Lentivirus-Mediated Gene Transfer in Primary T Cells Is Enhanced by a Central DNA ﬂap

Home , Lentivirus

Gene Therapy (2001) 8, 190–198  2001 Nature Publishing Group All rights reserved 0969-7128/01 $15.00 www.nature.com/gt RESEARCH ARTICLE Lentivirus-mediated gene transfer in primary T cells is enhanced by a central DNA ﬂap

V Dardalhon1, B Herpers1, N Noraz1, F Pflumio2, D Guetard3, C Leveau3, A Dubart-Kupperschmitt2, P Charneau3 and N Taylor1 1Institut de Geńe´tique Molećulaire de Montpellier, UMR 5535/IFR 22; 2INSERM U474, Maternite´ de Port Royal; and 3Unite´ d’Oncologie Virale, Institut Pasteur, Paris, France

Retroviral vectors have become the primary tool for gene vated T cells (51 ± 12.7% versus 15 ± 1.4%). CD4+ and delivery into hematopoietic cells, including T lymphocytes. CD8+ T cells were transduced at equivalent levels. Lentiviral vectors offer an advantage over Moloney murine Importantly, freshly isolated T cells stimulated only during leukemia virus (MuLV) vectors because of their ability to the 12-h transduction period could be efficiently transduced translocate across an intact nuclear membrane and inte- with this new flap-containing lentiviral vector, but not with grate into the genome of nonproliferating cells. We have the parental lentiviral vector nor an MuLV vector. Transgene recently demonstrated that a central strand displacement expression in the flap-containing lentiviral vector, under the event, controlled by the central polypurine tract (cPPT) and control of either an internal cytomegalovirus or the elong- the central termination sequence (CTS), results in the forma- ation factor-1 alpha (EF1␣) promoter, was significant and tion of a central DNA flap which acts as a cis-determinant expression remained elevated in resting T cells. Thus, this of HIV-1 genome nuclear import. Here, we show that inser- system allows stable expression of transgenes in T lympho- tion of this DNA determinant in a classical lentiviral vector cytes following a short ex vivo transduction protocol. Gene resulted in a significantly higher level of transduction in acti- Therapy (2001) 8, 190–198.

Keywords: lentiviral vector; DNA ﬂap; gene transfer; primary T cells

Introduction lines.8–10 Additionally, it has been demonstrated that T cell transduction can be enhanced by pseudotyping Genetic manipulation of T lymphocytes may significantly virions with the gibbon ape leukemia virus (GALV) benefit patients with various genetic or acquired hemato- envelope rather than an amphotropic envelope.9,11 logic disorders. Transfer of immunomodulatory and anti- Finally, co-localization of retrovirus and target cells on viral genes into T cells will enable the development of specific adhesion domains of a recombinant fibronectin innovative therapies for cancer patients and human molecule has resulted in augmented transduction of T immunodeficiency virus type 1 (HIV-1) patients, respect- cells.12–14 1–4 ively. Moreover, transfer of wild-type genes into T cells Despite these advances, a significant barrier to the use with defects in enzymatic genes such as adenosine deam- of MuLV-based vectors for T cell gene therapy remains inase may allow the correction of these types of dis- the quiescent state of the majority of circulating T cells. 5–7 orders. At this time, the most widely used method for The integration of oncoretroviruses such as MuLV gene transfer into human T cells has involved Moloney requires mitosis, due to an inability to translocate the murine leukemia virus (MuLV)-based vectors. MuLV- viral genome across an intact nuclear membrane.15–17 based vectors allow stable gene transfer as the genetic Thus, primary T cells need to be stimulated before any material contained within the vector is integrated into MuLV-based gene transfer. This is clearly a hindrance to target cell genomes. While this methodology holds prom- gene therapy applications for both practical and theoreti- ise, optimism has been moderated by the difficulties cal reasons: cells need to be activated during an extended encountered in achieving high levels of gene transfer in period of ex vivo culture and the functional capacity of T clinical trials. Nevertheless, new techniques have signifi- cells may change following ex vivo stimulation. The cantly improved transduction efficiencies. Optimisations development of lentiviral vectors can theoretically over- include collection of cell-free viral supernatants after cul- come this limitation since lentiviruses have evolved a ° ture at 32 C, co-centrifugation of supernatant with cells, mitosis-independent nuclear import strategy allowing multiple exposures of cells to virus as well as techniques them to integrate in nondividing cells. Indeed, vectors where T cells are cocultivated with retrovirus producer derived from the HIV-1 genome have recently been shown to transduce terminally differentiated nondividing cells.18–23 Nevertheless, lentiviral transduction of Correspondence: N Taylor, Institut de Geńe´tique Molećulaire de 24 Montpellier, 1919 Route de Mende, 34293 Montpellier, Cedex 5, France quiescent T cells has not yet been achieved. It is not The first two authors contributed equally to this work clear whether quiescent T cells have not been transduced Received 17 August 2000; accepted 25 October 2000 because the previously employed lentiviral vectors were Lentiviral gene transfer in primary T cells V Dardalhon et al 191 not optimized or alternatively, because integration of a lentiviral vector in these cells requires cell cycle progression. This latter hypothesis is based on the finding that wild-type HIV-1 infection of quiescent T cells is not successful due to a block in reverse transcription.25 It has recently been demonstrated that the classical lentiviral vectors, which have been used to date, are not optimized for nuclear import. The viral determinants responsible for active nuclear import remain contro- versial, but we have shown that a plus-strand overlap, the DNA flap, acts as a cis-determinant of HIV-1 genome nuclear import. This central DNA flap is created because the reverse transcription strategy of HIV-1 involves the synthesis of two discrete half-genomic fragments with a central strand displacement event controlled in cis by the central polypurine tract (cPPT) and the central termination sequence (CTS).26,27 The linear DNA molecule, thus formed, bears a 99 nucleotide overlap at its center. The addition of this sequence to the classical lentiviral vector, in which the viral coding sequences between the LTRs are deleted, results in significantly enhanced nuclear import in dividing as well as non-dividing HeLa cells.28 Figure 1 Schematic representation of lentiviral- and MuLV-based vectors. Moreover, we and others have found that comp- The TRIP-CMV vector was derived from the previously described HR- lementation of this nuclear import defect results in CMV vector by the insertion, immediately upstream of the internal CMV + enhanced transduction of human CD34 progenitor promoter, of a 178 bp HIV-1 DNA fragment containing the cPPT and cells.29,30 Here, we have assessed the capacity of this CTS central cis-active sequences.28 Derivative TRIP vectors in which the U3 region was deleted (TRIP⌬U3-CMV), and in which the CMV pro- enhanced flap-containing lentiviral vector to transduce ␣ ⌬ ␣ quiescent and activated primary T lymphocytes. moter was substituted by the EF1 promoter (TRIP U3-EF1 ) are indicated (see Materials and methods). The previously described MuLV-based vector containing an internal ribosome entry site (IRES) is also shown.14 All vectors harbor the EGFP coding sequence and are denoted by the Results promoter driving its expression.

Conditions for transduction of T cells with a TRIP lentiviral vector We were first interested in determining optimal conditions for transduction of primary T cells using TRIP lentiviral vectors containing the central DNA flap. A schematic representation of the TRIP lentiviral vectors used in these experiments is shown in Figure 1. In the TRIP⌬U3-EF1␣ vector, the U3 LTR sequence is deleted and EGFP transgene expression is driven from an internal EF1␣ promoter. T cells were activated for 48 h with immobilized anti-CD3 and anti-CD28 monoclonal antibodies and transductions were performed with the VSV-G-pseudotyped TRIP⌬U3-EF1␣ vector. Transduc- tions with different quantities of virion particles, normalized by measuring the P24 content of supernatants, demonstrated a linear dose response. Specifically, following a single 12-h exposure to virion particles corresponding to 10–100 ng of P24/ml, gene transfer levels increased in a linear fashion, from approximately 10% to greater Figure 2 Dose response transduction experiments of primary T cells with than 50% (Figure 2). the TRIP⌬U3 lentiviral vector. Human peripheral blood lymphocytes were Co-localization of retrovirus and target cells on a activated for 48 h with immobilized anti-CD3 and anti-CD28 antibodies. recombinant fibronectin fragment (CH-296) has been Cells were then exposed to TRIP⌬U3 lentiviral particles, corresponding shown to significantly increase gene transfer of ampho- to 10–100 ng of P24/ml, for a 12-h period. Transductions were performed tropic and GALV-pseudotyped MuLV vectors into T in the presence or absence of recombinant fibronectin (CH-296) and poly- ␮ cells.12–14 Transduction efficiencies with the VSV-G- brene (4 g/ml). Transductions were performed in duplicate and gene transfer efficiency was assessed 48 h later by FACS analysis. The mean pseudotyped TRIP lentiviral vector were only augmented numbers of EGFP-expressing cells are indicated. by 10–20% in the presence of fibronectin. However, it is important to note that this increase was consistent and was observed at all tested concentrations of virion par- to T cells, an inhibitory effect cannot be excluded. Based ticles (Figure 2). Addition of the polybrene cation did not on these data, all subsequent experiments were perfor- enhance gene transfer efficiencies, but rather transduction med with a concentration of virion particles correspond- was slightly decreased. Although the dose of polybrene ing to 50 ng P24/ml in the presence of the recombinant used in these experiments, 4 ␮g/ml, did not appear toxic fibronectin fragment.

Gene Therapy Lentiviral gene transfer in primary T cells V Dardalhon et al 192 Transduction of quiescent and activated T cells with MuLV vector. It is nevertheless noteworthy that after a lentiviral and MuLV vectors short 12-h prestimulation with ␣CD3/␣CD28 antibodies, As wild-type HIV-1 cannot infect circulating peripheral T cells could be efficiently transduced by the TRIP-CMV 25 ± T lymphocytes, most of which are in a G0 resting state, vector (39.5 0.7%), but not by the HR-CMV vector (11 it was of significant interest to assess whether quiescent T ± 1.4%). The high levels of transduction attained with cells could be transduced with the enhanced TRIP vector. TRIP-CMV were also reached following exposure of T Since variation in the efficiency of transduction can be cells to the newly generated TRIP⌬U3-CMV and observed between different batches of virions, all com- TRIP⌬U3-EF1␣ vectors (Figure 3a and data not shown). parative experiments between the parental HR-CMV vec- MuLV-based gene transfer remained very low (6.8 ± tor and the various TRIP vectors were performed with 0.1%), likely because mitosis is only induced after a stocks of virions produced concurrently (50 ng P24/ml). longer stimulation with activating antibodies.14 The aug- It is important to underline that the reported percentages mentation in gene transduction by the TRIP vectors was of EGFP+ cells represent transduction, and not pseudo- also observed following a 24-h prestimulation with transduction or transient expression from nonintegrated ␣CD3/␣CD28 antibodies. Transduction efficiencies by vector DNA, because EGFP levels remained stable for the TRIP-CMV, TRIP⌬U3-EF1␣ and HR-CMV vectors more than 2 weeks in culture. were 51 ± 12.7%, 71 ± 3.8% and 15 ± 1.4%, respectively. In agreement with results obtained with wild-type We find that this augmented level of gene transfer is HIV-1, freshly isolated resting T cells were not trans- specifically due to the presence of the DNA flap because duced with the parental HR-CMV vector or with the transduction efficiencies obtained with a TRIPinv-CMV TRIP-CMV vector (Figure 3a). As expected, these cells vector, where the central region is inserted in a reverse were also not transduced with a GALV-pseudotyped nonfunctional orientation,28 were as low as those observed with the parental HR-CMV vector (data not shown). All together, these data demonstrate that the presence of the central DNA flap in the lentiviral vector enhanced transduction of T cells which were prestimulated for only a very short time period, 12 h. Notably though, the central DNA flap was also required for optimal lentivirus- mediated gene transfer in dividing T cells. This latter conclusion is based on our finding that at a time-point when MuLV-based gene transfer was very high (24–36 h) (Figure 3a), and as such the cells were imperatively undergoing mitosis, the presence of the DNA flap con- tinued to significantly augment lentiviral gene transfer. For many experimental applications, it is desirable to transduce T cells immediately following their isolation. While our results obtained using the TRIP vector were very encouraging, they nonetheless precluded the immediate transduction of fresh T lymphocytes. In order to assess whether a concurrent exposure of T lymphocytes to virion particles and ␣CD3/␣CD28 antibodies would result in gene transfer, freshly isolated T lymphocytes were exposed to TRIP vector inoculums for 12 h in the presence of the stimulating anti-CD3 and anti-CD28 antibodies. Notably, these T lymphocytes were efficiently transduced with levels of 31 ± 0.5% in a representative experiment (Figure 3b). In contrast, equivalently treated cells transduced by HR or MuLV-based vectors were transduced at insignificant levels (Figure 3b). Thus, we have established experimental conditions under which the TRIP vector can support gene transfer into freshly isolated lymphocytes.

Figure 3 Kinetics of transduction of primary T cells with TRIP⌬U3- Gene transfer efficiencies in CD4+ and CD8+ T cells EF1␣, TRIP-CMV, HR-CMV and MuLV-5ЈLTR retroviral vectors. Fol- using TRIP and MuLV vectors lowing activation of freshly isolated T lymphocytes with immobilized anti- The ability to transduce both CD4+ and CD8+ T cells has CD3/CD28 antibodies for the indicated time periods, cells were removed from the activating antibodies and exposed to cell-free retroviral super- important implications for gene therapy strategies. It has natants for 12 h. Cells were exposed to 50 ng P24/ml of TRIP or HR recently been reported that the classical lentiviral vectors + viral stocks, or alternatively to GALV-pseudotyped MuLV virions. All transduce CD8 T cells at significantly lower efficiencies transductions were performed on recombinant fibronectin-coated plates in than CD4+ T cells (20% versus 50%).24 Additionally, it has the presence of recombinant IL-2 (a). Freshly isolated T cells were trans- been suggested that GALV-pseudotyped MuLV-based duced as described above except that the 12-h exposure to cell-free retrovi- virions are limited in their abilities to transduce primary ral supernatants was performed in the presence of anti-CD3/CD28 anti- CD8+ T cells.31 We therefore compared transduction of bodies (concurrent stimulation/transduction) (b). Experiments were + + performed in duplicate and the mean numbers of EGFP+ cells are indi- CD4 and CD8 T cells with both the TRIP vector and an cated. Results are representative of one of three independent experiments. MuLV vector. Following activation of T cells with anti-

Gene Therapy Lentiviral gene transfer in primary T cells V Dardalhon et al 193 CD3/CD28 antibodies for 24 h, lymphocytes were moter, in lieu of the CMV promoter in the U3-deleted exposed to TRIP or MuLV virions for a 12-h period. As TRIP vector, was associated with an approximately equi- demonstrated in Figure 4, both CD4+ and CD8+ T cells valent level of expression in primary T cells (MFI-209). were transduced to levels of greater than 50% by TRIP, Thus, both promoters can be used to achieve high-level as well as by MuLV vectors. transgene expression. The activation status of primary T lymphocytes may Transgene expression in mitotically active and resting T significantly affect transgene expression from an internal cells transduced with TRIP vectors promoter in MuLV-based vectors.12,33,34 Specifically, The major reason for generating TRIP vectors in which transgene expression has been shown to decrease in T the transcriptional activity of the LTR was abolished cells which return to resting phase following their acti- ⌬ (TRIP U3 vectors, Figure 1) was to prevent the emerg- vation. In order to assess whether expression of the EGFP ence of replication-competent recombinants. Our prelimi- transgene was stable following transduction with the nary experiments demonstrated that the ability of the ⌬ ␣ ⌬ TRIP U3-EF1 vector, cultures of transduced T cells initially constructed TRIP vector and the new TRIP U3 were maintained in media containing exogenous recom- vectors to transduce primary T cells was equivalent binant IL-2 and EGFP expression was monitored at vari- (Figure 3). Interestingly though, removal of the enhancer ous time-points. Notably, the percentage of EGFP+ cells and promoter from U3 in the 5Ј HIV LTR did result in a remained stable for greater than 15 days after transduc- change in transgene expression, with a greater than six- tion (Figure 6a). By 10 days after transduction, cells had fold increase in EGFP levels in T cells transduced with returned to a resting phase, as monitored by cell cycle TRIP⌬U3-CMV as compared with cells transduced with analysis (data not shown). Additionally, the level of the original TRIP-CMV vector (mean fluorescence intensity (MFI) of 34 versus 248) (Figure 5). EGFP expression, measured as the mean fluorescence Transgene expression from an internal promoter in a intensity, remained stable throughout the culture period retroviral vector is, at least partially, dependent upon the (Figure 6b). While these experiments were performed using TRIP⌬U3-EF1␣, equivalent results were obtained nature of the promoter itself. Thus, we next compared ⌬ transgene expression driven from the CMV immediate– using the TRIP U3-CMV vector (data not shown). early promoter with that driven from the EF1␣ promoter, Together, these data show that transgene expression from ␣ because although the former has been shown to be the CMV and EF1 internal promoters in the TRIP lenti- extremely efficient in many cells types, the latter is a virus-derived vector remains unchanged in activated and ubiquitous cellular promoter which is functional in resting phase primary T cells. quiescent cells.32 We find that insertion of the EF1␣ pro-

Figure 4 Transduction of CD4+ and CD8+ T cells subsets with the TRIP⌬U3-EF1␣ and MuLV retroviral vectors. Cells were activated for 24 h as described and exposed to retrovirus during a 12-h period. The phenotype of the transduced cells was assessed by simultaneously monitoring EGFP ﬂuorescence and CD4 or CD8 expression with PE-Cy5-conjugated antibodies. The percentage of cells in each quadrant is indicated.

Gene Therapy Lentiviral gene transfer in primary T cells V Dardalhon et al 194

Figure 5 EGFP transgene expression in T cells transduced with TRIP vectors. The level of EGFP expression in transduced cells, as assessed by the mean ﬂuorescence intensity (MFI), was monitored 48 h following transduction. The MFI of cells transduced with the TRIP-CMV vector, the TRIP⌬U3- CMV vector and the TRIP⌬U3-EF1␣ vector are shown. In all three histograms, the percentage of EGFP+ cells is between 33 and 36%.

Figure 6 Kinetics of EGFP expression in T cells transduced with the TRIP⌬U3-EF1␣ vector. T cells prestimulated with anti-CD3/CD28 antibodies were transduced with TRIP⌬U3-EF1␣ retroviral stocks (10–100 ng P24/ml). Cells were maintained in media containing recombinant IL-2 and the percentage of EGFP+ cells was monitored during a 15 day period. The mean numbers of EGFP+ cells are indicated and values varied by less than 10% (a). Histograms demonstrating the level of EGFP expression (MFI) at days 4, 6 and 11 after transduction with 50 ng P24/ml are shown (b). Data are representative of results obtained in three independent experiments.

Discussion sequence into the parental HR lentiviral vector results in Previous gene transfer studies utilizing lentiviral vectors a significantly higher level of nuclear import in human have been compromised because the vectors were not CD34+ hematopoietic cells.29 In these experiments, the optimized for nuclear import. Specifically, the central increased level of nuclear import translated into a strik- DNA flap in the center of HIV-1, which acts as a cis-deter- ing increase in gene transduction efficiency.29 Here, we minant of HIV-1 nuclear import28 was not present in first show that inclusion of this central DNA flap resulted in and second generation lentiviral vectors.19,23,35–37 We have a significant improvement in gene transfer into prestimu- recently demonstrated that introduction of this DNA flap lated CD4+, as well as CD8+ human T cells. While we

Gene Therapy Lentiviral gene transfer in primary T cells V Dardalhon et al 195 detected gene transfer efficiencies of greater than 50% ability of the enhanced TRIP vector to transduce nondi- with the flap-containing TRIP vector, it is important to viding cells does not appear to be limited to T lympho- note that the majority of the experiments reported here cytes since we also detected efficient gene transfer in were done under conditions where the quantity of virion CD34+ cells which had not undergone division during the particles was limiting (50 ng P24/ml). Experiments per- transduction period.29 It remains to be determined formed with higher concentrations of virions (500 ng whether entry into the G1b phase of the cell cycle is P24/ml) resulted in a T cell transduction efficiency of required for lentiviral transduction of all cell types, or greater than 80% (data not shown). alternatively, is a specific requirement for the trans- The ability to achieve high levels of gene transfer in duction of T lymphocytes. primary T cells following a short ex vivo transduction per- The successful transduction of both CD4+ and CD8+ T iod is highly desirable for many gene therapy strategies. cells is important for many clinical applications. A pre- In this context, lentiviral vectors are superior to onco- vious study using lentiviral vectors lacking the central retroviral vectors because they can reportedly integrate DNA flap reported a decreased level of transduction of into the genome of nonproliferating cells and most circul- CD8+ T cells versus CD4+ T cells.24 These authors sug- ating peripheral T cells are not in cycle. While lentiviral gested that the decreased level of CD8+ T cell transduc- vectors have been shown to transduce noncycling macro- tion may be due to the secretion of various inhibitory phages and HeLa cells arrested in G1-S and G2 phases factors such as IFNs and/or an absence of factors which of the cell cycle,19,22 it was not clear whether they could promote efficient lentiviral integration. In this context, it integrate into nonstimulated T cells. This question arose is notable that the CD8+ cells were transduced 4–6 days because the successful infection of quiescent CD4 lym- following T cell receptor (TCR) activation, a period dur- phocytes by wild-type HIV-1 requires a transition into ing which there is a significant secretion of cytokines.41,42 38 the G1b stage of the cell cycle. Although gene transfer In contrast, in our lentiviral transduction experiments into quiescent T cells has not been achieved with lentivi- which were performed at an earlier time-point, 24 h post- ral vectors lacking the central DNA flap,24 it was possible TCR activation, and in those reported by Costello and that transduction efficiencies would be augmented by the colleagues,43 performed 48 h after activation, gene trans- inclusion of this sequence in the TRIP vector. However, fer efficiencies in CD4+ and CD8+ T cells were equivalent. we did not observe transduction of freshly isolated T cells Moreover, Uckert et al31 found that CD8+ T cells were using the enhanced TRIP vector. An alternative expla- transduced very inefficiently with a GALV-pseudotyped nation is that this resting population of T cells was trans- MuLV vector following a relatively long prestimulation duced with the TRIP vector, but EGFP expression was (4–6 days), while we detected high MuLV-mediated not detected because the transgene cannot be expressed transduction of CD8+ T cells following a 24-h stimulation. in quiescent cells. This latter possibility is nevertheless Thus, it is possible that transduction of CD8+ T cells with unlikely because the EF1␣ promoter present in both VSV-G-pseudotyped lentiviral vectors and GALV- the TRIP⌬U3-EF1␣ promoter is known to be highly pseudotyped MuLV-based vectors decreases when cells functional in quiescent cells.32 are prestimulated for greater than 48 h. Further studies In apparent contradiction with our results, Case et al39 will help to determine whether lentiviral- and MuLV- recently reported that approximately 12% of CD34+ cells mediated gene transfer into CD8+ T cells, but not CD4+ in the G0 phase of the cell cycle could be transduced with T cells, decreases following long-term TCR stimulation. a non-FLAP-containing lentiviral vector. Several possi- If this is the case, it will be interesting to identify the bilities can account for this difference: (1) the CD34+ cells factor(s) which differentially regulate lentiviral trans- + + progressed from G0 to G1b during the transduction per- duction in CD4 and CD8 T cells. + iod; (2) transgene expression in the CD34 cells in G0 rep- Stable high level expression of the transgene of interest resented pseudotransduction; or (3) unlike primary T is a crucial parameter for gene therapy and as such, we + cells, CD34 cells can be transduced during the G0 phase monitored transgene expression from various lentiviral of the cell cycle. Although we found that T cells in G0 vector constructs in primary T cells. It was very aus- could not be transduced with the TRIP vector, cell picious that deletion of the promoter and enhancer from division itself was not required for integration and sub- U3, which were eliminated in order to create a self-inacti- sequent transgene expression. Using the fluorescent dye vating TRIP vector with improved biosafety, resulted in carboxyfluorescein succinimidyl ester (CFSE) which a significantly higher level of transgene expression in allows distinct generations of progeny to be dis- both T cells and CD34+ progenitor cells (this report and tinguished,40 we determined that less than 20% of T cells FP, AD, and PC, manuscript in preparation). Specifically, had undergone division by 48 h following ␣CD3/␣CD28 transgene expression from the internal CMV promoter stimulation (VD, unpublished observations). The high was approximately six-fold higher in primary T cells and level of transduction observed in freshly isolated T cells three-fold higher in CD34+ cells following deletion of the which were immediately exposed to TRIP virions for 12 h U3 sequences. The elimination of a transcriptional inter- in the presence of activating antibodies strongly indicates ference between the LTR and the internal CMV promoter that gene transfer was achieved in T cells which had not likely accounts for this augmented transgene expression. yet undergone division. Importantly, cell cycle entry was Notably, these results are in agreement with those not modulated in T cells stimulated with ␣-CD3/␣-CD28 recently reported by Kung et al.44 Specifically, they found antibodies only during this short 12-h period. It is also that in T cells, transgene expression driven from an essential to underline that the EGFP expression detected internal MuLV-based promoter in a lentiviral vector was in these cells reflects transduction, and not pseudotrans- increased following deletion of U3 sequences. Zufferey duction or transient expression from non-integrated vec- and colleagues37 found that deletion of U3 in a second tor DNA, because the percentage of EGFP+ cells remained generation lentiviral vector resulted in more modest essentially constant for more than 2 weeks in culture. The changes in CMV promoter-driven transgene expression,

Gene Therapy Lentiviral gene transfer in primary T cells V Dardalhon et al 196 ranging from 0.5-fold to 2.0-fold. The apparent discrep- second round of PCR was performed on these two PCR ancy between their data and those reported here is prob- products using the external EcoRI-TRIP+ and BamHI- ably due to the nature of the target cell in which trans- EF1− primers generating a single PCR product. The TRIP- gene expression was monitored (293T, HeLa, and 3T3 EF1␣ plasmid was generated by replacing the EcoRI– cells versus CD34+ and T cells). BamHI fragment of TRIP-CMV with the above-described The ability of two internal promoters to drive trans- EcoRI–BamHI-digested PCR product containing the flap gene expression in the context of the TRIP lentiviral vec- sequence and the EF1␣ promoter. In this construct, an tor varied in distinct cell types. Expression of EGFP in MluI site has been inserted between the triplex sequence hepatocytes and HeLa cells was two- to three-fold higher and the EF1␣ promoter. To generate the TRIP⌬U3-EF1␣ when driven from the CMV promoter as compared with plasmid, the EcoRI–BamHI fragment of TRIP-EF1␣, con- the EF1␣ promoter (PC, unpublished observations). In taining the flap sequence and the EF1␣ promoter, was contrast, EGFP expression in primary T cells, driven from swapped with the EcoRI–BamHI fragment of TRIP⌬U3- the CMV and EF1␣ promoters, was approximately equiv- CMV. alent. Further studies are required to determine whether transcription factors which specifically bind and activate Virus production the EF1␣ promoter are differentially expressed in T lym- Virion particles containing the HR or TRIP vectors were phocytes, as compared with these other cell types. Never- produced by transient calcium phosphate co-transfection theless, it is important to note that transgene expression of 293T cells with the vector plasmid, an encapsidation from both promoters, in the context of the TRIP⌬U3 vec- plasmid lacking Vif, Vpr, Vpu and Nef accessory HIV- tor, was stable in resting T cells. Thus, the TRIP⌬U3 vec- 1 proteins (p8.91),22,37 and a vesicular stomatitis virus-G tor offers an advantage in comparison with MuLV-based protein (VSV-G) envelope expression plasmid (pHCMV- vectors as transgene expression from the latter appears G),46 as previously described.19 Since some variation is to be sensitive to the activation status of the cell.33,34,45 observed in the efficiency of transduction between differ- The data presented here demonstrate that primary ent batches of virions, all comparative experiments CD4+ and CD8+ T cells can be efficiently transduced by between HR and TRIP vectors were performed with a VSV-G-pseudotyped TRIP lentiviral vector, even after stocks of virions produced in parallel. The concentration a brief exposure to virion particles. Inclusion of the cen- of virion particles was normalized by measuring the P24 tral DNA flap in the lentiviral vector was crucial for high (HIV-1 capsid protein) content of supernatants by ELISA. gene transfer levels in freshly isolated, as well as pre- Production of supernatants containing gibbon ape leuke- stimulated T cells. In conclusion, the use of this vector mia virus envelope (GALV)-pseudotyped MuLV vector strategy for T cell-based gene therapy of genetic has been previously described.14,47 disorders, viral infections and cancers appears very promising. Preparation of primary T cells and lymphocyte activation Peripheral blood samples were obtained from healthy adult donors. Peripheral blood mononuclear cells Materials and methods (PBMC) were separated by Ficoll–Hypaque (Sigma Ald- rich, St Louis, MO, USA) density gradient centrifugation. Construction of TRIP⌬U3-CMV and TRIP⌬U3-EF1␣ T cell were grown in RPMI medium (GibcoBRL, Grand vectors Island, NY, USA) with 10% FCS supplemented with The KpnI–XbaI fragment of the initially described TRIP- recombinant IL-2 (75 U/ml) (Chiron, Emeryville, CA, CMV vector,28 containing the polypurine tract and the USA). Before prestimulated transduction, lymphocytes 3ЈLTR, was subcloned into pUC18, generating the plas- (1 × 106/well) were stimulated with an immobilized anti- mid pUC-LTR. The EcoRI site was destroyed, creating CD3 antibody (UCHT1, 1 ␮g/ml; the generous gift of D pUC-LTR/RI-. A diverging PCR was performed on pUC- Cantrell, ICRF, London, UK) together with an anti-CD28 LTR/RI- in order to amplify the plasmid, but exclude the antibody (9.3, 1 ␮g/ml; kindly provided by C June, Uni- promoter and enhancer of the U3 sequence, using the oli- versity of Pennsylvania School of Medicine, Philadephia, gonucleotides: DU3-: 5Ј-CGGAATTCGGATCCGCGGCC PA) on non-tissue culture-treated plates for the indicated GCATCGATCTTGTCTTCGTTGGGAGTG-3Ј and DU3+: time period. 5ЈCGGAATTCAGCCGTCTCGAGAGATGCTGCATATA AGCAGC-3Ј.AnEcoRI site was inserted in the 5Ј and 3Ј Retroviral transductions primer sequences such that the resulting pUC- At the indicated time-points, cells were removed from LTR⌬U3/RI- plasmid contained an EcoRI site in lieu of plates containing the immobilized antibodies and trans- the 400 nucleotide U3 sequence. The TRIP⌬U3-CMV plas- duced on fibronectin-coated plates essentially as mid was constructed by replacing the KpnI–NheI frag- described by Moritz et al.48 The recombinant fibronectin ment of TRIP-CMV, containing the 3Ј LTR, with the fragment (CH-296) which contains the connecting seg- KpnI–XbaI fragment of pUC-LTR⌬U3/RI-. ment, cell-binding domain, and heparin-binding The EF1␣ promoter was kindly provided by Dr S domain,49 was kindly provided by Takara Shuzo (Otsu Nagata, Osaka Bioscience Institute, Japan. It was ampli- Shiga, Japan). Non-tissue culture-treated 24-well plates fied using the following two primers: MluI-EF1+:5Ј- were coated with fibronectin (8 ␮g/cm2) for 2 h at room CTGATACGCGTCGTGAGGCTCCGGTG-3Ј and BamHI- temperature. The fibronectin was then removed, plates EF1−:5Ј-CGGGATCCTGTGTTCTGGCGGCAAAC-3Ј.The were blocked with 1% bovine serum albumin (BSA) for flap sequence was amplified with the two primers: EcoRI- 20 min at 37°C, and subsequently washed once with PBS TRIP+:5Ј-GTCGTCGGCGCCGAATTCACAAATGGCA before use. In some experiments polybrene (Sigma- GTATTCATCC-3Ј and MluITRIP−:5Ј-AGCCTCAC- Aldrich) (4 ␮g/ml) was added to cells during the trans- GACGCGTATCAGCCAAAGTGGATCTCTGCTG-3Ј.A duction period while in other experiments, cells were

Gene Therapy Lentiviral gene transfer in primary T cells V Dardalhon et al 197 transduced on fibronectin-coated plates in the presence 10 Rudoll T et al. High-efficiency retroviral vector mediated gene + of immobilized anti-CD3/anti-CD28 antibodies. transfer into human peripheral blood CD4 T lymphocytes. Gene Lymphocytes (1 × 106) were then incubated with lenti- Therapy 1996; 3: 695–705. viral vector particles, at concentrations corresponding to 11 Lam JS et al. Improved gene transfer into human lymphocytes 10–100 ng of viral P24/ml, for 12 h in RPMI medium con- using retroviruses with the gibbon ape leukemia virus envelope. Hum Gene Ther 1996; 7: 1415–1422. taining recombinant IL-2. Alternatively, lymphocytes 12 Pollok KE et al. High-efficiency gene transfer into normal and were exposed to 0.5 ml of cell-free MuLV retroviral adenosine deaminase-deficient T lymphocytes is mediated by supernatant under equivalent conditions. Although titers transduction on recombinant fibronectin fragments. J Virol 1998; of GALV-pseudotyped MuLV particles and VSV-G- 72: 4882–4892. pseudotyped lentiviral particles cannot be easily com- 13 Fehse B et al. Highly-efficient gene transfer with retroviral vec- pared, all comparative experiments were performed tors into human T lymphocytes on fibronectin. Br J Haematol using concentrations of virions that allowed a 50% infec- 1998; 102: 566–574. tion of cycling T cells. After a 12-h exposure to retrovirus 14 Dardalhon V et al. Green fluorescent protein as a selectable at 37°C, cells were centrifuged and resuspended in fresh marker of fibronectin-facilitated retroviral gene transfer in pri- medium with IL-2. Transduction efficiencies were mary human T lymphocytes. Hum Gene Ther 1999; 10: 5–14. assessed 48 h later or as indicated. Transduced cells were 15 Roe T et al. Integration of murine leukemia virus DNA depends identified by their FL-1 autofluorescence on a FACScan on mitosis. EMBO J 1993; 12: 2099–2108. (Becton Dickinson, San Jose, CA, USA). 16 Miller DG, Adam MA, Miller AD. Gene transfer by retrovirus vectors occurs only in cells that are actively replicating at the time of infection. Mol Cell Biol 1990; 10: 4239–4242. Monoclonal antibodies and cell surface staining 17 Lewis PF, Emerman M. Passage through mitosis is required for PE-Cy5-conjugated anti-CD4 and anti-CD8 antibodies oncoretroviruses, but not for the human immunodeficiency were obtained from Immunotech (Marseille, France). virus. J Virol 1994; 68: 510–516. Cells were incubated with antibodies for 20 min and then 18 Miyoshi H et al. Stable and efficient gene transfer into the retina washed in PBS before FACS analysis. using an HIV-based lentiviral vector. Proc Natl Acad Sci USA 1997; 94: 10319–10323. 19 Naldini L et al. In vivo gene delivery and stable transduction of Acknowledgements nondividing cells by a lentiviral vector. Science 1996; 272: 263– We are grateful to M Sitbon for critical insights during 267. the course of this work. Dr Ikunoshin Kato and Setsuko 20 Poeschla E, Corbeau P, Wong-Staal F. Development of HIV vec- Yoshimura of Takara Shuzo Co Ltd are generously tors for anti-HIV gene therapy. Proc Natl Acad Sci USA 1996; 93: acknowledged for providing the recombinant fibronectin 11395–11399. fragment and for their continuing assistance. We are 21 Uchida N et al. HIV, but not murine leukemia virus, vectors mediate high efficiency gene transfer into freshly isolated indebted to Drs C June, D Cantrell and S Nagata for gen- G0/G1 human hematopoietic stem cells. Proc Natl Acad Sci USA erously providing reagents. Supported by grants from 1998; 95: 11939–11944. the ANRS and AFM (NT and PC), March of Dimes Grant 22 Zufferey R et al. Multiply attenuated lentiviral vector achieves No. 6.FY99-406 and ARC (NT). efficient gene delivery in vivo. Nat Biotechnol 1997; 15: 871–875. 23 Reiser J et al. Transduction of nondividing cells using pseudo- References typed defective high-titer HIV type 1 particles. Proc Natl Acad Sci USA 1996; 93: 15266–15271. 1 Plavec I et al. High transdominant RevM10 protein levels are 24 Unutmaz D et al. Cytokine signals are sufficient for HIV-1 infec- required to inhibit HIV-1 replication in cell lines and primary tion of resting human T lymphocytes. J Exp Med 1999; 189: T cells: implication for gene therapy of AIDS. Gene Therapy 1997; 1735–1746. 4: 128–139. 25 Zack JA. The role of the cell cycle in HIV-1 infection. Adv Exp 2 Roth T et al. Synthesis and biological activity of novel non- Med Biol 1995; 374: 27–31. nucleoside inhibitors of HIV-1 reverse transcriptase. 2-Aryl-sub- 26 Charneau P, Alizon M, Clavel F. A second origin of DNA plus- stituted benzimidazoles. J Med Chem 1997; 40: 4199–4207. strand synthesis is required for optimal human immunodefi- 3 Hwu P et al. Lysis of ovarian cancer cells by human lymphocytes ciency virus replication. J Virol 1992; 66: 2814–2820. redirected with a chimeric gene composed of an antibody vari- 27 Charneau P et al. HIV-1 reverse transcription. A termination step able region and the Fc receptor gamma chain. J Exp Med 1993; at the center of the genome. J Mol Biol 1994; 241: 651–662. 178: 361–366. 28 Zennou V et al. HIV-1 genome nuclear import is mediated by a 4 Chinen J et al. Protection of primary human T cells from HIV central DNA flap. Cell 2000; 101: 173–185. infection by Trev: a transdominant fusion gene. Hum Gene Ther 29 Sirven A et al. The HIV-1 central DNA flap is a crucial determi- 1997; 8: 861–868. nant for lentiviral vector nuclear import and gene transduction 5 Blaese RM et al. T lymphocyte-directed gene therapy for ADA- Blood 96 SCID: initial trial results after 4 years. Science 1995; 270: 475–480. of human hematopoietic stem cells. 2000; : 4103–4110. 6 Bordignon C et al. Gene therapy in peripheral blood lympho- 30 Follenzi A et al. Gene transfer by lentiviral vectors is limited by cytes and bone marrow for ADA-immunodeficient patients. nuclear translocation and rescued by HIV-1 pol sequences. Nat Genet 2000; 25: 217–222. Science 1995; 270: 470–475. + 7 Kohn DB et al. T lymphocytes with a normal ADA gene accumu- 31 Uckert W et al. Efficient gene transfer into primary human CD8 late after transplantation of transduced autologous umbilical T lymphocytes by MuLV-10A1 retrovirus pseudotype. Hum cord blood CD34+ cells in ADA-deficient SCID neonates. Nature Gene Ther 2000; 11: 1005–1014. Med 1998; 4: 775–780. 32 Uetsuki T et al. Isolation and characterization of the human 8 Mavilio F et al. Peripheral blood lymphocytes as target cells of chromosomal gene for polypeptide chain elongation factor-1 retroviral vector-mediated gene transfer. Blood 1994; 83: 1988– alpha. J Biol Chem 1989; 264: 5791–5798. 1997. 33 Quinn ER, Lum LG, Trevor KT. T cell activation modulates 9 Bunnell BA et al. High-efficiency retroviral-mediated gene trans- retrovirus-mediated gene expression. Hum Gene Ther 1998; 9: fer into human and nonhuman primate peripheral blood lym- 1457–1467. phocytes. Proc Natl Acad Sci USA 1995; 92: 7739–7743. 34 Agarwal M et al. Scaffold attachment region-mediated enhance-

Gene Therapy Lentiviral gene transfer in primary T cells V Dardalhon et al 198 ment of retroviral vector expression in primary T cells. J Virol 42 Levine BL et al. CD28 ligands CD80 (B7–1) and CD86 (B7–2) 1998; 72: 3720–3728. induce long-term autocrine growth of CD4+ T cells and induce 35 Douglas J et al. Efficient transduction of human lymphocytes similar patterns of cytokine secretion in vitro. Int Immunol 1995; and CD34+ cells via human immunodeficiency virus-based gene 7: 891–904. transfer vectors. Hum Gene Ther 1999; 10: 935–945. 43 Costello E et al. Gene transfer into stimulated and unstimulated 36 Page KA, Landau NR, Littman DR. Construction and use of a T lymphocytes by HIV-1-derived lentiviral vectors. Gene Therapy human immunodeficiency virus vector for analysis of virus 2000; 7: 596–604. infectivity. J Virol 1990; 64: 5270–5276. 44 Kung SK, An DS, Chen IS. A murine leukemia virus (MuLV) 37 Zufferey R et al. Self-inactivating lentiviral vector for safe and long terminal repeat derived from rhesus macaques in the con- efficient in vivo gene delivery. Multiply attenuated lentiviral text of a lentivirus vector and MuLV gag sequence results in vector achieves efficient gene delivery in vivo. J Virol 1998; 72: high-level gene expression in human T lymphocytes. J Virol 9873–9880. 2000; 74: 3668–3681. 38 Korin YD, Zack JA. Progression to the G1b phase of the cell 45 Pollok KE et al. Costimulation of transduced T lymphocytes via cycle is required for completion of human immunodeficiency T cell receptor–CD3 complex and CD28 leads to increased tran- virus type 1 reverse transcription in T cells. J Virol 1998; 72: scription of integrated retrovirus. Hum Gene Ther 1999; 10: 3161–3168. 2221–2236. 39 Case SS et al. Stable transduction of quiescent CD34(+)CD38(−) 46 Yee JK et al. A general method for the generation of high-titer, human hematopoietic cells by HIV-1-based lentiviral vectors. pantropic retroviral vectors: highly efficient infection of primary Proc Natl Acad Sci USA 1999; 96: 2988–2993. hepatocytes. Proc Natl Acad Sci USA 1994; 91: 9564–9568. 40 Wells AD, Gudmundsdottir H, Turka LA. Following the fate of 47 Dardalhon V et al. Highly efficient gene transfer in naive human individual T cells throughout activation and clonal expansion. T cells with a murine leukemia virus-based vector. Blood 2000; Signals from T cell receptor and CD28 differentially regulate the 96: 885–893. induction and duration of a proliferative response. J Clin Invest 48 Moritz T, Patel VP, Williams DA. Bone marrow extracellular 1997; 100: 3173–3183. matrix molecules improve gene transfer into human hematopo- 41 Riley JL et al. Naive and memory CD4 T cells differ in their ietic cells via retroviral vectors. J Clin Invest 1994; 93: 1451–1457. susceptibilities to human immunodeficiency virus type 1 infec- 49 Kimizuka F et al. Production and characterization of functional tion following CD28 costimulation: implications for trans- domains of human fibronectin expressed in Escherichia coli. J mission and pathogenesis. J Virol 1998; 72: 8273–8280. Biochem 1991; 110: 284–291.

Gene Therapy