Gene Therapy (2002) 9, 46–52  2002 Nature Publishing Group All rights reserved 0969-7128/02 $25.00 www.nature.com/gt RESEARCH ARTICLE engraftment of autologous transduced with a lentiviral flap vector: an approach to complement brain dysfunctions

E Mordelet1, K Kissa2, C-F Calvo3, M Lebastard4, G Milon4, S van der Werf1, C Vidal1 and P Charneau5 1Unite de Genetique Moleculaire des Virus Respiratoires, Institut Pasteur, Paris, France; 2Unite d’Embryologie Moleculaire, Institut Pasteur, Paris, France; 3INSERM U114, College de France, Paris, France; 4Unite d’Immunophysiologie et de Parasitisme Intracellulaire, Institut Pasteur, Paris, France; and 5Groupe de Virologie Moleculaire et Vectorologie, Institut Pasteur, Paris, France

Transplantation of ex vivo gene-corrected autologous cells expression remained stable for 1 month without cytopathic represents an attractive therapeutic approach for brain dis- effect. In vivo, transplants of transduced macrophages into eases. Among the cells of the , brain the striatum of adult rats exhibited long-term expression of macrophages are promising candidates due to their role in GFP up to 3 months. Transduced macrophages were tissue and their implication in several neuro- observed around the brain injection site and exhibited the logical diseases. Up to now, gene transfer into macrophages brain / phenotype. There was no sig- has proven difficult by most currently available gene delivery nificant sign of around the graft. These results methods. We describe herein, an efficient transduction of rat confirm the potential of lentiviral vectors for efficient and bone marrow-derived and brain macrophages with an HIV- stable ex vivo transduction of macrophages. Moreover, 1-derived vector containing a central DNA flap and encoding transduced autologous macrophages appear as a valuable the GFP reporter gene (TRIP-⌬U3-GFP). In primary cultures vehicle for long-term and localized gene expression into of macrophages our results show that more than 90% of the the brain. cells were transduced by the TRIP vector and that GFP Gene Therapy (2002) 9, 46–52. DOI: 10.1038/sj/gt/3301591

Keywords: macrophages; lentiviral vector; brain; ex vivo gene therapy; rat

Introduction phages into the brain.9,11–14 In order to target the thera- peutic action of genetically modified macrophages on The transfer of therapeutic genes into CNS-homing cells brain tissues, we have used an alternative approach by is a promising approach for brain therapy. Numerous cell injecting transduced autologous macrophages directly types such as neural progenitors, , fibroblasts into the brain. and immortalized cells have been genetically modified Transduction of cells must combine efficient gene 1–6 for their further transplantation into the rat brain. For delivery and long-term gene expression without example, in a rat model of Parkinson’s disease, the intra- inflammatory process. Herpes simplex and adenoviral striatal graft of dopaminergic embryonic neurons vectors provide efficient cell transduction, but their use expressing the human Cu/Zn superoxide dismutase may be restricted in vivo due to their inherent immuno- gene, was found to improve cell survival and to correct genicity.15,16 Recently, Costantini et al17,18 have shown that 7,8 motor symptoms partially. In the same model, the graft a herpes simplex virus/adeno-associated virus amplicon of astrocytes expressing tyrosine hydroxylase signifi- vector injected directly into the brain allowed stable 1 cantly improved motor deficits. expression of the transgene without inflammatory pro- Other CNS diseases like peroxisomal or abnormal lyso- cess. Other studies have used retroviral vectors derived somal storage are known to involve macrophage from Moloney murine leukemia virus19,20 or lenti- 9,10 defects. In the murine Gaucher disease, it has been viruses.21,22 The latter vectors are valuable tools due to reported that glucocerebrosidase-transduced mono- their ability to transduce non-dividing cells, such as nuclear phagocytes injected intravenously migrated into neurons and macrophages.21–24 9,11 the brain. However, only partial improvement of the We have reported that the insertion of the DNA flap neurological deficits were observed, due to the limited sequence into a previously described HIV-1 derived vec- penetration of systemically delivered transduced macro- tor22 strongly stimulates in vitro transduction.25,26 The high efficiency of gene transfer is provided by the triple- stranded DNA flap sequence which increases the nuclear Correspondence: E Mordelet, Groupe de Virologie Moleculaire et Vectorologie, Institut Pasteur, 25–28 rue du Dr Roux, 75724 Paris Cedex import rate of the TRIP vector DNA into the nucleus of 15, France target cells. Received 5 March 2001; accepted 1 October 2001 In the present study, we show that a self-inactivating Brain grafts of transduced macrophages E Mordelet et al 47 TRIP vector, expressing the GFP reporter gene, trans- (2.5 × 107 cells/ml) were injected into the brain. The GFP- duced macrophages derived from bone marrow (BMDM) positive BMDM or bM cells were injected into the right or embryonic brain (bM) in an efficient and stable man- striatum, whereas control untransduced cells (Figure 3a ner in vitro. In vivo, transplants of transduced macro- and c (BMDM); e and g (bM)) or PBS were injected into phages into the striatum of adult rats showed long-term the contralateral striatum. Animals were killed at 5, 15, expression of GFP up to 3 months without any significant 30 and 90 days after transplantation and brain sections sign of astrogliosis. Thus, such transduced macrophages were analyzed by confocal microscopy as described in appear as particularly attractive vehicles for long-term Materials and methods. gene expression into the brain. For each transplanted cell type, the GFP fluorescence was detected around the injection site at day 5 (Figure 3b and f), and at day 15 and 30 (data not shown). At day Results 90, GFP-positive cells surrounding the injection site were still present, thus indicating sustained GFP expression Highly efficient gene transfer into primary cultures of and long-term cell survival, together with no or limited brain (bM) or bone marrow-derived macrophages cell migration (Figure 3d and h). It is noteworthy that at (BMDM) day 90 as compared with day 5 most of the transplanted Non-dividing terminally differentiated bM and BMDM bM cells exhibited an intense ramified phenotype as were transduced with a TRIP vector deleted in the U3 shown on Figure 3h. The BMDM adopted the same region of the LTR and expressing the GFP reporter gene phenotype, but the ramifications were less extended than 27 under the control of the CMV internal promoter. TRIP for the bM (Figure 3d). vector particles were pseudotyped using the VSV-G envelope protein.28 Three, 10 and 15 days after transduc- Immunohistochemical analysis of transplanted cells tion, FACS analysis was performed to assess gene trans- The phenotype of transplanted BMDM or bM cells was fer by detection of GFP fluorescence. We observed that characterized using ED1 and Mac-1 antibodies. At day 5, more than 90% of both BMDM (Figure 1a) or bM (Figure we observed that bM expressing GFP were both ED1 and 1b) were transduced using a concentration of TRIP vector Mac-1 positive (Figure 4a and b). Additional ED1 and particles corresponding to as little as 30 ng/ml of p24. Mac-1-positive cells were also observed which might cor- Pseudotransduction of GFP was assayed by treating cells, respond to resident macrophages or monocytes that have ␮ before the transduction, with 1 m nevirapine infiltrated the tissue as a consequence of a local inflam- (Boehringer Ingelheim, Ridgefield, CT, USA), an HIV-1 matory process. reverse transcriptase inhibitor. The weak signal observed At day 90, all the GFP-expressing cells were strongly 3 days after transduction in the presence of nevirapine Mac-1-positive (Figure 3d), but the ED1 staining became undetectable 15 days after transduction, whereas decreased dramatically (Figure 3c). This observation sug- a strong GFP signal was still detected in the absence of gests that the resting ramified phenotype adopted by the nevirapine up to 1 month after transduction (data not bM cells at day 90 is the result of the integration of the shown). The sustained GFP expression in vitro and the cells in the brain microenvironment. Moreover, no longer cell viability for up to 1 month after transduction was local inflammation process was detected indicating the indicative of the absence of lentiviral vector toxicity in presence of local and transient monocytes/macrophages the primary BMDM or bM cultures. infiltration due to the inflammation caused by the injection needle. Characterization of the population of transduced cells For BMDM cells, similar results were obtained both at The purity of primary cultures of BMDM and bM cells day 5 and at day 90 after transplantation (data not was assessed using macrophage monoclonal antibodies. shown). ED1 recognizes a single chain glycoprotein expressed predominantly on the lysosomal membrane of macro- Assessment of the astrocytic gliosis phages and Mac-1 labels, a cell surface protein present on In numerous brain transplantation studies in animals, resident or activated macrophages. As shown in Figure 2, local astrocytic gliosis is a well-known phenomenon whole cells were stained with ED1 and Mac-1 antibodies, which needs to be monitored for gene therapy appli- indicating that they belonged to the mononuclear phago- cations. We therefore examined each transplanted brain cyte lineage. Almost all cells expressed GFP when section for astrocytes reactivity as an assessment of local observed by confocal microscopy, thus confirming the astrocytic gliosis using the GFAP astrocytic marker. highly efficient gene transduction by the TRIP vector. The Enhanced GFAP staining was observed around the injec- BMDM cells exhibited a rounded cell morphology with a tion site 5 days after PBS injection (data not shown), as large cytoplasm (Figure 2a and b), while bM cells showed well as after the graft of untransduced (Figure 5a) or highly branched extensions (Figure 2c and d). Both cell transduced cells (Figure 5b). Extensive GFAP staining population labeled with ED1 (Figure 2a and c) and was no longer detectable at days 15 and 30 (data not Mac-1 (Figure 2b and d) antibodies were strongly GFP shown), nor at day 90 (Figure 5c and d). Therefore, the fluorescent. local and transient astrocytic gliosis can be interpreted as the consequence of the mechanical trauma caused by the Long-term survival of transplanted cells into the brain injection needle. The highly efficient transduction of primary macro- phages allowed us to investigate the effects of intracer- Discussion ebral engraftment of macrophages into the rat striatum. In order to minimize the volumetric constraints of injec- In the present study, we showed that an HIV-1-derived tions into the brain, 2 ␮l of a concentrated cell suspension TRIP vector, including the triple-stranded DNA flap,

Gene Therapy Brain grafts of transduced macrophages E Mordelet et al 48

Figure 1 FACS analysis of TRIP-̅U3-GFP vector transduction efficiency in primary bone-marrow derived (BMDM) or brain (bM) macrophages. Primary BMDM (a) or bM (b) macrophages were not transduced (A) or transduced (C) with the TRIP-⌬U3-GFP vector in the presence (B) or in the absence of nevirapine (C) as described in Materials and methods. They were subjected to FACS analysis 3, 10 and 15 days after transduction.

allowed highly efficient ex vivo transduction of primary adaptation of the transduced cells to the brain tissue macrophages derived from bone marrow, as well as from environment. fetal brain, without cytopathic effect. In vivo, following Among currently available methods of gene delivery, injection of the transduced macrophages into the rat stria- lentiviruses represent a valuable tool for genetic engin- tum, stable transgene expression was observed from 5 to eering of cells.29,30 We have recently established that the 90 days. At day 5, GFAP staining revealed local astro- insertion of the DNA flap sequence, a determinant of len- gliosis around the injection site, which was no longer tiviral genome nuclear import, into a previously detectable at days 15, 30 or 90. At day 90, transplanted described HIV-1 vector22 increases transduction cells exhibited an intense ramified phenotype in accord- efficiency.25 As demonstrated by this study, the high ance with the absence of inflammatory process or astro- transduction rate of macrophages make them attractive gliosis. Moreover, grafted macrophages remained vehicles for autologous transplantation into the brain. localized around the injection site. These results are Macrophages play a crucial role in many brain patho- indicative of the long-term survival and successful logies including neurodegenerative and neurological dis-

Gene Therapy Brain grafts of transduced macrophages E Mordelet et al 49

Figure 2 Immunocytochemical characterization of primary bone marrow derived (BMDM) or brain (bM) macrophages transduced with the TRIP- ⌬U3-GFP vector. Primary BMDM (a and b) or bM (c and d) macrophages were subjected to immunocytochemistry with ED1 (a, c) or Mac-1 (b, d) antibodies 3 days after transduction. Fluorescence was analyzed on a con- focal microscope using 488 or 543 wave lengths for GFP and ED1 or Mac-1 staining, respectively. Scale bar = 20 ␮m.

eases.31,32 Repopulation of the CNS mononuclear system is a critical issue in the treatment of lysosomal storage diseases by bone marrow transplantation.9,11,31,33 Recent efforts in the treatment of these diseases have focused on the concept of gene therapy, in which the patient’s genetically deficient bone marrow cells are transduced in vitro and reinjected by the systemic route to the subject.33 Despite substantial progress towards this goal, ex vivo gene therapy of the brain using such an approach has proven to be limited by the poor penetration of macro- phages through the blood–brain barrier and by their slow turnover into the brain. Direct injection into the brain of stably transduced macrophages would represent an alter- native way for intracerebral functional complementation. Such an approach is in progress, acccording to the recent reports of successful brain therapy using direct injection Figure 3 GFP expression by transplanted bone marrow-derived (BMDM) of lentiviral vector delivering GDNF or intracerebral neu- and brain (bM) macrophages into the rat striatum. Rats were grafted with ron complementation in neurodegenerative diseases.34,35 not transduced BMDM (a, c) or bM (e, g) or with TRIP-⌬U3-GFP trans- As opposed to what has been described in the case of duced BMDM (b, d) or bM (f, h). Brain sections were analyzed for GFP 36 fluorescence by 3-D view on a confocal microscope (10 ␮m) at 5 (a, b, e, f) transplantation, we observed that macro- × phages did not migrate at a great distance from the injec- or 90 days (c, d, g, h) after intracerebral engraftment. Magnification 40. tion site. This feature is important for targeting specific brain areas in pathologies where the genetic delivery sys- tem must be localized such as in Parkinson’s, Tay-Sachs Materials and methods or Sandhoff’s diseases.31 As shown here, modified-macro- phage transplants could provide local delivery of the Primary cultures of brain-derived macrophages (bM) gene of interest in the target CNS structure. and bone marrow-derived macrophages (BMDM) Gene therapy methods using macrophage transplants Primary mixed brain (bM) cell cultures were generated offer several advantages. Macrophages are an easy and from cerebral cortices and striata of 17-day-old (E17) Wis- abundant cell source and can be efficiently transduced tar rat embryos (IFFA CREDO, L’Arbresle, France) as pre- with HIV-1 derived TRIP vectors. viously described.37 Briefly, the cell suspension was cul-

Gene Therapy Brain grafts of transduced macrophages E Mordelet et al 50 cells/ml). Cells were left to adhere for 3 days before being transduced. Bone marrow cells were collected from femurs and tibias of old male Long-Evans rats (Janvier, Le Genest-St- Isle, France) as previously described.37,38 Briefly, cells were flushed out with a 25-gauge needle into ice-cold phosphate-buffered saline (PBS) without Mg++,Ca++. The marrow plugs were centrifuged (1500 g, 10 min, 4°C). After dissociation of the cells and elimination of red cells, bone marrow cells were seeded into six-well dishes (4 × 106 cells/ml; Costar, Cambridge, MA, USA) in RPMI 1640 (Seromed, Berlin, Germany) supplemented as above and containing 5% of a conditioned medium from L929, as a source of CSF-1.39 At day 3, adherent and differentiated BMDM were then used for transduction. For immunofluorescence studies, glass coverslips were placed into the wells of culture plates. For in vivo trans- plantation, the cells were detached with trypsin-EDTA Figure 4 Immunohistochemical characterization of bM transduced with ⌬ 1X, washed and resuspended in serum and CSF-1-free the TRIP- U3-GFP vector and transplanted into the striatum. Brain sec- × 7 tions were analyzed at 5 (a, b) or 90 days (c, d) after transplantation using medium at 2.5 10 cells/ml. ED1 (a, c) or Mac-1 (b, d) antibodies. Fluorescence (488 wave length) and immunocytochemistry labellings (543 wave length) were analyzed by 3- Lentiviral vector production D view (10 ␮m) with a confocal microscope. Magnification ×40. Lentiviral vector stocks were produced as previously described,22 with minor modifications.25 P8.2 was used as packaging plasmid.22 Vector particles were pseudotyped with the VSV envelope protein using pCMV-G expression plasmid.28 Vector RNA genomes were pro- vided by co-transfection of pTRIP-⌬U3-CMV-GFP, a self inactivating version of pTRIP.27

Cell transduction and quantitation of gene transfer efficiency In vitro transduction experiments were conducted in six- well plates. BMDM and bM cells were transduced at a vector particle concentration corresponding to 30 ng of p24 per milliliter. Transduction efficiency was evaluated using flow cyto- metry. GFP fluorescence was measured in BMDM and bM on a FACS and analyzed with the CellQuest software (Becton-Dickinson, Le Pont de Claix, France). Pseudo- transduction of GFP was quantified after transduction of cells treated by 1 mm nevirapine, a non-nucleosidic HIV- 1 reverse transcriptase inhibitor. After suitable gating, 10 000 events were collected for each experiment. R1 rep- resents the mononuclear phagocyte population selected except and dead cells. In FL1-height GFP autofluorescence activity was gated while FL2-height counted Mac-1/TRITC labeled macrophage/microglial cells. In plots the black bars were defined from the dou- ble-negative population for FL1 and FL2-height. Statistics plots determined the rate of gated cells and fluorescence Figure 5 Assessment of astrocyte reactivity after transplantation of bM cells. Rats were grafted into the striatum with not transduced bM (a, c) intensity in the four different quadrants (LL, low left; LR, or TRIP-⌬U3-GFP transduced bM (b, d). Brain sections were analyzed low right; UL, up left; and UR, up right). at 5 (a, b) or 90 days (c, d) after transduction by immunohistochemical labelling of the GFAP antibody. Fluorescence was evaluated using a flu- Intracerebral macrophage transplantation orescence microscope (magnification ×10). Autologous engraftments were performed using bM or BMDM transplanted into the brain of Wistar (n = 20) or Long-Evans (n = 20) rats. Male adult rats (250–300 g) were tured in DMEM supplemented with 2 mml-glutamine, anesthetized with sodium pentobarbital (50 mg/kg; San- ␮ 20 mm NaHCO3, 5 U/ml penicillin, 50 g/ml strepto- ofi, France), and positioned in a stereotactic frame as pre- mycin and 10% heat-inactivated fetal calf serum. All cell viously described.40,41 Holes were drilled in the appropri- culture reagents were from Life Technologies (Cergy ate locations and a 10-␮l Hamilton syringe was used to Pontoise, France). Floating bM were isolated from a 2- inject 2 ␮l of cell suspension into the striatum (AP +1.8; week culture, by shaking for 30 min at room temperature, L ±2.6; DV Ϫ 4.5) at a slow time course (0.2 ␮l/min). At and replated on to uncoated six-well dishes (106 the cessation of the injection, the needle was left in place

Gene Therapy Brain grafts of transduced macrophages E Mordelet et al 51 for 15 min before withdrawal. GFP-transduced macro- primary fibroblasts genetically modified to produce L-dopa. phages were injected into the right striatum and control Neuron 1991; 6: 371–380. untransduced cells or PBS were injected into the left stria- 7 Nakao N et al. Overexpressing Cu/Zn superoxide dismutase tum. No behavioral abnormalities in daily activities or enhances survival of transplanted neurons in a rat model of Par- changes in body weight could be observed following kinson’s disease. Nat Med 1995; 1: 226–231. 8 Barkats M et al. Intrastriatal grafts of embryonic mesencephalic transplantation over the study period. rat neurons genetically modified using an adenovirus encoding For each type of engraftment, five rats were perfused human Cu/Zn superoxide dismutase. Neuroscience 1997; 78: intracardially with 4% PFA at day 5, 15, 30 and 90, 703–713. respectively. were removed, dipped in 20% 9 Krall WJ et al. Cells expressing human glucocerebrosidase from sucrose solution, frozen in Tissue-tek OCT (Miles, a retroviral vector repopulate macrophages and central nervous Elkhart, IN, USA) and stored at Ϫ80°C. system microglia after murine bone marrow transplantation. Blood 1994; 83: 2737–2748. Immunocytochemical procedures 10 Steven A. Origin of macrophages in the central nervous tissue. In vitro, BMDM and bM were washed, fixed in 4% PFA J Neurol Sci 1993; 118: 117–122. for 10 min at room temperature and rinsed with PBS. The 11 Krivit W, Shapiro EG, Lackman LA. Microglia: the effector cell for reconstitution of the central nervous system following bone cells were then treated with a blocking solution contain- marrow transplantation for lysosomal and peroxisomal storage ing 10% goat serum and 0.2% Triton X-100 in PBS. For diseases. Cell Transplant 1995; 4: 385–392. characterization of macrophages, the cells were incubated 12 Walker SU, Dobrennis K, Huang M. Bone marrow transplan- with mouse IgG anti-rat ED1 or Mac-1 (Serotec, Oxford, tation corrects the enzyme defect in numerous of the central ner- UK, diluted 1:100 in blocking solution).42,43 Next, the sec- vous sytem in a lysosomal storage disease. Proc Natl Acad Sci ondary antibody Cy3-conjugated affiniPure goat anti- USA 1994; 91: 2970–2974. mouse IgG (Beckman Coulter, Roissy, France) diluted 13 Yeager AM, Hart C, Pardoll DM. Repopulation by donnor- 1:500 in blocking solution was added to the cells. The derived macrophages in the murine central nervous system glass coverslips were mounted in Fluosave Reagent (CNS) after congenic bone marrow transplantation (BMT): a (Calbiochem, Meudon, France) and stored at Ϫ20°C. quantitative study. Blood 1992; 80 (Suppl. 1): Abstr. 269a. 14 Kennedy DW, Abkowitz JL. Mature monocytic cells enter For in situ immunohistochemistry, coronal OCT-coated ␮ tissues and engraft. Proc Nal Acad Sci USA 1998; 95: 14944–14949. brain slices (cryostat, 25 m) were rehydrated in PBS and 15 Byrnes AP, Wood MJ, Charlton HM. Role of T cells in inflam- treated for immunohistochemistry as described above. mation caused by adenovirus vectors in the brain. Gene Therapy The antibody specific for the astrocytic marker GFAP 1996; 3: 644–651. (diluted 1:400 in blocking solution, Chemicon, France) 16 Ugolini G. Transneuronal transfer of herpes simplex virus type was used to detect inflammatory process.44 1 (HSV 1) from mixed limb nerves to the CNS: I. Sequence of transfer from sensory, motor, and sympathetic nerve fibres to Confocal microscopy the . J Comp Neurol 1992; 326: 527–548. For immunohistochemical analysis, a confocal scanning 17 Costantini LC et al. Gene transfer to the nigrostriatal system by laser microscope (Zeiss, Le Pecq, France) was used for hybrid herpes simplex virus/adeno-associated virus amplicon vectors (published erratum appears in Hum Gene Ther 2000; 11: fluorescence detection. 2-D optical sections were acquired 981). Hum Gene Ther 1999; 10: 2481–2494. with the LSM 510 software (Zeiss). For double-label 18 Costantini LC, Bakowska JC, Breakefield XO, Isacson O. Gene immunofluorescence, the 488 and 543 lines were used in therapy in the CNS. Gene Therapy 2000; 7:93–109. alternating acquisition mode. 19 Baszler TV, Zachary JF. Murine retroviral neurovirulence corre- lates with an enhanced ability of virus to infect selectively, repli- cate in, and activate resident microglial cells (published erratum Acknowledgements appears in Am J Pathol 1991; 138: 1058). Am J Pathol 1991; 138: E Mordelet was supported by fellowships from the Past- 655–671. eur-Weizmann Foundation and from SIDACTION. This 20 Zachary JF, Baszler TV, French RA, Kelley KW. Mouse Moloney leukemia virus infects microglia, but not neurons even though work was supported by ANRS and AFM. We thank R it induces disease. Molec Psych 1997; 2: 104–106. Hellio and P Roux for assistance at the laser confocal 21 Naldini L. Lentiviruses as gene transfer agents for delivery to microscope which was purchased with the donation from non-dividing cells. Curr Opin Biotechnol 1998; 9: 457–463. Marcel and Liliane Pollack. 22 Naldini L et al. In vivo gene delivery and stable transduction of nondividing cells by a lentiviral vector. Science 1996; 272: 263– 267. References 23 Blomer U et al. Highly efficient and sustained gene transfer in 1 Cortez N, Trejo F, Vergara P, Segovia J. Primary astrocytes retro- adult neurons with a lentivirus vector. J Virol 1997; 71: 6641– virally transduced with a tyrosine hydroxylase transgene driven 6649. by a glial-specific promoter elicit behavioral recovery in experi- 24 Zufferey R et al. Self-inactivating lentivirus vector for safe and mental parkinsonism. J Neurosci Res 2000; 59:39–46. efficient in vivo gene delivery. J Virol 1998; 72: 9873–9880. 2 Corti O et al. A single adenovirus vector mediates doxycycline- 25 Zennou V et al. HIV-1 genome nuclear import is mediated by a controlled expression of tyrosine hydroxylase in brain grafts of central DNA flap. Cell 2000; 101: 173–185. human neural progenitors. Nat Biotechnol 1999; 17: 349–354. 26 Sirven A et al. The human immunodeficiency virus type 1 cen- 3 Gage FH, Kawaja MD, Fisher LJ. Genetically modified cells: tral DNA flap is a crucial determinant for lentiviral vector applications for intracerebral grafting. Trends Neurosci 14; 1991: nuclear import and gene transduction of human hematopoietic 328–333. stem cells Blood; 2000; 96: 4103–4110. 4 Martinez-Serrano A. Immortalized neural progenitor cells for 27 Sirven A, Ravet E, Charneau P. Enhanced transgene expression CNS gene transfer and repair. Trends Neurosci 1997; 20: 530–538. in cord blood CD34+-derived hematopoietic cells, including 5 Mulligan R. The basis science of gene therapy. Science 1993; 260: developing T cells and NOD/SCID-mouse repopulating cells, 926–932. following transduction with modified TRIP lentiviral vectors. 6 Fisher LJ et al. Survival and function of intrastriatally grafted Molec Ther 2001; 3: 438–448.

Gene Therapy Brain grafts of transduced macrophages E Mordelet et al 52 28 Yee J, Laporte P, Bouic K. A general method for the generation 36 Ridet JL, Serguera C, Zennou V. Primary astrocytes: a potent of high-titer, pantropic retroviral vectors: highly efficient infec- and safe vehicle for ex vivo gene transfer to the CN. (Submitted tion of primary hepathocytes. Proc Natl Acad Sci USA 1994; 91: for publication). 9564–9568. 37 Calvo C-F, Gelman M. Production of monocyte chemotactic pro- 29 Corbeau P, Kraus G, Wong-Staal F. Transduction of human tein-1 by rat brain macrophages. Eur J Neurosci 1996; 8: 1725– macrophages using a stable HIV-1/HIV-2-derived gene delivery 1734. system. Gene Therapy 1998; 5:99–104. 38 Tushinski RJ et al. Survival of mononuclear phagocytes depends 30 Schroers R et al. Transduction of human PBMC-derived den- on a lineage-specific growth factor that the differentiated cells dritic cells and macrophages by an HIV-1-based lentiviral vector selectively destroy. Cell 1982; 28:71–81. system. Molec Ther J Am Soc Gene Ther 2000; 1: 171–179. 39 Stanley ER. Factors regulating macrophage production and 31 Wada R, Proia RL. Microglia activation precedes acute neurod- growth. J Biol Chem 1977; 252: 4305–4312. egeneration in Sandhoff disease and is suppressed by bone mar- 40 Altmeyer R, Mordelet E, Girard M, Vidal C. Expression and row transplantation. Proc Natl Acad Sci USA 2000; 97: 10954– detection of macrophage-tropic HIV-1 gp120 in the brain using 10959. conformation-dependent antibodies. Virology 1999; 259: 314–323. 32 Persidsky Y, Zheng J, Miller D, Gendelman HE. Mononuclear 41 Mordelet E, Altmeyer R, Girard M, Vidal C. HIV-1 gp120: phagocytes mediate blood–brain barrier compromise and neu- expression and subcellular localisation in macrophages/ ronal injury during HIV-1-associated . J Leuk Biol 2000; microglia in vitro and in vivo in the rat brain. J Neurovirol 1998; 68: 413–422. 4: 360. 33 Kennedy DW, Abkowitz JL. Kinetics of central nervous system 42 Barclay AN. Different reticular elements in rat lymphoid tissue microglial and macrophage engraftment: analysis using a trans- identified by localization of Ia, TH-1 and MRC OX 2 . genic bone marrow transplantation model. Blood 1997; 90: Immunology 1981; 44: 727–736. 986–993. 43 Dijkstra CD. The heterogeneity of mononuclear phagocytes in 34 Bachoud-Levi AC et al. Motor and cognitive improvements in lymphoid organs: distinct macrophage subpopulations in the rat patients with Huntington’s disease after neural transplantation. recognized by monoclonal antibodies ED1, ED2 and ED3. Immu- Lancet 2000; 356: 1975–1979. nology 1985; 54: 589–599. 35 Kordower JH, Bloch J. prevented by lentivi- 44 Balasingam V, Collier B, Yong VW. Attenuation of astroglial ral vector delivery of GDNF in primate models of Parkinson’s reactivity by IL-10. Soc Neurosci Abstr 1995; 21: 304. disease. Science 2000; 290: 767–773.

Gene Therapy