doi: /mthe , available online at on IDEAL A B FIG. 1. Efficient and sustained expression of the transgene in

Transcription

1 Maximal Lentivirus-Mediated Gene Transfer and Sustained Transgene Expression in Human Hematopoietic Primitive Cells and Their Progeny Sophie Amsellem, * Emmanuel Ravet, * Serge Fichelson, * Françoise Pflumio, * and Anne Dubart-Kupperschmitt *, Institut Cochin, INSERM U567, CNRS UMR 8104, Hematology Department, Maternité de Port-Royal, 123 Boulevard de Port-Royal, Paris, France * These authors contributed equally to this work. To whom correspondence and reprint requests should be addressed. Fax: dubart@cochin.inserm.fr. Gene therapy using permanent modifications of hematopoietic stem cells (HSC) has increasing potential applications for both genetic and acquired diseases. Considerable progress has been made recently in gene transfer to HSC by the use of lentivirus-derived vectors, which have the capacity to transduce noncycling cells. However, overall efficiency of HSC transduction reported so far is still not sufficient for numerous applications. We describe here an improved HSC transduction protocol, using the previously described lentiviral vector, that leads to more than 90% transduction of human CD34 + cells from cord blood, including NOD-LtSz-scid/scid repopulating cells. Moreover, under the same conditions, we transduce more than 75% and 80% of CD34 + cells mobilized in peripheral blood and from bone marrow, respectively. We further show that transgene expression is stable through time and hematopoietic cell differentiation in vitro as well as in vivo. Such a high HSC transduction efficiency opens new opportunities for both gene therapy applications and functional studies of regulator proteins of hematopoiesis. Key Words: hematopoietic stem cell, gene transfer, lentiviral vector, cell therapy, gene therapy, hematopoiesis regulation INTRODUCTION Gene therapy using genetic modifications targeted to hematopoietic stem cells (HSC) holds promise for the treatment of a wide variety of inherited and acquired human diseases. The potential field of applications for such an approach could be widened with the recent observations of adult stem cell plasticity. Indeed, numerous reports suggest some plasticity of adult tissue stem cells that can, in the course of in vivo regeneration, repopulate heterologous cell systems and therefore would possess a surprisingly broad spectrum of differentiation potential [reviewed in 1,6]. However, the definitive demonstration of such a potential remains to be brought since alternative explanations for transdifferentiation observations are possible [7,10]. Identification of increasing numbers of genes involved in human diseases, together with the development of new lentivirus-derived vectors, which, in contrast to widely used murine oncoretroviral vectors, can transduce noncycling cells, has resulted in significant improvements in the prospects for gene therapy. However, gene transfer efficiency must be improved before the full therapeutic potential of these vectors can be used. We have concentrated our efforts on improving gene transfer efficiency and transgene expression using a HIV-1-derived lentiviral vector and have demonstrated that this lentiviral vector was able to efficiently transduce the whole hierarchy of hematopoietic cells, including multipotent NOD-LtSz-scid/scid (NOD-SCID) repopulating cells, which are the most immature testable human stem cells [11. This vector also allows long-term sustained expression of transgenes through the ubiquitously expressed EF1α cellular promoter [12]. However, the proportion of transduced HSC obtained was still not satisfying for numerous gene therapy applications and not sufficient to address fundamental questions aimed at exploring the regulatory function of proteins by modulating their expression /02 $

2 doi: /mthe , available online at on IDEAL A B FIG. 1. Efficient and sustained expression of the transgene in transduced CD34 + and CD34 + -derived cells. (A) Histograms of EGFP expression from FACS analyses of representative transduction experiments of CD34 + cells from CB, BM, and MPB. (B) Illustrations of FACS analyses from representative experiments that show, as histograms, the distribution of lineage-marker-positive cells as a function of EGFP expression within CD61 + megakaryocytes, CD14 + monocytes, CD15 + granulocytes, GPA + erythroid cells, CD19 + B cells, CD56 + NK cells, and CD4 + T cells. The CD4 + T cells illustrated correspond to lane 5 of Table 2, and all the others correspond to lane 3 of Table 2 and derive from transduced CD34 + cells illustrated in A (left). Percentage of EGFP + cells is indicated above the histograms and dot plots representing the analysis of lineage marker versus EGFP expression in the totality of the cultured cells are shown as insets in the top right corners. Here, we describe new improvements in the HSC transduction protocol, using the same lentiviral vector, that lead to more than 90% transduction of human CD34+ cells from cord blood (CB), including NOD-SCID repopulating cells (SRC), and more than 75% and 80% transduction of CD34+ cells mobilized in peripheral blood (MPB) and from bone marrow (BM), respectively. Transgene expression was sustained with time and throughout hematopoietic cell differentiation, both in vitro and in vivo, in myeloid and lymphoid cells. These results represent a breakthrough for gene therapy applications since we demonstrate that the most frequently used therapeutic cells, that is, cells from BM and MPB, can be very efficiently transduced, which is of prime importance especially when corrected cells have no selective advantage. RESULTS AND DISCUSSION Improved Transduction Efficiency Using the New Protocol We transduced CD34 + cells purified from CB, during an period of 72 hours, with the TRIPΔU3-EF1α/EGFP vector [12] (2500 ng/ml viral P24), which was added to the cells immediately after they were seeded at 10 6 cells/ml in the presence of cytokines (stem cell factor (SCF) 100 ng/ml, Flt3-Ligand (FL) 100 ng/ml, interleukin (IL)-3 60 ng/ml, thrombopoietin-mimetic peptide (TPOmp) 10 nm; final concentrations). The same amount of vector was added 24 hours later and the cells were cultured for 48 additional hours. We then washed the cells and immediately seeded some of them in methylcellulose for 2 weeks to obtain clonogenic progenitor-derived colonies. We performed PCR studies for vector integration by selective amplification of 674

3 TABLE 1: Percentages of EGFP expressing cells after transduction of CD34 + cells from various sources of HSC % of EGFP+ cells Cord blood Bone marrow MPB ± ± 6.5 Percentages of EGFP + cells were evaluated by FACS analysis 72 hours after the end of the transduction protocol. Results were obtained with different cord blood (n = 8), bone marrow (n = 4), and mobilized in peripheral blood (MPB) (n = 2) samples. Mean ±SD are noted at the bottom of each column. the vector-encoded enhanced green fluorescent protein (EGFP) sequence on DNA extracted from cells of individually picked day 14 colonies. In three experiments, a total of 218 colonies, including BFU-E, CFU-GM, and mixedcolonies, were tested, of which 95 ± 5% were positive for vector integration. This assay, which reflects the overall transduction efficiency in all types of progenitors, including the most primitive human NOD-SCID repopulating cells [11], indicated that most of the CB CD34 + cells are transduced using this protocol. At the end of the transduction protocol, we also cocultured the cells with MS5 cells for 72 hours in liquid medium supplemented with five cytokines. At this time point, the transgene expression had reached its maximum and most of the cells were still CD34 +. We thus analyzed them by cytofluorometry for EGFP expression (Fig. 1A, left), and, using the same TRIPΔU3- EF1α/EGFP vector and eight different cord blood samples, we obtained an average of 92.5 ± 4.7% (range 86 98%) cells that express EGFP 72 hours after the end of transduction (Table 1, column 1). These improved results are in accordance with the previously observed very good correlation between vector integration and A C EF1 promoter-driven transgene expression shortly after transduction [12]. EGFP Expression in Transduced CD34 + Cells Isolated from Various Sources of HSC As most potential therapeutic applications of gene transfer will preferentially use HSC from BM and MPB cells, we evaluated the transduction efficiency using our protocol in these cells. We obtained an average of 85.5 ± 6.5% (n = 4, range 80 95%) transduction with CD34 + cells from BM, as measured by EGFP expression 72 hours after the end of the transduction protocol, and 82 and 75% with two different samples of CD34 + MPB cells (Fig. 1A and Table 1). These results most likely reflect the variability between samples and the different activation status of CD34 + cells depending on their origins [13,14]. Sustained in Vitro EGFP Expression in Transduced CD34 + -Derived CB Cells After transduction, we cultured CD34 + CB cells under conditions allowing differentiation of the various hematopoietic cell lineages. The proportions of CD34 + -derived differentiating cells that express both EGFP and differentiation markers are shown in Table 2 and illustrated in Fig. 1B. Whatever the myeloid lineage considered, the proportion of lineage-marker-positive cells that expressed EGFP remained comparable to that of the input transduced CD34 + CB cells (Table 2, column 1, lane 3; Fig. 1A, left, and Fig. 1B, top). Indeed, CD61 + MK, CD36 + /GPA + erythroid, CD15 + granulocytic, and CD14 + macrophagic B D FIG. 2. Sustained high levels of transgene expression in human cells from the BM of NOD-SCID mice transplanted with transduced CD34 + CB cells. Immediately after transduction, we injected CD34 + cells into NOD-SCID mice and tested the human cell engraftment 15 weeks later. We labeled BM cells with monoclonal antibodies directed specifically against human CD45, CD19, and CD14/CD15 markers and analyzed them by flow cytometry. Results from a representative mouse are shown. Cells were gated according to their FSC and SSC parameters (gate morphology ; A). EGFP expression was measured in the total human CD45 + population (B), in the lymphoid CD19 + B cell population (C), and in the myeloid CD14 + /CD15 + cell population (D). Percentages of cells in quadrants are indicated and the mean fluorescence intensity (MFI) of EGFP, in upper right quadrants, is given in (C) and (D). 675

4 doi: /mthe , available online at on IDEAL TABLE 2: EGFP expression in transduced CD34+ cells and their differentiated progeny % of EGFP + cells CD34 + MK Erythroid cells Granulo-monocytic Thymocytes Lympho Lympho cells cells (FTOC) B NK Total cells CD61 CD36 GPA CD15 CD14 CD4 CD19 CD ± ± 3 90 ± ± 7 88 ± 6 86 ± 3 54 ± 8 55 ± 9 After the end of the transduction protocol, EGFP expression was measured by FACS analysis on day 3 of culture for CD34 + cells, on day 7 for megakaryocytic (MK), on day 11 for erythroid and granulo-monocytic cells, on day 21 for B and NK lymphocytes and on day 28 for thymocytes. Results obtained for six independent experiments are shown. Mean ± SD are indicated at the bottom of each column., Not determined. cells expressed EGFP with means of 91 ± 3%, 90 ± 7%, 88 ± 6%, and 86 ± 3%, respectively, while the mean for input CD34 + cells was 93.5 ± 4.5% (Table 2). Interestingly, CD34 + -derived lymphoid cells expressed EGFP in lower proportions with means of 54 ± 8%, 55 ± 9%, and about 57% for T, B, and NK cells, respectively (Fig. 1B, bottom). This result might be related to a weaker expression of the transgene in lymphoid versus myeloid cells due to the presence of the 5 noncoding region of the EGFP mrna, which comprises a 5 -terminal oligopyrimidine tract (5 TOP) of the EF1 mrna leader sequence. This cis-regulating element has been described as selectively repressing translation in quiescent cells [15] like lymphoid cells. Sustained High Levels of Transgene Expression in Human Cells Recovered from NOD-SCID Mice Transplanted with Transduced CD34 + CB Cells To definitely demonstrate that the most immature human HSC had been highly transduced in our protocol and that their progeny expressed the transgene long term, we transplanted transduced CD34 + CB cells (equivalent of 10 5 day 0 CD34 + cells/mouse) into five sublethally irradiated NOD- SCID mice. Gene transfer efficiency in the transplanted CD34 + cells was 95% as measured 72 hours after the end of the transduction. We analyzed the human cell engraftment levels and transgene expression 15 weeks after transplantation in the BM and in the spleen of the mice. All five mice displayed human CD45 + cells in BM and four of five showed very homogeneous human cell engraftment with a mean of 52.1 ± 14.6% (Fig. 2B) and 31.5 ± 8.5% (data not shown) of human CD45 + cells, detected in the BM and spleen, respectively, suggesting that equivalent numbers of SRC clones had engrafted these four mice. As expected in this model, B cells were predominant, with 85% of the human CD45 + cells being CD19 +, whereas less than 12% were myeloid CD14 + and/or CD15 + cells (Figs. 2C and 2D). Importantly, 96 ± 2.4% of the engrafted human CD45 + cells expressed EGFP, demonstrating that the expression of the transgene was sustained over a long period of time (Fig. 2B). Interestingly, in contrast with what was observed with CD34 + -derived CB cells in vitro, lymphoid B cells and myeloid cells expressed EGFP in identical proportions, that is, 98 ± 2% and 94 ± 1%, respectively (Figs. 2C and 2D), ruling out the hypothesis of a preferential transduction of myeloid vs lymphoid progenitors. However, the mean fluorescence intensity of EGFP was lower in lymphoid B cells (MFI = 1600 ± 400, n = 4) compared to myeloid cells (MFI = 4000 ± 700, n = 4; Figs. 2C and 2D), as previously observed in vitro and potentially with identical explanations. Altogether, these results indicate that a 72-hour protocol that includes two additions of TRIPΔU3-EF1α/EGFP vector at 24-hour intervals in the presence of cytokines allows us to transduce and express the vector-encoded transgene in more than 90% of CB CD34 + cells and in a slightly lower proportion of CD34 + cells MPB and from BM. The transgene expression is sustained during several months in the lymphoid and myeloid progeny of the CD34 + cells. It should be emphasized that these very high transduction and expression efficiencies are also obtained with TRIPΔU3-EF1 vectors encoding HOXB4 and TAL-1/SCL proteins (manuscripts in preparation), which indicates that there is no bias coming from the use of the EGFP transgene in HSC. The protocol described here is longer than the one we previously described [12] and uses higher concentrations of cytokines, but it results in much better transduction efficiency. Furthermore, as already demonstrated under similar culture conditions [16,17], it is not detrimental to HSC potentials, based on the results of the NOD-SCID experiments presented here. MATERIALS AND S Lentiviral vector. We have previously described the lentiviral vector TRIP U3-EF1 /EGFP encoding EGFP [12]. Briefly, this vector is derived from the HR vector [18] by addition of the central polypurine tract and central termination sequence, which control the formation of the central DNA flap during the reverse transcription process of the native HIV-1 virus, and deletion of most of the U3 region of the 3 -LTR, leading to transcriptionally silent LTRs in the target cells. Moreover, the cytomegalovirus internal promoter was replaced by the ubiquitously expressed EF1 cellular promoter to improve the expression of the EGFP transgene. 676

5 Lentiviral vector supernatants. Vector particles were produced by transient calcium phosphate cotransfection of 293T cells as previously described [12]. The viral titer measured on MT4, a very permissive T lymphoid human cell line, was /ml and P24 concentration was 240 g/ml for the vector preparations using the TRIPΔU3-EF1α/EGFP vector. When an amount of vector corresponding to 2500 ng of viral P24/ml was used on 10 6 CD34 + cells/ml, this corresponded to a multiplicity of infection of 280. Collection and fractionation of hematopoietic CD34 + cells. Cord blood, bone marrow, and mobilized in peripheral blood cell samples were collected with the informed consent of the mothers or patients, according to approved institutional guidelines. We purified CD34 + cells by immunomagnetic selection (Miltenyi Biotec, Paris, France) as previously described [19] and used them either immediately or after storage in liquid nitrogen. Transduction protocol. We plated human CD34 + populations at 10 6 cells/ml in RM B00 serum-free medium (RTM, Tourcoing, France) in the presence of recombinant human (rhu) SCF (kindly provided by Amgen, Neuilly sur Seine, France), FL (Immunex, Seattle, WA, USA), IL-3 (Novartis France, Rueil- Malmaison, France), and TPOmp [20] (Genosys Biotechnologies, St Quentin en Yvelines, France) together with concentrated lentiviral vector particles. After transduction, cells were washed and cultured under different conditions. PCR analysis. Integration of the TRIPΔU3-EF1 /EGFP vector in the cellular genome was examined by PCR analysis on genomic DNA extracted from individual colony-forming cell (CFC)-derived colonies as previously described [21]. Primer sequences used to amplify part of the vector-encoded EGFP sequence were 5 -CCCTCGAGCTAGAGTCGCGGCCG-3 for the sense primer and 5 -CCGGATCCCCACCGGTCGCCACC-3 for the antisense primer. The amplification was performed for 35 cycles at an annealing temperature of 62 C, resulting in an 800-bp PCR product. Hematopoietic cell cultures. CFC were assayed as described [22]. We tested lymphoid NK and myeloid (granulomonocytic) potentials by culture on MS5 stromal cells in the presence of rhu-scf, -FL, and -TPO (50 ng/ml each) and rhu-il-2 and -15 (5 and 10 ng/ml, respectively), as described [19]. We obtained lymphoid B cells by culture of CD34 + cells on MS5 stromal cells in the presence of RPMI medium supplemented with 10% human AB serum, 5% FCS, and rhu-fl (10 ng/ml). Erythrocytic and megakaryocytic potentials were assessed under specific culture conditions. For erythroid differentiation, cells were cultured for 7 days in the presence of IL-3 (10 ng/ml), IL-6 (10 ng/ml), and SCF (25 ng/ml) and then for 4 additional days in the presence of the same cytokines supplemented with 2 U/ml Epo [23]. EGFP expression was examined by FACS analysis at the end of the culture (day 11), either in erythroid progenitors expressing the CD36 surface marker or in more mature erythroid precursors expressing glycophorin A (GPA). We obtained megakaryocytic differentiation by culture of CD34 + cells in the presence of SCF (5 ng/ml) and TPOmp (25 nm final concentration) for 7 days. We assessed human T-cell potential using NOD-SCID mouse fetal thymic organ cultures [24]. We analyzed CD34 + -derived cells by flow cytometry, after 3 (CD34 + ), 7 (megakaryocytes), 11 (erythrocytes and granulomonocytic cells), 21 (B and NK lymphocytes), and 28 days of culture (T lymphocytes), for expression of both EGFP and specific differentiation markers, using the following mouse monoclonal antibodies (MoAbs): CD19-PE (Becton Dickinson, Pont de Claix, France); CD15-, CD14-, CD36-, GPA-, and CD61-PE and CD4-PE-Cy5 (Pharmingen, Pont de Claix, France); and CD56- and CD34-PE-Cy5 (Immunotech, Villepinte-Roissy CDG, France). In all the FACS analyses, quadrants were positioned to distinguish specific from nonspecific staining according to labeling with irrelevant mouse IgG and IgM control monoclonal antibodies. Transplantation into NOD-SCID mice. NOD-LtSz-scid/scid mice were produced in the animal facilities of the INSERM U506, Paul Brousse Hospital (Villejuif, France). Transduced CD34 + cells were intravenously injected into sublethally irradiated (3.5 Gy) NOD-SCID mice immediately after the end of the transduction. Every mouse received the equivalent of 10 5 day 0 cells. Fifteen weeks later, BM and spleen cells were harvested from individual mice and the presence of human cells was detected by cytofluorometry using the following mouse anti-human monoclonal antibodies: CD45-PE (clone J.33), CD19-PE-Cy5 (clone J4.119), CD14-PE (clone RM052), and CD15-PE (clone 80H5) (all from Immunotech). ACKNOWLEDGMENTS We thank Brigitte Izac and Monique Titeux for expert technical assistance; Ibrahim Cazal and all the staff from the animal facilities of the INSERM U506 (Villejuif, France) for help in producing NOD-SCID mice; and Dr. Rouquet and the midwives from the Clinique des Noriets (Vitry-sur-Seine, France) for providing cord blood samples. We also thank Amgen for rhu-scf, Cilag for rhu-epo, and Immunex for rhu-flt3-ligand. E.R. is supported by a grant from the Ministère de la Recherche et de la Technologie and S.A. by a poste d accueil from the Institut National de la Santé et de la Recherche Médicale (INSERM). This work was supported by grants from the INSERM, Association Française contre les Myopathies (ATG ), and Association de la Recherche contre le Cancer (ARC 7551). RECEIVED FOR PUBLICATION JUNE 3; ACCEPTED AUGUST 21, REFERENCES 1. Orkin, S. H. (2000). Stem cell alchemy. Nat. Med. 6: Lowell, S. (2000). Stem cells show their potential. Trends Cell. Biol. 10: Wei, G., Schubiger, G., Harder, F., and Muller, A. M. (2000). Stem cell plasticity in mammals and transdetermination in Drosophila: Common themes? Stem Cells 18: Anderson, D. J., Gage, F. H., and Weissman, I. L. (2001). Can stem cells cross lineage boundaries? Nat. Med. 7: Blau, H. M., Brazelton, T. R., and Weimann, J. M. (2001). The evolving concept of a stem cell: Entity or function? Cell 105: Morrison, S. J. (2001). Stem cell potential: Can anything make anything? Curr. Biol. 11: R Dorshkind, K. (2002). Multilineage development from adult bone marrow cells. Nat. Immunol. 3: Orkin, S. H., and Zon, L. I. (2002). Hematopoiesis and stem cells: Plasticity versus developmental heterogeneity. Nat. Immunol. 3: Terada, N., et al. (2002). Bone marrow cells adopt the phenotype of other cells by spontaneous cell fusion. Nature 416: Ying, Q., Nichols, J., Evans, E., and Smith, A. (2002). Changing potency by spontaneous fusion. Nature 416: Sirven, A., et al. (2000). The human immunodeficiency virus type-1 central DNA flap is a crucial determinant for lentiviral vector nuclear import and gene transduction of human hematopoietic stem cells. Blood 96: Sirven, A., et al. (2001). Enhanced transgene expression in cord blood cd34 + -derived hematopoietic cells, including developing T cells and nod/scid mouse repopulating cells, following transduction with modified trip lentiviral vectors. Mol. Ther. 3: Lansdorp, P. M., Dragowska, W., and Mayani, H. (1993). Ontogeny-related changes in proliferative potential of human hematopoietic cells. J. Exp. Med. 178: Oh, I. H., Lau, A., and Eaves, C. J. (2000). During ontogeny primitive (CD34 + CD38 + ) hematopoietic cells show altered expression of a subset of genes associated with early cytokine and differentiation responses of their adult counterparts. Blood 96: Avni, D., Shama, S., Loreni, F., and Meyuhas, O. (1994). Vertebrate mrnas with a 5 terminal pyrimidine tract are candidates for translational repression in quiescent cells: Characterization of the translational cis-regulatory element. Mol. Cell. Biol. 14: Piacibello, W., et al. (1997). Extensive amplification and self-renewal of human primitive hematopoietic stem cells from cord blood. Blood 89: Kobari, L., et al. (2000). In vitro and in vivo evidence for the long-term multilineage (myeloid, B, NK, and T) reconstitution capacity of ex vivo expanded human CD34 + cord blood cells. Exp. Hematol. 28: Naldini, L., et al. (1996). In vivo gene delivery and stable transduction of nondividing cells by a lentiviral vector. Science 272: Robin, C., Pflumio, F., Vainchenker, W., and Coulombel, L. (1999). Identification of lymphomyeloid primitive progenitor cells in fresh human cord blood and in the marrow of nonobese diabetic-severe combined immunodeficient (NOD-SCID) mice transplanted with human CD34(+) cord blood cells. J. Exp. Med. 189: Cwirla, S. E., et al. (1997). Peptide agonist of the thrombopoietin receptor as potent as the natural cytokine. Science 276: Marandin, A., et al. (1998). Retrovirus-mediated gene transfer into human CD low primitive cells capable of reconstituting long-term cultures in vitro and nonobese diabetic-severe combined immunodeficiency mice in vivo. Hum. Gene Ther. 9: Issaad, C., Croisille, L., Katz, A., Vainchenker, W., and Coulombel, L. (1993). A murine stromal cell line allows the proliferation of very primitive human CD34 ++ /CD38 progenitor cells in long-term cultures and semisolid assays. Blood 81: Freyssinier, J. M., et al. (1999). Purification, amplification and characterization of a population of human erythroid progenitors. Br. J. Haematol. 106: Robin, C., Bennaceur-Griscelli, A., Louache, F., Vainchenker, W., and Coulombel, L. (1999). Identification of human T-lymphoid progenitor cells in CD34 + CD38 low and CD34 + CD38 + subsets of human cord blood and bone marrow cells using NOD-SCID fetal thymus organ cultures. Br. J. Haematol. 104: