, Germany), 200 nm (Sartorius), 100 nm (Millipore, Bedford, MA, USA) and 20 nm (Millipore). After each filtration, aliquots of filtrates were examined by titer determination on HeLa cells and by virus particle immunostaining. To determine the virus particle number, 110-nm-diameter red fluorescent carboxylate-modified microspheres (FluoSpheres; Molecular Probes, Eugene, OR, USA) were added to the filtrates at a final concentration of microspheres/ml. Alternatively, the same concentration of microspheres was added to the viral supernatant before filtration, in order to confirm the retention capacity of the filters. Mixtures of virus particles and fluorescent microspheres were immobilized on glass slides and processed for immunostaining as described above. The virus particle number in the vector preparation was calculated as the product of the microsphere concentration and the ratio of counts of fluorescent dots for virus and microsphere particles. Taqman PCR Real-time PCR was carried out using the ABI Prism 7700 Sequence detection system (Applied Biosystems, Foster City, CA, USA). Viral RNA was extracted from a 10- l aliquot of viral vector (QIAamp viral RNA mini kit, Qiagen, Hilden, Germany). Five l of eluate was DNaseI 937

7 938 treated (DNA-free, Ambion) in a total of 50 l. Five l of this solution was used as template per 25 l one step RT-PCR reaction using TaqMan One-Step RT-PCR master mix reagents (Applied Biosystems, Austin, TX, USA). Duplicate reactions were carried out both with and without reverse transcriptase. After initial incubations at 48 C for 30 min and 95 C for 10 min, 40 cycles of amplification were carried out for 15 s at 95 C followed by 1 min at 60 C. Primers/probe (Applied Biosystems) were as follows: GFPF, 5 CAACAGCCACAACGTCTATATCAT 3 (300 nm); GFPR, 5 ATGTTGTGGCGGATCTTGAAG 3 (900 nm); GFPP, 5 (FAM) CCGACAAGCAGAA- GAACGGCATCAA (TAMRA) 3 (150 nm). Acknowledgements This work was supported by United Kingdom Medical Research Council and Cancer Research Campaign. References 1 Akkina RK et al. High-efficiency gene transfer into CD34+ ce lls with a human immunodeficiency virus type 1-based retroviral vector pseudotyped with vesicular stomatitis virus envelope glycoprotein G. J Virol 1996; 70: Follenzi A et al. Gene transfer by lentiviral vectors is limited by nuclear translocation and rescued by HIV-1 pol sequences. Nat Genet 2000; 25: Naldini L et al. In vivo gene delivery and stable transduction of nondividing cells by a lentiviral vector. Science 1996; 272: Kotsopoulou EA et al. Rev-independent human immunodeficiency virus type 1 (HIV-1)-based vector that exploits a codonoptimized HIV-1 gag-pol gene. J Virol 2000; 74: Kim VN, Mitrophanous K, Kingsman SM, Kingsman AJ. Minimal requirement for a lentivirus vector based on human immunodeficiency virus type 1. J Virol 1998; 72: Johnston JC et al. Minimum requirements for efficient transduction of dividing and nondividing cells by feline immunodeficiency virus vectors. J Virol 1999; 73: Wang G et al. Feline immunodeficiency virus vectors persistently transduce nondividing airway epithelia and correct the cystic fibrosis defect. J Clin Invest 1999; 104: Poeschla EM, Wong-Staal F, Looney DJ. Efficient transduction of nondividing human cells by feline immunodeficiency virus lentiviral vectors. Nat Med 1998; 4: Curran MA, Kaiser SM, Achacoso PL, Nolan GP. Efficient transduction of nondividing cells by optimized feline immunodeficiency virus vectors. Mol Ther 2000; 1: Mitrophanous K et al. Stable gene transfer to the nervous system using a non-primate lentiviral vector. 1999; 6: Olsen JC. Gene transfer vectors derived from equine infectious anemia virus. 1998; 5: Mazarakis ND et al. Rabies virus glycoprotein pseudotyping of lentiviral vectors enables retrograde axonal transport and access to the nervous system after peripheral delivery. Hum Mol Genet 2001; 10: Taube R et al. Interactions between equine cyclin T1, Tat and TAR are disrupted by a leucine-to-valine substitution found in human cyclin T1. J Virol 2000; 74: Hofmann W et al. Species-specific, postentry barriers to primate immunodeficiency virus infection. J Virol 1999; 73: Zufferey R et al. Multiply attenuated lentiviral vector achieves efficient gene delivery in vivo. Nat Biotechnol 1997; 15: Pizzato M, Marlow SA, Blair ED, Takeuchi Y. Initial binding of murine leukemia virus particles to cells does not require specific Env-receptor interaction. J Virol 1999; 73: Abe A, Chen ST, Miyanohara A, Friedmann T. In vitro cell-free conversion of noninfectious Moloney retrovirus particles to an infectious form by the addition of the vesicular stomatitis virus surrogate envelope G protein. J Virol 1998; 72: Sharma S, Miyanohara A, Friedmann T. Separable mechanisms of attachment and cell uptake during retrovirus infection. J Virol 2000; 74: Rohll JB et al. The design, production, safety, evaluation and clinical applications of non-primate lentiviral vectors. Meth Enzymol 2002; 346: Porter CD et al. Cationic liposomes enhance the rate of transduction by a recombinant retroviral vector in vitro and in vivo. J Virol 1998; 72: Neil S, Martin F, Ikeda Y, Collins M. Post-entry restriction to human immunodeficiency virus-based vector transduction in human monocytes. J Virol 2001; 75: