High-Level Transgene Expression in Primary Human T Lymphocytes and Adult Bone Marrow Cd34ϩ Cells Via Electroporation-Mediated Gene Delivery

Home , Transgene

Gene Therapy (2000) 7, 1431–1437  2000 Macmillan Publishers Ltd All rights reserved 0969-7128/00 $15.00 www.nature.com/gt NONVIRAL TRANSFER TECHNOLOGY BRIEF COMMUNICATION High-level transgene expression in primary human T lymphocytes and adult bone marrow CD34ϩ cells via electroporation-mediated gene delivery

VFI Van Tendeloo1,2, R Willems1, P Ponsaerts1, M Lenjou1, G Nijs1, M Vanhove3, P Muylaert3, P Van Cauwelaert3, C Van Broeckhoven2, DR Van Bockstaele1 and ZN Berneman1 1Laboratory of Experimental Hematology, Antwerp University Hospital (UIA/UZA); 2Laboratory of Molecular Genetics, University of Antwerp (UIA), Department of Molecular Genetics, Flanders Interuniversity Institute for Biotechnology (VIB); and 3Department of Cardiac Surgery, Middelheim General Hospital, Antwerp, Belgium

The design of effective gene delivery systems for gene stimulation. Kinetic analysis of EGFP expression revealed transfer in primary human blood cells is important both for that activated T lymphocytes maintained transgene fundamental hematopoiesis research and for cancer gene expression at high levels for a prolonged period. In addition, therapy strategies. Here, we evaluated electroporation as a fresh unstimulated BM CD34+ cells were consistently trans- nonviral means for transfection of activated human T lym- fected (5.2 ± 0.4%) with minimal cytotoxicity (Ͻ5%), even phocytes and adult bone marrow (BM) CD34+ cells. We without preliminary CD34+ cell puriﬁcation. Both T cells and describe optimal culture and electroporation parameters for CD34+ cells retained their phenotype and functional capacity efﬁcient gene delivery in prestimulated T lymphocytes (16.3 after electroporation. These results demonstrate that elec- ± 1.3%), as well as 2-day cultured adult BM CD34+ cells troporation is a suitable nonviral transfection technique that (29.6 ± 4.6%). PHA-stimulated T cells were most receptive may serve applications in gene therapy protocols using T for transfection after 48h of in vitro culture, while T cells lymphocytes or CD34+ cells. Gene Therapy (2000) 7, stimulated by CD3 cross-linking and interleukin (IL)-2 1431–1437. achieved maximum transfection levels after 72 h of pre-

Keywords: gene transfer; CD34+ cells; T cells; EGFP; ﬂow cytometry; electroporation

The design of effective nonviral gene transfer protocols is a standard technique for nonviral transfection of both for primary human hematopoietic cells, including CD34+ primary and transformed human hematopoietic cells,12,13 hematopoietic progenitor cells (HPC) and T lymphocytes, and has potential for gene therapy applications ex vivo14 is a challenging approach towards gene therapy of cancer and in vivo.15 As it is a physical transfection method, gene and infectious diseases such as the acquired immuno- transfer does not require cell-type-specific receptors, deficiency syndrome (AIDS).1–3 To date, gene transfer in unlike viral vectors. In addition, electroporation is highly primary human HPC and T cells has mainly focused on reproducible, time-saving and cost-effective. Recently, we retroviral vectors based on a replication-incompetent have demonstrated efficient transient transfection by Moloney murine leukemia virus (MuLV).4,5 However, electroporation in primary CD34+ progenitor-derived MuLV-based vectors do not efficiently transduce dendritic cells (DC) and Langerhans cells (LC).16 Here, unstimulated human HPC as they require dividing cells we want to extend our electroporation protocol to trans- for effective transduction.6 Therefore, nonviral vectors fect HPC and primary T lymphocytes. Recently, Satoh et are being explored as alternative strategies to achieve al17 demonstrated that peripheral blood CD34+ HPC short-term7 and long-term8 transgene expression in HPC. could be transfected by electroporation by EBV-based Although cationic lipids have been shown to be effec- vectors. Also, Ebert et al18 demonstrated that electropor- tive as nonviral vectors in vitro and in vivo,9 limited suc- ation of lymphocytes, unlike lipofection and receptor- cess has been obtained in liposome-mediated transfection mediated transfection, induced no significant TNF␣ pro- of HPC or primary cultures of lymphocytes.10,11 This duction correlated with induction of apoptosis. However, illustrates the need and potential utility of an alternative very little data on cell viability and transfection efficiency and efficient nonviral protocol for transient gene were provided due to the use of enzymatic bulk assays expression in HPC and primary T cells. Electroporation or PCR analysis to detect reporter gene expression. Here, we want to explore the efficiency of transfection by electroporation in primary human CD34+ cells and T lymphocytes using a plasmid vector encoding the enhanced Correspondence: ZN Berneman, Laboratory for Experimental Hematology (UIA/UZA), Antwerp University Hospital, Wilrijkstraat 10, B-2650 green fluorescent protein gene (EGFP). This EGFP Edegem, Belgium reporter system allowed optimization of the electropor- Received 26 October 1999; accepted 3 May 2000 ation procedure towards efficiency, viability and trans- Electroporation of primary T cells and CD34+ cells VFI Van Tendeloo et al 1432 gene expression on a single-cell level by flow cytometric (FCM) analysis.19 In the process of optimizing transient electroporation of hematopoietic cells, voltage and capacitance were found to be the most critical parameters since increasing voltages result in increased numbers and size of pores, while capacitance is linearly correlated with pulse time (according to the equation ␶ = R × C, where ␶ is pulse time, R resistance and C capacitance) which determines the duration of the pore-formation.20 In preliminary experiments using leukemic cell lines, we observed that leukemic cells could be successfully transfected at lower voltage pulses (200–300 V) in combination with a relatively high capacitance (1050–1500 ␮F), which is in accordance with a previous report.21 Electroporation efficiency in leukemic cell lines (Jurkat, Sup-T1, Molt-3, HSB-2, K562, KG-1, U-937 and HL-60, all obtained from the American Type Culture Collection (ATCC), Rockville, MD, USA) was in the range of 28–52%. Decreasing the volume in the electroporation cuvette or further increasing capacitance to for example 3000 ␮F led only to longer pulse times and excessive cell death (data not shown). At increasing voltages the percentage of transfected cells within the viable population increased but overall efficiency at higher voltages decreased as cell survival Figure 1 Optimization of electroporation in primary T cells. Peripheral became a limiting factor. By balancing survival rate and blood was obtained from hemachromatosis patients after informed consent electroporation efficiency – in terms of absolute numbers + and peripheral blood mononuclear cells (PBMC) were isolated by density of viable EGFP cells – optimal parameters were determ- gradient centrifugation using Lymphocyte Separation Medium (ICN ined for primary human T lymphocytes. Peripheral blood Biomedicals, Costa Mesa, CA, USA). PBMC were resuspended in AIM- lymphocytes (PBL) were prestimulated with PHA for 2 V medium (Gibco BRL, Paisley, UK) and allowed to attach to plastic ° days and electroporated at various settings. We estab- dishes in order to remove monocytes. After 2 h at 37 C, the non-adherent ␮ cell fraction was used as peripheral blood lymphocytes (PBL). Non-adher- lished that electroporation at 300 V (and 1050 F) was ent PBMC were grown in AIM-V (Gibco BRL) supplemented with 10% optimal for electroporation of activated T lymphocytes. autologous plasma unless stated otherwise and stimulated for 2 days with Absolute transient efficiency in 2-day prestimulated T 1 ␮g/ml phytohemagglutinin (PHA; Sigma). Then, PBL (Ͼ95% CD2) lymphocytes peaked at 16% and cell mortality at these were resuspended in Opti-MEM (Gibco BRL) to a final concentration of electrical settings was 70% (Figure 1). Due to this initial 30 × 106 cells/ml. Subsequently, 500 ␮l of cell suspension was mixed with ␮ toxic effect, further culture of T cells electroporated at 300 20 g of pEGFP-N1 (Clontech Laboratories) containing an enhanced green fluorescent protein (EGFP) reporter gene under the control of a V in the presence of IL-2 for 1 week resulted in a some- cytomegalovirus (CMV) promoter, incubated for 1 min at room tempera- what lower expansion (eight- to 10-fold) as compared ture and electroporated with an Easyject Plus apparatus (EquiBio). After with non-electroporated T cells (12- to 15-fold). Several electroporation, cells were immediately resuspended in fresh AIM-V attempts to electroporate unstimulated T cells resulted in medium, 10% autologous plasma supplemented with 20 U/ml IL-2 (R& undetectable EGFP expression, which emphasizes the D Systems, Minneapolis, MN, USA). Electroporation was carried out at ␮ need for a mitogenic stimulus before transfection of T a capacitance of 1050 F with varying voltages. The resulting pulse time 22 is linearly correlated with the applied capacitance (according to the equ- cells. Moreover, transfection of activated T cells by ation ␶ = R × C, where ␶ is pulse time, R resistance and C capacitance) DMRIE-C-mediated lipofection did not result in detect- and was automatically and empirically calculated after the pulse by the able EGFP expression. electroporation device. The internal resistance shunt of the electroporation In order to investigate the effect of prestimulation time device was always set on +ϱ in order to allow a maximum electrical cur- on transfection efficiency, peripheral blood lymphocytes rent to pass through the cuvette (optimal when working with highly ionic (PBL) were stimulated with PHA for 1 to 3 days (Ͼ95% electroporation buffers). One day after electroporation, transfected cells were washed once in PBS supplemented with 1% FCS and resuspended of cells bore T cell markers after prestimulation) and sub- in 0.5 ml of PBS supplemented with 1% BSA and 0.1% azide. Ethidium sequently transfected by electroporation at optimal para- bromide (EB) at a final concentration of 10 ␮g/ml was added directly meters. We obtained high-efficiency gene transfer by elec- before FCM analysis, as described previously.16 On a FACScan analytical troporation in 1- to 3-day PHA-stimulated CD2+ T cells. flow cytometer (Becton Dickinson), lymphocytes were gated based on their A representative dot plot of EGFP-transfected 1-day scatter profile and cells were evaluated for EGFP expression. Viability was PHA-prestimulated T cells is shown in Figure 2a, con- determined by ethidium bromide exclusion using FCM analysis (triangles, + + ± right y-axis). Transfection efficiency was determined 24 h after electropor- firming the presence of CD2 EGFP T cells ( 12%). ation by FCM analysis of EGFP expression and is given as the percentage Maximum transfection levels were observed after a 2-day of EGFPϩ cells (left y-axis), either as overall efficiency (black bars, % prestimulation (Figure 2b). FCM analysis revealed 16.3 ± EGFPϩ cells of total population including dead cells) or as relative 1.3% EGFP+ cells 24 h after electroporation, which was efficiency (grey bars, % EGFPϩ cells within the viable population). Data significantly higher than 1-day (9.6 ± 2.9%) and 3-day represent means of three separate experiments (s.d. Ͻ5%). (10.5 ± 1.9%) stimulated T cells. As the use of PHA should be avoided in clinical trials, we also investigated electroporation efficiency of anti- CD3/IL-2-stimulated PBL. Substitution of PHA activation by CD3 cross-linking followed by IL-2 stimulation

Gene Therapy Electroporation of primary T cells and CD34+ cells VFI Van Tendeloo et al 1433

Figure 2 Electroporation of primary human T lymphocytes. (a) Following 24 h of PHA activation, T lymphocytes were electroporated at 300 V and 1050 ␮F, as described in the legend to Figure 1, and FCM analysis of EGFP fluorescence combined with CD2-PE staining was performed 24 h after electroporation. For immunophenotypic analysis, 5 × 105 transfected T cells were washed once in PBS supplemented with 1% FCS and stained for 30 min at 4°C with 1 ␮g of monoclonal antibody in 150 ␮l of PBS containing 1% BSA and 0.1% azide and analyzed on a FACScan flow cytometer. In order to allow combination with EGFP analysis, a phycoerythrin (PE)-conjugated monoclonal CD2 antibody (Becton Dickinson) was used. To set quadrants, a nonreactive isotype-matched monoclonal antibody (Becton Dickinson) was used. A representative dot plot is shown. (b) PBL were activated with PHA or anti-CD3 for 24–72 h, followed by IL-2 stimulation. At different time-points during culture (24 h, 48 h and 72 h after prestimulation), T lymphocytes were transfected by electroporation. Transfection efficiency was determined by FCM analysis of EGFP expression and is expressed as the percentage of EGFPϩ cells 24 h after electroporation. Data are shown as mean ± s.d. of five independent experiments. (c) For kinetic analysis of EGFP expression in 2-day stimulated T lymphocytes, PBL were PHA-stimulated for 48 h and subsequently electroporated at 300 V and 1050 ␮F, as described in the legend to Figure 1. EGFP expression was monitored by FCM analysis as a function of time. Marker regions indicating the EGFPϩ cell fraction (normal lines) were set using control-transfected T lymphocytes as background controls (bold lines). Note the vast increase of the percentage of EGFP-expressing cells accompanied by a decrease of the mean fluorescence intensity in the EGFP+ population over time.

Gene Therapy Electroporation of primary T cells and CD34+ cells VFI Van Tendeloo et al 1434 resulted in a significantly lower transfection efficiency in procedure was performed on 2-day cultured CD34+ cells, 1-day (5.3 ± 2% with anti-CD3/IL-2 versus 9.6 ± 2.9% with using the same approach as described for electroporation PHA, P Ͻ 0.05) and 2-day (8.9 ± 2.5% with anti-CD3/IL- of T lymphocytes. Highest absolute numbers of viable 2 versus 16.3 ± 1.3% with PHA, P Ͻ 0.01) prestimulated EGFP+ CD34+ cells were obtained when electroporation T cells, but we observed a slight but significant increase was performed at 260 V with a capacitance of 1050 ␮F in efficiency in 3-day anti-CD3/IL-2-stimulated as com- (data not shown). As Flt-3 ligand (FL) has been reported pared with 3-day PHA-activated T cells (13.5 ± 1.3% ver- to be an early-acting growth factor for HPC,27 we first sus 10.5 ± 1.9%, respectively, P Ͻ 0.05), as depicted in investigated the effect of the presence of FL during CD34+ Figure 2b. Interestingly, transgene expression kinetics cell culture on transfection efficiency. As the presence of showed that the median number of cells positive for FL during culture significantly increased electroporation EGFP increased steadily over time, reaching almost 50% efficiency (21.7 ± 1.5% without FL versus 28.5 ± 2.1% with at day 5, while the mean fluorescence intensity (MFI) of FL; P Ͻ 0.01), FL was routinely added to all further cul- the EGFP+ population almost linearly decreased from tures. Kinetic analysis of EGFP expression in CD34+ cells 2525 upon 24 h to 85 after 1 week (Figure 2c). As IL- transfected after 2 days of culture in IL3+IL-6+SCF+FL 2-stimulated lymphocytes tend to grow in clusters, we revealed peak expression from 24 to 48 h after electropor- hypothesize that there is a high degree of passive transfer ation. Four days after gene delivery, 50% of the peak of EGFP in these clusters upon cell division. Although expression was still present in the cells. Subsequently, this phenomenon warrants further investigation, our there was a steady decline to background levels 1 week results indicate that electroporation of primary activated after transfection (Figure 3). Figure 4a shows a represen- human T lymphocytes results in relatively high levels of tative FACS profile of 2-day stimulated CD34+ cells. Table transgene expression for an extended period of time in a 1 shows the outcome of electroporation of cultured substantial portion of cells. CD34+ cells at different time-points in more detail. Trans- In order to verify that electroporation had no adverse fection of 2-day cultured CD34+ cells was most optimal effects on the function of T cells, we electroporated an with regard to efficiency (29.6 ± 4.7% at day 2 versus 11.3 IL-2-dependent CD8+ tumor infiltrating lymphocyte (TIL) ± 1.0% at day 1; P Ͻ 0.001) despite the significant lower + 1235 clone recognizing the HLA-A2-restricted Melan25–37 survival rate compared with 1-day-stimulated CD34 epitope.23 Following electroporation of the TIL clone at cells. Although transfection efficiency, in terms of per- 300 V and 1050 ␮F, we detected 9 ± 3.5% EGFP+ TIL. centage of viable EGFP+ cells, after 4 days was not sig- Electroporated TIL responded as vigorously by IFN-␥ nificantly lower than at day 2, the levels of CD34 + + release to Melan-A25–37 peptide-pulsed HLA-A2 T2 cells expression favored the use of 2-day stimulated CD34 (kindly provided by Pierre Van der Bruggen, Ludwig cells (Table 1). Institute for Cancer Research, Brussels Branch), as control A recent report on nonviral transfection of CD34+ cells non-electroporated TIL, also exposed to the same targets by a gene gun technique demonstrated successful trans- (25.6 ± 4 versus 28 ± 3.5 IU/ml IFN-␥, respectively, n = fection of cord blood CD34+ cells and mentioned a trans- 3). These results show that electroporation does not fection efficiency of approximately 6% in LacZ-trans- adversely affect T cell function and suggest that our elec- fected 4-day cultured CD34+ cells,7 which is substantially troporation protocol might serve applications in adoptive lower than the transfection efficiency obtained by our immunotherapy strategies in which T cells are genetically protocol (26%), using the same culture conditions. Never- modified with immune-enhancing cytokines modulating theless, their optimal time-point of transfection (upon anti-tumor immune responses.24 This also suggests the day 2 of culture), as well as the CMV-driven luciferase possibility of establishing stable transfectants of primary expression kinetics, were in concordance with our obser- T cells or T cell clones by consecutive EGFP-based fluor- vations. escence-activated cell sorting (FACS), since T cells can be Strikingly, electroporation of unstimulated, fresh cultured and expanded in vitro for months allowing inte- gration of the electroporated plasmid, selection of stable transfectants by FACS and subsequent expansion.25 Using a FACStarPLUS cell sorter, we routinely obtained between 2 and 5 × 105 CD34+ cells starting from 8 to 12 ml of a fresh adult bone marrow sample. FCM analysis showed purities Ͼ98% CD34 positivity on the day of iso- lation (data not shown). CD34+ cells undergo substantial proliferation and differentiation to committed myeloid progenitors, precursors and differentiated cells when maintained in serum- or cytokine-containing medium.26 This differentiation is paralleled by a loss of CD34 expression. During culture in medium supplemented with IL-3+IL-6+SCF, CD34+ cells underwent substantial proliferation (four-fold and 15-fold expansion by days 4 and 7, respectively). Here, we evaluated electroporation as an alternative for transfection of both freshly isolated + Figure 3 EGFP expression in cultured CD34ϩ cells as a function of time. and 1- to 4-day cultured CD34 cells. The EGFP reporter ϩ system in combination with FCM analysis allowed accur- CD34 cells were sorted from total bone marrow mononuclear cells, cultured for 48 h in medium containing IL-3, IL-6, SCF and FL, and electro- ate determination of transfection efficiency in combi- porated as described in Table 1. At regular time-points after electropor- nation with simultaneous CD34 antigen staining at a sin- ation, transfected CD34ϩ cells were checked for EGFP expression by FCM gle-cell level. Optimization of the electroporation analysis as a function of time (n = 3).

Gene Therapy Electroporation of primary T cells and CD34+ cells VFI Van Tendeloo et al 1435

Figure 4 FACS profiles of fresh and cultured EGFP-transfected CD34ϩ cells. CD34ϩ cells were sorted from total bone marrow mononuclear cells and electroporated as described in Table 1, after 48 h culture in medium containing IL-3, IL-6, SCF and FL (a) or directly after sorting (b). Histogram plots of EGFP fluorescence in CD34ϩ cells (left side) show the overall percentage of EGFPϩ cells. Control-transfected CD34ϩ cells (bold) were used to set marker regions in order to determine the percentage of EGFP-expressing cells (histograms left). Dot plot analysis (right side) shows EGFP-transfected CD34ϩ cells simultaneously stained with PE-labeled CD34 antibody 24 h after electroporation. To set quadrants, a nonreactive isotype-matched monoclonal antibody (Becton Dickinson) was used. (c) Fresh BM mononuclear cells were resuspended at 50 × 106 c/ml in Opti-MEM and electroporated at 260 V and 1050 ␮F, as described in Table 1. One day after electroporation, cells were stained with a PE-labeled monoclonal CD34 antibody and CD34ϩ cells were gated (= R1; left dot plot). EGFP analysis was performed to assess gene transfer efficiency within the CD34ϩ compartment 24 h after electroporation (right dot plot). To set quadrants, a nonreactive isotype-matched monoclonal antibody (Becton Dickinson) was used. Representative dot plots are shown.

CD34+ cells resulted in a marked EGFP expression in ated CD34+ cells (147 ± 40 versus 134 ± 30 CFC, respect- approximately 5% of the CD34+ cells (depicted in Figure ively, n = 4). 4b). Interestingly, electroporation of fresh CD34+ cells As the CD34+ cell fraction comprises only 1–5% of was accompanied by only a minimal mortality (less than adult BM mononuclear cells (BMMC), we also investi- 5%) in contrast to a classical 20–50% mortality rate in gated the efficiency of transfection by electroporation of other cell types. Phenotypic analysis revealed that 95% of total BMMC. While an overall transfection efficiency of the cells were still CD34+ 24 h after electroporation. We nearly 2% was observed (data not shown), Figure 4c also investigated whether electroporated CD34+ cells clearly demonstrates that transient electroporation were still functional in terms of their colony-forming efficiency within the CD34+ compartment is similar to capacity in vitro. Using a colony-forming unit (CFU)- that obtained with purified CD34+ cells (5% EGFP+ cells). assay in methylcellulose semi-solid medium, we These results clearly highlight electroporation as a valid observed no significant difference between the colony- and effective nonviral method for gene delivery into both forming capacity of electroporated versus non-electropor- fresh and cultured CD34+ cells, even without preliminary

Gene Therapy Electroporation of primary T cells and CD34+ cells VFI Van Tendeloo et al 1436 Table 1 Outcome of electroporation in fresh and cultured CD34+ cells

Time-point of transfection % efﬁciency % viability % CD34

Fresh CD34+ cells day 0 (n = 4) 5.2 ± 0.4 95.8 ± 2.5 95 Cultured CD34+ cells day 1 (n = 3) 11.3 ± 1.0 85.0 ± 1.4 88 day 2 (n = 7) 29.6 ± 4.7 78.0 ± 3.4 66 day 4 (n = 5) 26.3 ± 3.9 74.7 ± 4.9 45

Adult bone marrow (BM) samples were collected after informed consent by sternal aspiration from hematologically normal patients undergoing cardiac surgery. Mononuclear BM cells were isolated from adult BM by density gradient centrifugation using Lymphocyte Separation Medium and stained with supernatant of the 43A1 hybridoma (anti-CD34, kindly donated by Dr H-J Bu¨hring, University of Tu¨ bingen, Germany),33 followed by fluorescein isothiocyanate (FITC)-conjugated rabbit anti-mouse immunoglobulins (DAKO, Glostrup, Denmark). The CD34-labeled cells were then sorted on a FACStarPLUS cell sorter (Becton Dickinson, Erembodegem, Belgium) as described elsewhere.34 Sort windows were set to include cells with low side scatter and with positive green fluorescence (CD34+). Purities of Ͼ98% were routinely obtained. Sorted CD34+ cells were plated at 105 cells/ml in 2 ml IMDM supplemented with 1% bovine serum albumin (BSA; Sigma, Bornem, Belgium), 100 ng/ml interleukin (IL)-3 (kindly provided by Dr SC Clark, Genetics Institute, Cambridge, MA, USA), 100 ng/ml IL-6 (Roche Molecular Biochemicals, Mannheim, Germany), 50 ng/ml stem cell factor (SCF) (Biosource, Nivelle, Belgium) and 100 ng/ml Flt-3 ligand (FL) (kindly provided by Dr SC Lyman, Immunex Corporation, Seattle, WA, USA). At different time-points after initiation of the cell culture with IL-3+IL-6+SCF+FL, fresh or cultured CD34+ cells were resuspended at 2 × 106 cells/ml and 500 ␮l of cell suspension was mixed with 20 ␮g of pEGFP-N1 (Clontech Laboratories, Palo Alto, CA, USA) containing an enhanced green fluorescent protein (EGFP) reporter gene, incubated for 1 min at room temperature and electroporated with an Easyject Plus apparatus (EquiBio, Kent, UK) at optimized parameters (260 V, 1050 ␮F). One day after electroporation, transfected CD34+ cells were washed once in PBS supplemented with 1% FCS and resuspended in 0.5 ml of PBS supplemented with 1% BSA and 0.1% azide. Ethidium bromide (EB) at a final concentration of 10 ␮g/ml was added directly before FCM analysis. Electroporation efficiency is expressed as the mean percentage of EGFP+ cells 24 h after electroporation. Viability was determined by FCM analysis of ethidium bromide exclusion and is expressed as the percentage of ethidium bromide-negative cells. Data are shown as mean ± s.d. The level of CD34 expression was determined simultaneously with EGFP fluorescence 24 h after transfection at the indicated time-point by immunophenotypic analysis, as described previously.16 Briefly, 5 × 105 transfected cells were washed once in PBS supplemented with 1% FCS and stained for 30 min at 4°C with 1 ␮g of monoclonal antibody in 150 ␮l of PBS containing 1% BSA and 0.1% azide and analyzed on a FACScan flow cytometer. In order to allow combination with EGFP analysis, a phycoerythrin (PE)-conjugated monoclonal CD34 antibody (Becton Dickinson) was used. To set quadrants, a nonreactive isotype-matched monoclonal antibody (Becton Dickinson) was used (see also Figure 4) and is expressed as the mean of three independent experiments.

CD34+ cell sorting. Together, our results show that elec- expression of transgenes in both human CD34+ HPC and troporation can markedly transfect not only cultured primary activated T lymphocytes and may ultimately CD34+ HPC at different stages but also unstimulated, find clinical applications in BM transplantation and/or freshly isolated CD34+ cells. These observations confirm gene therapy protocols that might benefit from short- that electroporation is able to transfect primary unstimu- term expression of therapeutic gene products in T lym- lated CD34+ cells, unlike retroviral MuLV-based vec- phocytes and CD34+ cells. tors.28 Interestingly, we observed a similar transfection efficiency within the CD34+ compartment of the total BMMC fraction as compared with that of highly purified Acknowledgements CD34+ cells. Aside from saving time and being cost-effective, this feature could be of importance for future appli- This work was supported by grant No. 3.0109.96 and No. cations of this electroporation-based protocol, as it might G.0157.99 of the Fund for Scientific Research, Flanders, eliminate the need for preliminary cell selection pro- Belgium (FWO), by a grant of the Scientific Committee cedures, such as FACsorting. The current electroporation of the ASLK-financed Cancer Research and by gifts of technique may thus provide a highly effective and rapid Mrs Ellie Van Belleghem to the HEBA Foundation means to transfect CD34+ cells with no or minimal pre- (Foundation for Hematology, Bone Marrow Transplan- stimulation, thereby avoiding prolonged ex vivo culti- tation and Blood Transfusion, Antwerp) in memory of vation and/or exposure to differentiation-promoting her late mother Louisa De Saedeleer. We would like to cytokines, which might be detrimental for ex vivo gene thank Dr SC Lyman and Dr EK Thomas (Immunex Cor- transfer applications using primitive hematopoietic stem poration, Seattle, WA, USA) for providing us with Flt-3 cells, present within the CD34+ population.29 However, it ligand, Dr John Wunderlich (NIH, Bethesda, USA) for the remains to be determined whether hematopoietic stem 1235 Melan-A-specific TIL clone. VVT is a research assist- cells were included in the transfected CD34+ population. ant of the FWO. RW is a holder of a research grant of Taken together, electroporation of HPC or T cells may the Vlaamse Kankerliga. PP is holder of a research grant be more beneficial in some gene therapy settings, such as of the Institute for Science and Technology (IWT). transient expression of bone marrow-protecting genes (eg MDR, cytokines);30 short-term expression of cytokines and/or growth factors;31 oncogene antisense oligonucleo- References 32 tides in bone marrow purging settings; or transient 1 Nienhuis AW et al. Gene transfer into hematopoietic cells. Stem expression of viral receptor genes in combination with Cells 1997; 15 (Suppl. 1): 123–134. 7 viral transduction protocols. 2 Anderson WF. Human gene therapy. Nature 1998; 392: 25–30. We conclude that electroporation represents a valuable 3 Moritz T, Williams DA. Gene transfer into the hematopoietic nonviral transfection tool for high-level short-term system. Curr Opin Hematol 1994; 1: 423–428.

Gene Therapy Electroporation of primary T cells and CD34+ cells VFI Van Tendeloo et al 1437 4WuAGet al. Improvement of gene transduction efficiency in T electroporation protocol applicable to a wide range of cell lines. lymphocytes using retroviral vectors. Hum Gene Ther 1999; 10: Biotechniques 1994; 17: 1058–1062. 977–982. 22 Cron RQ, Schubert LA, Lewis DB, Hughes CC. Consistent tran- 5 Glimm H et al. Efficient serum-free retroviral gene transfer into sient transfection of DNA into non-transformed human and primitive human hematopoietic progenitor cells by a defined, murine T-lymphocytes. J Immunol Meth 1997; 205: 145–150. high-titer, nonconcentrated vector-containing medium. Hum 23 Reeves ME et al. Retroviral transduction of human dendritic Gene Ther 1998; 9: 771–778. cells with a tumor-associated antigen gene. Cancer Res 1996; 56: 6 Hatzfeld A et al. Increased stable retroviral gene transfer in early 5672–5677. hematopoietic progenitors released from quiescence. Hum Gene 24 Hwu P et al. Functional and molecular characterization of Ther 1996; 7: 207–213. tumor-infiltrating lymphocytes transduced with tumor necrosis 7YeZQet al. Cytokine transgene expression and promoter usage factor-alpha cDNA for the gene therapy of cancer in humans. J in primary CD34+ cells using particle-mediated gene delivery. Immunol 1993; 150: 4104–4115. Hum Gene Ther 1998; 9: 2197–2205. 25 Gallot G et al. Human HLA-specific T-cell clones with stable 8 Verma S et al. Gene transfer into human umbilical cord blood- expression of a suicide gene: a possible tool to drive and control derived CD34+ cells by particle-mediated gene transfer. Gene a graft-versus-host graft-versus-leukemia reaction? Blood 1996; Therapy 1998; 5: 692–699. 88: 1098–1103. 9 Ledley FD. Nonviral gene therapy: the promise of genes as 26 Dorrell C et al. Expansion of human cord blood CD34(+)CD38(−) pharmaceutical products. Hum Gene Ther 1995; 6: 1129–1144. cells in ex vivo culture during retroviral transduction without a 10 Harrison GS et al. Optimization of gene transfer using cationic corresponding increase in SCID repopulating cell (SRC) fre- lipids in cell lines and primary human CD4+ and CD34+ hema- quency: dissociation of SRC phenotype and function. Blood 2000; topoietic cells. Biotechniques 1995; 19: 816–823. 95: 102–110. 11 Floch V et al. Cationic phosphonolipids as non viral vectors for 27 Kiem HP et al. Improved gene transfer into baboon marrow DNA transfection in hematopoietic cell lines and CD34+ cells. repopulating cells using recombinant human fibronectin frag- Blood Cells Mol Dis 1997; 23: 69–87. ment CH-296 in combination with interleukin-6, stem cell factor, 12 Featherstone C. Electroporation: an effective technique for drug FLT-3 ligand, and megakaryocyte growth and development fac- delivery and gene therapy. Am Biotechnol Lab 1993; 11: 16. tor. Blood 1998; 92: 1878–1886. 13 Keating A, Toneguzzo F. Gene transfer by electroporation: a 28 Smith C. Retroviral vector-mediated gene transfer into hemato- model for gene therapy. Prog Clin Biol Res 1990; 333: 491–498. poietic cells: prospects and issues. J Hematother 1992; 1: 155–166. 14 Matthews KE, Dev SB, Toneguzzo F, Keating A. Electroporation 29 Yonemura Y, Ku H, Lyman SD, Ogawa M. In vitro expansion for gene therapy. Meth Mol Biol 1995; 48: 273–280. of hematopoietic progenitors and maintenance of stem cells: 15 Muramatsu T, Nakamura A, Park HM. In vivo electroporation: comparison between FLT3/FLK-2 ligand and KIT ligand. Blood a powerful and convenient means of nonviral gene transfer to 1997; 89: 1915–1921. tissues of living animals (review). Int J Mol Med 1998; 1: 55–62. 30 Cooper MJ. Noninfectious gene transfer and expression systems 16 Van Tendeloo V et al. Nonviral transfection of distinct types of for cancer gene therapy. Semin Oncol 1996; 23: 172–187. human dendritic cells: high-efficiency gene transfer by electrop- 31 Gomez NJ, Curiel DT, Douglas JT. Gene therapy for cancer. Eur oration into hematopoietic progenitor- but not monocyte- J Cancer 1999; 35: 867–885. derived dendritic cells. Gene Therapy 1998; 5: 700–707. 32 Bergan R et al. Electroporation of synthetic oligodeoxynucleot- 17 Satoh E et al. Successful transfer of ADA gene in vitro into ides: a novel technique for ex vivo bone marrow purging. Blood human peripheral blood CD34+ cells by transfecting EBV-based 1996; 88: 731–741. episomal vectors. FEBS Lett 1998; 441: 39–42. 33 Buhring HJ et al. The product of the proto-oncogene c-kit (P145c- 18 Ebert O et al. Lymphocyte apoptosis: induction by gene transfer kit) is a human bone marrow surface antigen of hemopoietic techniques. Gene Therapy 1997; 4: 296–302. precursor cells which is expressed on a subset of acute non- 19 Cheng L et al. A GFP reporter system to assess gene transfer lymphoblastic leukemic cells. Leukemia 1991; 5: 854–860. and expression in human hematopoietic progenitor cells. Gene 34 Lardon F et al. Generation of dendritic cells from bone marrow Therapy 1997; 4: 1013–1022. progenitors using GM-CSF, TNF-alpha, and additional cyto- 20 Weaver JC. Electroporation theory. Concepts and mechanisms. kines: antagonistic effects of IL-4 and IFN-gamma and selective Meth Mol Biol 1995; 55: 3–28. involvement of TNF-alpha receptor-1. Immunology 1997; 91: 21 Baum C, Forster P, Hegewisch BS, Harbers K. An optimized 553–559.

Gene Therapy