Proc. Natl. Acad. Sci. USA Vol. 93, pp. 11842-11847, October 1996 Medical Sciences

Efficient infection of a human T-cell line and of human primary peripheral blood leukocytes with a pseudotyped retrovirus vector SANJAI SHARMA*, MARK CANTWELLt, THOMAS J. Kippst, AND THEODORE FRIEDMANN*t Departments of *Pediatrics and tMedicine, University of California School of Medicine at San Diego, La Jolla, CA 92093-0634 Communicated by Helen M. Ranney, Alliance Pharmaceutical Corp., San Diego, CA, July 29, 1996 (received for review April 26, 1996)

ABSTRACT Peripheral blood lymphocytes (PBLs) are an permit highly efficient infection of CD4+ T cells and other important target for gene transfer studies aimed at human classes of lymphocytes in vitro, and possibly in vivo. Studies gene therapy. However, no reproducibly efficient methods are toward this goal have included a number of cell culturing currently available to transfer foreign, potentially therapeutic manipulations to improve transduction efficiency, but such genes into these cells. While vectors derived from murine approaches have limited utility for many potential applica- retroviruses have been the most widely used system, their low tions, including in vivo gene transfer. More efficient applica- infection efficiency in lymphocytes has required prolonged in tion of retroviruses to gene transfer into lymphocytes would vitro culturing and selection after infection to obtain useful also benefit from improved retrovirus vector technology, numbers of genetically modified cells. We previously reported including the preparation of vectors to much higher titers than that retroviral vectors pseudotyped with vesicular stomatitis currently possible and the design of vectors able to infect G glycoprotein (VSV-G) envelope can infect a wide variety of quiescent cells. cell types and can be concentrated to titers of greater than 109 Our laboratory has recently taken a step toward overcoming infectious units/ml. In this present study, we examined the one of these technical barriers-i.e., that of low retrovirus vector ability of amphotropic and pseudotyped vectors expressing a titers. We have shown that the env gene product of Moloney murine cell surface protein, B7-1, to infect the human T-cell murine leukemia virus (MoMLV) can be replaced by the G line Jurkat or human blood lymphocytes. Limiting dilution glycoprotein of vesicular stomatitis virus (VSV-G) (12, 13) to analysis of transduced Jurkat cells demonstrated that the produce vectors with a markedly enhanced host range and pseudotyped vector is significantly more efficient in infecting expanded tissue tropism. Because the cellular receptors for T cells than an amphotropic vector used at the same multiplicity VSV-G, possibly including phosphatidylserine, phosphatidylino- of infection (moi). To identify the transduction efficiency on sitol, and GM3 ganglioside (14, 15) appear to be very abundant PBLs, we examined the levels of cell surface expression of the and ubiquitous membrane components of most mammalian cells, B7-1 surface marker 48 to 72 hr after infection. The transduction the mechanisms of cell entry and infection with the pseudotyped efficiency ofPBLs with the pseudotyped vector increased linearly vectors are entirely different from those associated with tradi- with increasing moi to a maximum of approximately 16-32% at tional amphotropic retrovirus vectors. Most important, however, an moi of40. This relatively high efficiency ofinfection ofa T-cell is the fact that the VSV-G pseudotyped virus can be concentrated line and of blood lymphocytes with VSV-G pseudotyped virus to high titers, thereby permitting the use of much higher values of demonstrates that such modified pseudotyped retrovirus vectors multiplicity of infection (moi) than have been previously possible. may be useful reagents for studies of gene therapy for a variety In the present study, we present evidence that a T-cell line and of genetic or neoplastic disorders. primary PBLs are readily susceptible to infection with pseudotyped virus and a high percentage of cells express the T lymphocytes are relatively long-lived (1) cells of the hema- transgene. topoietic system that regulate immune responses to infectious or neoplastic disease. T-cell function is impaired in a variety of MATERIALS AND METHODS genetic or acquired diseases that affect lymphocytes, resulting in immunodeficiency, severe chronic infections, or neoplasia. Plasmids. The cDNA encoding the murine immune acces- Since most circulating lymphocytes are T cells, the peripheral sory molecule B7-1 (muB7) cDNA, a gift from James Allison blood lymphocyte compartment is an important target for (University of California, Berkeley), was subcloned into genetic modification. Unfortunately, T cells are refractory to pBluescript polylinker. A HindIII-XbaI fragment with the most current viral and non-viral gene transfer techniques. Only muB7 cDNA was then blunt- ended and inserted into the retrovirus vectors have been found useful in transferring blunt-ended BamHI site of vector plasmid pLPONL (16), foreign, potentially therapeutic genes into a variety of lym- which expresses the neomycin phosphotransferase gene from phocytes in vitro, including peripheral blood lymphocytes the Moloney retroviral 5' long terminal repeat by virtue of the (PBLs) and tumor-infiltrating lymphocytes (2-8). This has led presence of a poliovirus ribosome reentry sequence (PO). The to several clinical gene therapy studies involving gene transfer resulting plasmid was designated pLPONL-muB7. into such cells (6-11). Retrovirus Vector Preparation. Vector pLPONL-muB7 The recent results with adenosine deaminase deficiency plasmid DNA was transfected into the packaging cell line (ADA) reconstitution studies demonstrated that a limited PA317 (17) by -using the established calcium phosphate trans- degree of gene transfer and prolonged gene expression in fection method (18). Forty-eight hours after transfection, the lymphocytes are possible with current retroviral vectors. How- culture medium containing amphotropic LPONL-muB7 virus ever, prolonged periods ofin vitro culture of infected cells were was used to infect 293GP cells, which express high levels of required to obtain useful numbers of genetically modified cells Moloney murine leukemia virus gag and pol (12). After because of the poor transduction efficiencies of retroviruses. This mandates improvements in gene transfer technology to Abbreviations: PBL, peripheral blood lymphocyte; VSV-G, G glyco- protein of vesicular stomatitis virus; moi, multiplicity of infection; muB7, murine immune accessory molecule B7-1; IL-2, interleukin 2; The publication costs of this article were defrayed in part by page charge FACS, fluorescence-activated cell sorter; PE, phycoerythrin; FITC, payment. This article must therefore be hereby marked "advertisement" in fluorescein isothiocyanate; TE, transduction efficiency. accordance with 18 U.S.C. §1734 solely to indicate this fact. tTo whom reprint requests should be addressed. 11842 Downloaded by guest on October 2, 2021 Medical Sciences: Sharma et al. Proc. Natl. Acad. Sci. USA 93 (1996) 11843

selection in G418-containing medium (400 ,tg/ml) for 2 weeks, FACScan (Becton Dickinson). Other antibodies used to char- colonies were picked and the clones (muB7/293GP cells) acterize the phenotype of infected cells included an anti-CD3 expressing high levels of the muB7 transgene, as assessed by antibody [clone UCHT-1, fluorescein isothiocyanate (FITC) flow cytometric analysis (see below), were pooled. conjugate, Sigma], an anti-CD4 antibody (clone Q4120, FITC The preparation of high-titer VSV-G-pseudotyped LPONL- conjugate, Sigma), and an anti-CD45RO antibody (clone muB7 vector was carried out by methods described previously UCHL-1 FITC conjugate, Sigma). Cells were stained with (13). The muB7/293GP cells were transfected with plasmid propidium iodide (1 ,ug/ml) to label dead cells, allowing for pHCMV-G expressing VSV-G from the human cytomegalo- their exclusion from FACS analysis. virus (HCMV) promoter (13) by using the calcium phosphate Limiting Dilution Analysis of Infected Cells. For limiting method. Culture media were collected at 48 and 72 hr after dilution analysis (19), cells were infected with the amphotropic transfection, filtered through a 0.45-,um pore diameter filter, and pseudotyped vectors at an moi of 1 as described above, and and concentrated by ultracentrifugation as described (13). after 3 days the cells were replated in complete medium at a Virus was titered on 208F cells by infection with virus dilutions density of 100, 25, 10, or 1 cell per well in 96-well plates. in the presence of Polybrene (8 ,ug/ml) followed by G418 Duplicate cultures either were maintained in complete growth selection (400 ,g/ml) starting 24 hr after infection. G418- medium or were grown under selective conditions consisting of resistant colonies were counted 10-12 days after infection. complete medium containing G418 at 800 ,ug/ml. Wells with Aliquots of virus stocks were frozen at -70°C. To obtain viable cells were counted 2 weeks after plating. retrovirus enveloped with an amphotropic envelope, plasmid Efficiency of Transduction. The transduction efficiency was pHCMV-ampho-env was constructed by excising the BamHI calculated by dividing the negative natural log of the G418- fragment VSV-G cDNA from pHCMV-G and replacing it with selected fraction of wells that did not contain cells by the a 2.8-kb EagI fragment from plasmid PA14 (from A. Rein, number of cells plated in that well (19). For instance, in wells National Cancer Institute-Frederick Cancer Research and containing 25 cells per well, 100% of the wells were positive in Development Center) containing the complete amphotropic the absence of selection, while 90% ofwells were positive in the envelope gene. Plasmid pHCMV-ampho-env was also trans- presence of selection. The negative natural log of the nonse- fected into muB7/293GP cells and culture medium was col- lected wells (10% or 0.1) is 2.3, and this number divided by 25 lected 48 hr thereafter, filtered through a 0.45-,um filter, and (cells per well plated) provides a transduction efficiency (TE) stored at -70°C. The amphotropic virus was also titered by of 0.092 or 9.2%. infecting rat 208F cells, as described above. Cell Lines and Primary Lymphocytes. Jurkat cells were grown in complete medium [RPMI medium 1640 with 10% RESULTS fetal bovine serum and 10 mM Hepes (Sigma)], at 37°C in a 5% To develop a rapid and reliable assay for determining retro- CO2 incubator. Human PBLs were isolated from the buffy coat viral infection, we developed a retroviral vector encoding of donors by Ficoll-Hypaque gradient centrifugation (His- muB7. The cell surface expression of this muB7 costimulatory topaque 1077, Sigma). PBLs were cultured in complete me- molecule can be detected with a monoclonal antibody and dium supplemented with human recombinant interleukin 2 flow-cytometric analysis, thereby allowing quantitative deter- (IL-2; Sigma) at 100 units/ml. An enrichment column for mination of the efficiency of gene transfer and gene expres- CD4+ cells (R & D Systems) was used to isolate human CD4+ sion. The retrovirus producer cell line muB7/293GP was cells from PBLs by methods specified by the manufacturer. transfected with expression plasmids expressing either ampho- Infection of the Jurkat T-Cell Line. A total of 5 x 105 Jurkat tropic or VSV-G envelope. Titers of amphotropic virus ranged cells in complete medium were plated into 6-well plates in a from 1 to 3 x 105 colony-forming units/ml and that of total volume of 2 ml of medium per well. Twenty-four hours concentrated VSV-G pseudotyped virus from 2 to 4 x 108 after plating, the cells were infected by addition of ampho- colony-forming units/ml. tropic or VSV-G-pseudotyped LPONL-muB7 vector in the Infection of Jurkat Cells. As it is not feasible to concentrate presence of Polybrene at a final concentration of 8 ,ug/ml. The the amphotropic enveloped virus to titers comparable to those efficiency of vector infection was established by determination of the pseudotyped form, we initially infected the Jurkat cells of the number of wells with viable cells as assayed by limiting with the amphotropic vector and with nonconcentrated dilution analysis (see below) 2 weeks after infection. To pseudotyped vector, both at an moi of 1. Table 1 presents the determine the effect of different mois on infection efficiency, results of limiting dilution assays of two independent infections Jurkat cells were infected with dilutions of pseudotyped virus after maintenance of infected cells with and without G418. and analyzed 48 hr later for cell surface muB7 expression, as Under these conditions, the TE of the amphotropic virus was assayed by a fluorescence-activated cell sorter (FACS). 0.6% + 0.1%. In contrast, the pseudotyped vector had a TE of Infection of PBLs. Primary PBLs were cultured in complete 11.95% + 0.6%, a 22-fold difference. medium containing human recombinant IL-2 (100 units/ml) Because of its high titer, we were able to infect Jurkat cells for 48 hr. Cells were counted 48 hr after initial culturing and with the pseudotyped vector at higher moi values than was then replated at a density of 5 x 105 or 106 cells per ml in a possible with the amphotropic vector. We carried out the 12-well dish with 1 ml of complete culture medium per well. infection at an moi of 2, 10, 20, or 50 with the pseudotyped Virus and phytohemagglutinin (2 ,ug/ml) were added together retrovirus LPONL-muB7. The experiment was performed at the time of replating. Medium was changed 24 hr after three times, and Fig. 1 illustrates the percentage of cells infection to complete medium containing 100 units of recom- expressing the muB7 marker as a function of moi. The binant IL-2 per ml. Cells were stained at 48 or 72 hr after percentage of Jurkat cells expressing the muB7 transgene infection with the anti-muB7 antibody to identify infected cells increased linearly with respect to moi. At an moi of 2, muB7 by FACS detection of cell surface B7-1 expression, as de- marker was expressed on 2% of cells, and expression reached scribed below. a mean maximum of 18% at an moi of 50. We did not examine FACS Analysis of Gene Expression. For FACS analysis of mois greater than 50. B7-1 expression, the cells were collected from the plates, To determine the stability of transgene expression, we washed once with phosphate-buffered saline, and stained with infected Jurkat cells with pseudotyped LPONL-muB7 at an a phycoerythrin-conjugated murine anti-B7-1 antibody desig- moi of 20 and analyzed the muB7 expression at various times nated 1G1G-PE (phycoerythrin-conjugated, PharMingen), or a after infection. Transgene expression was detected at an early phycoerythrin-conjugated isotype control antibody (IgG-a PE; (36-hr) time point and persisted throughout the study period Caltag, South San Francisco, CA). Cells were analyzed on a of 11 days, with 20-38% of Jurkat cells expressing the trans- Downloaded by guest on October 2, 2021 11844 Medical Sciences: Sharma et al. Proc. Natl. Acad. Sci. USA 93 (1996)

Table 1. Limiting dilution analysis and efficiency of transduction Exp. 1 Exp. 2 % of wells positive % of wells positive Virus and Without With Without With cell density G418 G418 TE, % G418 G418 TE, % Amphotropic virus 100 cells per well - 100 36 0.4 25 cells per well 100 10 0.4 100 10 0.4 10 cells per well 100 0 - 96 3 0.9 1 cell per well 53 0 - 46 0 Pseudotyped virus 100 cells per well 100 100 25 cells per well 100 90 9 100 96 13 10 cells per well 100 70 12 100 73 13 1 cell per well 46 7 ll 53 10 13 Jurkat cells were infected at an moi of 1 with amphotropic and VSV-G pseudotyped viruses and plated into 96-well plates at 100, 25, 10, or 1 cell per well with and without G418 selection. Wells with viable cells were counted 2 weeks after starting selection. TE was calculated as described in ref. 19. A - indicates not done.

gene (data not shown). At all time points, the mean fluores- PE-conjugated anti-B7 antibody to identify T cells and the T cence of infected cells was 3- to 4-fold greater than that of cells that express muB7, respectively. Since the great majority noninfected cells (data not shown). of cells were CD3+ (90-95%) after IL-2 and phytohemagglu- Infection of PBLs and CD4+ Cells. Ninety-five percent of tinin stimulation (Table 2), the overall infection efficiency is cells isolated by means of the CD4 enrichment column ex- likely to reflect the percentage of viable CD3+ T cells infected pressed CD4, as assessed by flow cytometry. Total PBLs from by the vector. Infection efficiency varied from 16% to 32%. normal donors, or isolated CD4+ cells, were stimulated with Fig. 3 shows representative FACS profiles indicating expres- IL-2 and phytohemagglutinin, as described in Materials and sion of muB7 and CD3 by control (A) or infected (B) cells. Methods. These cells then were infected with pseudotyped Ninety-five percent of the cells stained with the anti-CD3 muB7 retrovirus at an moi of 5, 20, or 40 (Fig. 2). The monoclonal antibody (upper and lower right quadrants). Of percentage of cells expressing the muB7 transgene 48 hr after the CD3+ infected cells, 32% expressed muB7 (30.6% of the infection increased linearly with increasing moi, reaching a total). From this we conclude that 32% of the CD3+ cells maximum of 25-30% at an moi of 40 (Table 2 and Fig. 3). express muB7 48 hr after infection. Table 2 presents the multiparameter FACS data on four We examined the infected cells for expression of CD45RO, different PBL cell preparations infected at an moi of 40 with a surface marker associated with T cells (memory phenotype) pseudotyped LPONL-muB7 vector and analyzed 48 hr (PBL 3 that have an extended life-span in vivo and that demonstrate and 4) or 72 hr (PBL 1 and 2) after infection. Cells were a tropism for sites of active tissue inflammation in organs such costained with an FITC-conjugated anti-CD3 antibody and a as the lung, liver, or intestine (20). Preparations of PBL were infected with muB7 virus at an moi of 40, and then examined 30 48 hr later for expression of CD45RO and muB7. FACS analysis revealed that approximately 73% of cells were CD45RO positive. CD45RO and muB7 were coexpressed by 13% of cells, indicating that these cells were infected at an average infection efficiency of 18% (data not shown).

0 20 DISCUSSION IN Techniques for efficient and stable gene transfer into T cells 60 have not been generally available. Although retrovirus vectors have already been used in a number of gene transfer studies in U, T cells, prolonged periods of in vitro culture of infected cells 10 have been required to obtain useful numbers of genetically x modified cells for clinical application because of the poor

C,, transduction efficiencies of standard amphotropic retrovi- ruses. A central problem with previous reports of lymphocyte u infection with retrovirus vectors, however, stems from the use in different studies of markedly different assays for the detec- tion of infection as well as the use of vectors of different design 0 and mechanisms of gene expression. Early studies have re- 0 10 20 30 40 50 60 Table 2. Infection efficiency in PBLs multiplicity of infection PBL Starting cell no. % CD3+ % muB7+ preparation preinfection Final cell no. cells cells FIG. 1. Efficiency of infection and stability of gene expression in 1 5 x 105 6.0 X 105 (72 hr) 93 32 Jurkat cells infected with pLPONL-muB7 vector. The percentage of 2 5 x 105 10.8 x 105 (72hr) 95 16 exposed cells that express the muB7 gene product 48 hr after infection 3 106 9.2 x 105 (48 hr) 26 at mois of 2, 10, 20, and 50 is shown. Percentages represent mean values 4 106 9.0 x 105 (48 hr) - 31 and SDs of results from three experiments. Downloaded by guest on October 2, 2021 Medical Sciences: Sharma et al. Proc. Natl. Acad. Sci. USA 93 (1996) 11845

-0--- PBL 1 30- ...... 0-...... CD4+ cells

-.S n 20 - xl~

O-

10 -

I I I I I 0 10 20 30 40 50

multiplicity of infection

FIG. 2. Infection efficiency of a preparation of human PBLs and a preparation of enriched CD4+ cells (95% cells CD4+) at mois of 5, 20, and 40 with the pseudotyped vector LPONL-muB7. FACS analysis of muB7 expression was carried out 48 hr after infection.

ported infection efficiencies of T cells with amphotropic virus 40. In the present study, we demonstrate that the increased vectors up to 10-20% (4, 5). However, quantitative studies infection frequency in T cells of the VSV-G pseudotyped using FACS analysis to detect expression of the transgene vectors compared with amphotropic vectors is likely to stem product have reported efficiencies of only 1% per infection from two independent factors. The first involves the much cycle (5). Higher estimates have been derived from nonquan- higher titers and therefore the higher mois that are possible titative experiments (4), while still others have been obtained with the pseudotyped vectors compared with traditional am- from studies that required the use of extremely sensitive PCR photropic vectors that make possible the exposure of target methods to detect the presence, but not the expression, of cells to higher concentrations of vector. VSV-G-pseudotyped vector sequences (21). In a recent report, a retroviral vector retrovirus vectors can be concentrated to titers in the range of containing the gibbon leukemia virus envelope glycoprotein 109 colony-forming units/ml because of the stability of the instead of the Moloney leukemia virus counterpart was found VSV-G-containing enveloped virus particles to centrifugal to infect peripheral blood lymphocytes from human and forces, in contrast to the marked instability of traditional nonhuman primate sources, particularly under conditions of amphotropic retrovirus vectors to such conditions. phosphate depletion of the medium (21). The second factor is the different and apparently more A recent report (22) indicates that the infection efficiency efficient mechanism of virus entry into exposed cells of the with retroviruses in general is dependent on the concentration VSV-G pseudotyped vectors compared with amphotropic of retrovirus particles rather than on the moi per se, consistent vectors. Infection by VSV-G pseudotyped vectors is indepen- with the high efficiency of infection in our study at an moi of dent of the presence or exposure of retrovirus receptors on the

A B

CD3 FITC o CD3 FITC lo

FIG. 3. FACS analysis of human PBLs for expression of CD3 and of the marker muB7 gene. Results with noninfected control cells (A) and with PBLs infected with the pseudotyped muB7 vector at an moi of 40 (B) are presented. In both cases, cells were examined by FACS analysis 48 hr after infection. Downloaded by guest on October 2, 2021 11846 Medical Sciences: Sharma et al. Proc. Natl. Acad. Sci. USA 93 (1996) cell surface and seems to be related to other, more abundant, Since the eventual aim of all gene transfer studies is effi- and possibly more readily exposed components of the cells' ciency of both gene transfer and gene expression, the results lipid bilayer. A comparison of the transduction efficiency of described in the present study represent a useful and realistic the pseudotyped vector with that of the amphotropic retroviral quantitative assessment of both these parameters in lympho- vector at equal mois of 1 indicates that the VSV-G virus is cytes transduced with pseudotyped retrovirus vectors. markedly better able to infect Jurkat cells. Our data also provide evidence that the subset of CD45RO+ We have used two independent methods to estimate the T cells is also susceptible to infection with VSV-G- transduction efficiency of lymphocytes with pseudotyped ret- pseudotyped retrovirus vectors. In a recent report, mature rovirus vectors: a limiting dilution assay to detect G418- human T-cell preparations containing equal numbers of mem- resistant colonies and a FACS analysis of muB7 expression. ory (CD45RO+) T cells and virgin or naive T cells Interestingly, the FACS analysis fails to detect transduced (CD45RA+) were injected into severe combined immunod- Jurkat cells at an moi of 2, while the limiting dilution assay eficient (SCID) mice (20). When these SCID mice were indicates that an moi of 1 provides an infection efficiency of analyzed for the phenotype of the populating human T cells, 12% (Fig. 1, Table 1). This discrepancy may be due to the only the memory (CD45RO+) T cells were detected, indicating relative insensitivity of the FACS assay or the expression that this phenotype is involved in long-term in vivo repopula- level(s) of the different reporter genes, underscoring the tion of hosts. Since our studies show that these cells are difficulties of comparing infection frequencies in studies using susceptible to retrovirus infection, it becomes possible to different detection methods and different vectors. Neverthe- examine the effect of genetic modification of these cells on less, the FACS assay of transgene expression provides a rapid reconstitution studies in the SCID mouse model. measurement of the efficiency of infection and has advantages Obviously, many cell biology aspects of lymphocyte recon- over Southern blot analysis or semiquantitative PCR (4, 6, 7, stitution in gene therapy studies in humans still need to be 21), both of which provide information only on the presence, identified and solved. Nevertheless, we propose that some not on the expression, of the transgene in the infected cells. issues related to poor initial infection efficiency in lymphocytes Furthermore, cells in Southern assays must be grown to may begin to be resolved through the use of modified retro- sufficient numbers for DNA isolation, thereby allowing time virus vectors, such as the VSV-G pseudotyped vectors, that for selection of subpopulations of cells with growth advantages increased efficiency of infection, thereby making it and thereby possibly leading to errors in the estimation of permit initial transduction efficiency. easier to develop new approaches to gene therapy of some Although a direct comparison of the optimum conditions for genetic deficiency disorders and of acquired defects of T cells. T-cell infection with amphotropic virus and with the pseudotyped These studies were supported by Grants CA66000, R37 CA49870, virus would be desirable, it has not been technically feasible under HD20037, HL53680, and DK49023 from the National Institutes of the conditions of our experiments. The low titer of amphotropic Health. vectors prevents infection of cells at the very high mois that are achievable with the pseudotypedvectors, and the addition oflarge 1. Mclean, A. R. & Michie, C. A. (1995) Proc Natl. Acad. Sci. USA volumes of amphotropic vector to increase apparent mois would 92, 3707-3711. not allow high concentrations of vector to come into intimate 2. Friedmann, T. (1989) Science 244, 1275-1281. enough contact with the target cells to allow infection. In addi- 3. Culver, K., Cornetta, K., Morgan, R., Morecki, S., Aebersold, P., tion, quantitative estimates of infection efficiency by FACS Kasid, A., Lotze, M., Rosenberg, S. A., Anderson, W. F. & analysis under the conditions of these studies requires mois of Blaese, R. M. (1991) Proc. Natl. Acad. Sci. USA 88, 3155-3159. greater than 2, as indicated by our finding that infection of Jurkat 4. Kasid, A., Morecki, S., Aebersold, P., Cornetta, K., Culver, K., cells at an moi of 2 results in detectable expression of muB7 in Freeman, S., Director, E., Lotze, M. T., Blaese, R. M. & Ander- son, W. F. (1990) Proc. Natl. Acad. Sci. USA 87, 473-477. only 2% of cells. Titers of amphotropic vectors sufficiently high 5. Mavilio, F., Ferrari, G., Rossini, S., Nobili, N., Bonini, C., to take advantage of the linear dose response of infection Casorati, G., Traversari, C. & Bordignon, C. (1994) Blood 83, efficiency in T lymphocytes or CD4+ T cells (Fig. 2) are generally 1988-1997. not possible. 6. Bordignon, C., Notarangelo, L. D., Nobili, N., Ferrari, G., Ca- Because of markedly different conditions for cell prepara- sorati, G., Panina, P., Mazzolari, E., Maggioni, D., Rossi, C., tion and the use of different and often nonquantitative meth- Servida, P., Ugazio, A. G. & Mavilio, F. (1995) Science 270, ods for determining infection, it is not possible to provide 470-475. accurate quantitative comparisons of T-cell transductions re- 7. Blaese, R. M., Culver, K. W., Miller, A. D., Carter, C. S., ported in earlier studies with those of the present study. In our Fleisher, T., Clerici, M., Shearer, G., Chang, L., Chiang, Y., study, the infection efficiency of PBLs has been determined by Tolstoshev, P., Greenblatt, J. J., Rosenberg, S. A., Klein, H., a quantitative method for detecting expressed transgenes, and Berger, M., Mullen, C. A., Ramsey, W. J., Muul, L., Morgan, to after R. A. & Anderson, W. F. (1995) Science 270, 475-480. the efficiencies have ranged up 16-32% single cycles 8. Ojwang, J. O., Hampel, A., Looney, D. J., Wong-Staal, F. & of infection. What is clear, furthermore, from the present work Rappaport, J. (1992) Proc. Natl. Acad. Sci. USA 89, 10802-10806. is that VSV-G-pseudotyped retrovirus vectors do permit much 9. Matsuda, Z., Yu, X., Yu, Q. C., Lee, T. H. & Essex, M. (1993) more efficient infection ofT cells than is possible with identical Proc. Natl. Acad. Sci. USA 90, 3544-3548. vectors in an amphotropic envelope. The precise mechanisms 10. Chen, S. Y., Khouri, Y., Bagley, J. & Marasco, W. A. (1994) Proc. responsible for enhanced infection of lymphocytes with the Natl. Acad. Sci. USA 91, 5932-5936. pseudotyped retrovirus vectors are not entirely clear. Virus 11. Sun, L. Q., Pyati, J., Smythe, J., Wang, L., Macpherson, J., titers determined on one cell type might not be entirely Gerlach, W. & Symonds, G. (1995) Proc. Natl. Acad. Sci. USA 92, relevant to effective titers on another target cell type, since the 7272-7276. availability of appropriate retrovirus receptors or phosphati- 12. Burns, J. C., Friedmann, T., Driever, W., Burrascano, M. & Yee, dylserine or other VSV-G receptors may be entirely different J. K. (1993) Proc. Natl. Acad. Sci. USA 90, 8033-8037. case 13. Yee, J. K., Friedmann, T. & Burns, J. C. (1994) Methods Cell Biol. in the different cell types. In the of primary hepatocytes, 43, 99-112. which we know to be more readily infected with pseudotyped 14. Schlegel, R., Tralka, T. S., Willingham, M. C. & Pastan, I. (1983) vectors than with traditional Moloney virus-based vectors, the Cell 32, 639-646. enhanced infection is not associated with increased copy 15. Mastromarino, P., Conti, C., Goldoni, P., Hauttecoeur, B. & number of integrated viral genomes. We suspect that, in that Orsi, N. (1987) J. Gen. Virol. 68, 2359-2369. case, the increased transgene expression from the pseudotyped 16. Yee, J. K., Miyanohara, A., LaPorte, P., Bouic, K., Burns, J. C. & vector results from increased numbers of cells being infected. Friedmann, T. (1994) Proc. Natl. Acad. Sci. USA 91, 9564-9568. Downloaded by guest on October 2, 2021 Medical Sciences: Sharma et al. Proc. Natl. Acad. Sci. USA 93 (1996) 11847

17. Miller, A. D. & Buttimore, C. (1986) Mol. Cell. Biol. 6, 2895- 20. Tary-Lehmann, M. & Saxon, A. (1992)J. Exp. Med. 175, 503-516. 2902. 21. 18. Chen, C. & Okayama, H. (1987) Mol. Cell. Biol. 7, 2745-2752. Bunnel, B. A., Muul, L. M., Donahue, R. E., Blaese, R. M. & 19. Henry, C., Marbrook, J., Vann, D. C., Kodlin, D. & Wofsy, C. Morgan, R. (1995) Proc. Natl. Acad. Sci. USA 92, 7739-7743. (1980) in Selected Methods in Cellular Immunology, eds. Mishell, 22. Morgan, J. R., Ledoux, J. M., Snow, R. G., Tompkins, R. G. & B. B. & Shiigi, S. M. (Freeman, San Francisco), pp. 138-152. Yarmush, M. L. (1995) J. Virol. 69, 6994-7000. Downloaded by guest on October 2, 2021