and man - Europe PMC

0 downloads 0 Views 1MB Size Report
Dec 31, 1990 - helper T cells were transduced with the retroviral vector SAX to express both neomycin-resistance and human adenosine deaminase genes.
Proc. Nati. Acad. Sci. USA Vol. 88, pp. 3155-3159, April 1991 Medical Sciences

Lymphocytes as cellular vehicles for gene therapy in mouse and man (retroviral vector/tumor-infilltrating lymphocyte/interleukin 2/neomycin resistance/adenosine deaminase)

KENNETH CULVERt, KENNETH CORNETTA*, RICK MORGANt, SHOSHANA MORECKI§, PAUL AEBERSOLD§, ATTAN KASID§, MICHAEL LOTZE§, STEVEN A. ROSENBERG§, W. FRENCH ANDERSONf, AND R. MICHAEL BLAESEt tMetabolism and §Surgery Branches, National Cancer Institute, and tMolecular Hematology Branch, National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, MD 20892 Communicated by Thomas A. Waldmann, December 31, 1990

ABSTRACT The application of bone marrow gene therapy has been stalled by the inability to achieve stable high-level gene transfer and expression in the totipotent stem cells. We show that retroviral vectors can stably introduce genes into antigenspecific murine and human T lymphocytes in culture. Murine helper T cells were transduced with the retroviral vector SAX to express both neomycin-resistance and human adenosine deaminase genes. These cells were expanded in culture and selected for expression of neomycin resistance with G418. The gene insertion, selection, and culture expansion did not alter antigen specificity or growth characteristics of the T cells in vitro. To determine if cultured T cells might be used for gene therapy, their persistence and continued expression of the introduced genes was evaluated in nude mice transplanted with the SAX-transduced T cells. G418-resistant cells could be readily recovered from the spleens of recipients of transduced T cells for several months. In addition, recovered cells continued to produce human adenosine deaminase. Based on these observations, we studied cultured human tumor-infiltrating lymphocytes as a candidate cell for a trial of gene transfer in man. Exponential cultures of interleukin-2-stimulated tumorinfiltrating lymphocytes were efficiently transduced with the neomycin-resistance gene using the retroviral vector N2. Gene insertion and subsequent G418 selection did not substantially alter the growth characteristics, interleukin 2 dependence, membrane phenotype, or cytotoxicity profile of the transduced T cells. These studies provided a portion of the experimental evidence supporting the feasibility of the presently ongoing clinical trials of lymphocyte gene therapy in cancer as well as in patients with adenosine deaminase deficiency.

Retroviral-mediated gene therapy is a promising potential approach for the treatment of human disease (1). Initial attention centered on candidate diseases affecting the bone marrow such as the hemoglobinopathies and severe combined immunodeficiency. Unfortunately, attempts at bone marrow gene transfer in large mammals and primates have not resulted in satisfactory levels of expression of the introduced genes (2). As an alternative approach, lymphocytes have several features that make them potentially attractive cellular vehicles for gene therapy (3). They are readily available from peripheral blood and are easily manipulated in tissue culture. This adaptability to tissue culture permits time for selection procedures and time to test for gene expression and other properties of the gene-transduced cells prior to their return to the patient. Peripheral lymphocytes are more fully differentiated than bone marrow precursor cells and thus inserted genes might not be as susceptible to inactivation

as they would be in cells going through many steps of differentiation (4). Memory lymphocytes will proliferate when exposed to their appropriate antigen and thus the population of gene-treated lymphocytes could be specifically expanded in vivo by reimmunization. Finally, some populations of antigen-specific lymphocytes will "target" to sites in the body containing deposits of antigen. Therefore, genetreated antigen-specific lymphocytes might be used to deliver specific gene products directly to the site of pathology (such as a tumor) in a treated patient. This report demonstrates that retroviral-mediated gene transfer can be used to stably introduce exogenous genes into both murine and human T lymphocytes in culture. Cultureexpanded antigen-specific gene-transduced T cells transfused into mice were shown to persist for prolonged periods in vivo, to retain their antigen-specific helper function, and to continue to express the introduced genes. Similarly, primary cultures of human T cells transduced with the neomycinresistance (NeoR) gene retain normal phenotypic and cytotoxic properties as well as growth characteristics and cytokine dependence.

MATERIALS AND METHODS Retroviral Vectors. The Moloney murine leukemia virusbased retroviral vector N2 (5), a generous gift from Eli Gilboa (Sloan-Kettering Cancer Center, New York), contains the NeoR gene promoted by the retroviral long terminal repeat. The SAX vector is a modification of N2 with the NeoR gene promoted by the long terminal repeat and the human adenosine deaminase (hADA) gene promoted by an additional internal promoter derived from simian virus 40 (6). The N2 virus was produced by the amphotropic packaging cell line PA317 (7) and the SAX virus was produced by PA12 cells. The average titer of each virus preparation was 1 x 106

G418-resistant colony-forming units/ml. Growth and Transduction of Murine and Human T Lymphocytes. Long-term cultures of the sperm whale myoglobin (SWM)-specific murine T helper cell clone 14.1 were maintained by repeated cycles of stimulation in vitro with 4.0 AuM SWM in the presence of irradiated (3300 rad; 1 rad = 0.01 Gy) fresh BlO.D2 spleen cells as antigen-presenting cells (8). Four days after antigen stimulation, the cultures were placed on irradiated syngeneic spleen cells for a 10-day rest phase. To introduce exogenous genes into these 14.1 T cells in the initial experiments, the T cells were cultured with the SAX retroviral vector at a virus to cell ratio of 5 in the presence of Polybrene (8 Ag/ml) for two sequential 4-hr incubations Abbreviations: hADA, human adenosine deaminase; NK, natural killer; NeoR, neomycin resistance; rIL-2, recombinant interleukin 2; SWM, sperm whale myoglobin; TIL, tumor-infiltrating lymphocyte; NPT, neomycin phosphotransferase; PHA, phytohemagglutinin.

The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.

3155

3156

Medical Sciences: Culver et al.

beginning on the third day after antigenic stimulation. In later experiments, stimulation with phytohemagglutinin (PHA) and recombinant interleukin 2 (rIL-2) was used since this combination resulted in more sustained proliferation and, therefore, permitted more complete G418 selection. Twentyfour hours after exposure to vector the transduced 14.1 cells were placed under selection with the neomycin analogue G418 [Geneticin; GIBCO]; 0.3 mg/ml (active drug) for 14 days. The SAX-infected G418-selected cells were then expanded by repeated 14-day cycles of antigen stimulation and rest on syngeneic feeder cells or by stimulation with PHA and rIL-2 (100 units/ml; a generous gift from Cetus). The use of IL-2 resulted in much more rapid expansion of the genetransduced cells in vitro, permitting their reinfusion into recipient animals closer to the time of transduction. Tumor-infiltrating lymphocytes (TILs) were grown from tumor biopsies obtained from patients with metastatic malignant melanoma (9). Exponentially growing cultures were exposed to two sequential 4-hr exposures to N2 supernatant (multiplicity of infection, 3) in the presence of protamine sulfate (5 jig/ml) (10). Twenty-four hours later a sample was removed and placed under selection in G418 (0.3 mg/ml) for 10 days. TIL cultures were fed and split as required (one or two times per week) until the total growth reached 2-3 x 1011 cells, the number ordinarily infused for therapy. Analysis of Vector Integration and Gene Expression. DNA samples from recipient mice and 14.1*SAX and control cells were PCR-amplified with primers specific for a segment of the NeoR gene and the amplified samples were analyzed as described (11, 12). The presence of the NeoR gene product neomycin phosphotransferase (NPT) was determined as described (13). Human and murine ADA were separated by electrophoresis on cellulose acetate (14). This method could detect approximately 1% hADA-producing cells mixed with 14.1 cells. To measure functional G418 resistance, lymphocytes were cultured at 2 x 105 cells per well in flat-bottom microculture plates alone or with G418. Proliferation in response to stimulation with PHA (5 lug/ml; Wellcome) and rIL-2 (100 units/ml) was measured as the incorporation of [3H]thymidine after 72 hr of culture. Results presented are the means of triplicate cultures, which varied by less than 10%. Cell-Surface Phenotyping and Cytotoxicity Analyses. From 3 to 8 weeks after gene insertion, the TIL populations were

analyzed for their surface membrane phenotype by standard fluorescence-activated cell sorting analysis of 1 x 106 TILs stained with monoclonal antibodies. Antibodies to CD4 and CD8 (Leu3 and Leu2, respectively; Becton Dickinson) were used. TIL cytotoxic function for autologous melanoma tumor targets and the natural killer (NK)-sensitive target K562 was measured by a standard 4-hr 51Cr release assay at effector/ target ratios of 40:1, 10:1, and 2.5:1.

Proc. Natl. Acad. Sci. USA 88 (1991)

the introduced genes and then expanded in culture by repeated cycles of antigen stimulation to obtain sufficient numbers of transduced cells for transplantation. These selected and expanded cells showed expression of high concentrations of NPIT as well as production of hADA (Fig. 1, lane 2). Although this cell growth protocol was successful, a high degree of selection for NeoR expression was difficult to achieve using a growth protocol consisting of only cycles of antigen stimulation and rest because the cell number only increased 4- to 8-fold every 2 weeks. When rIL-2 (100 units/ml) was added to the cultures stronger sustained proliferation was induced and resulted in consistently higher levels of hADA expression because of more complete G418 selection. The SAX-transduced 14.1 cells (14.1*SAX) retained their CD3/CD4' phenotype after gene insertion and G418 selection and continued to be specifically responsive to the antigen SWM in vitro (data not shown). In the initial experiment, athymic nude mice were then injected intraperitoneally with 20 x 106 14.1*SAX T cells and immunized with 100 gg of soluble SWM. Athymic recipients were chosen because they have little endogenous T-cell activity and, therefore, any T cells recovered from these mice later should represent the original injected 14.1*SAX cells or their progeny. In this initial experiment, spleens from recipient mice were removed 37 days after transplantation and tested for G418 resistance and hADA. Fig. 2 shows the proliferative response of recovered splenic lymphocytes from a group of two mice stimulated with PHA and IL-2 and the effect of G418 on this proliferation. Although control nude mouse splenocytes did proliferate modestly to PHA plus IL-2, this response was abolished by low concentrations of G418. By contrast, lymphocytes recovered from the spleens of nude mice injected with 14.1*SAX T cells proliferated much more vigorously and were resistant to G418 at 1.0 mg/ml. By using G418 to selectively regrow and recover the SAX-transduced T cells, we tested the recovered cells for continued expression of the hADA gene by electrophoretic separation of endogenous murine ADA from the hADA. Fig. 1, lane 3, shows that recovered G418-reselected T cells also continued to express the hADA gene at high levels. Thus both inserted genes were expressed in easily detectable amounts for 37 days after cell transfer in this experiment. Over 20 recipient mice were studied in various experiments designed to evaluate the duration of cell survival and the duration of expression of the introduced genes. Spleens from one experimental group of 11 mice were tested at intervals for vector DNA by PCR analysis. In these animals, detection of the inserted gene was variable with four of six recipients positive to a level of at least one NeoRcontaining cell among 10,000 spleen cells 37 days after transplantation. At 56 days, gene-modified cells were found in two of two recipient spleens and, by day 83, one of three recipient spleens was

RESULTS Studies with Murine T Cells. To initially test the feasibility of employing lymphocytes for gene transfer, we used the retroviral gene transfer vector SAX to attempt to insert the genes for hADA and NeoR into the nontransformed murine T-cell line 14.1. The T-cell line 14.1 was derived from the draining lymph nodes of a B1O.D2 mouse previously immunized with SWM by repeated cycles of antigen stimulation of these lymphocytes in tissue culture. 14.1 T cells proliferate in vitro when challenged with SWM in the presence of histocompatible antigen-presenting cells (without exogenous IL-2) or when stimulated with PHA and rIL-2. From 5 to 30% of the 14.1 T cells in various experiments achieved stable integration of the SAX vector after a supernatant infection protocol. The population of SAX-transduced 14.1 T cells was selected in G418 to enrich for cells expressing

FIG. 1. Lysates of the G418-resistant cells regrown from spleens removed at 37 days (lane 3) were separated by electrophoresis on cellulose acetate, which was then stained for the presence of the murine and human isozymes of ADA. Included also are a nontransduced 14.1 T cell population (lane 4) and 14.1 cells transduced with the SAX vector and partially selected in G418 (lane 2). The human T-cell line CEM is included as a control for the migration of hADA (lane 1).

Medical Sciences: Culver et al.

Proc. Natl. Acad. Sci. USA 88 (1991)

3157

Splenocytes from 14.1 Transplanted Nudes

EAL

0

0

a.

c 0 C.

0 la

E

1000*

I--

0.0

0.2

0.4 0.6 G418 Concentration (mg/ml)

0.8

1.0

FIG. 2. Effect of G418 on the proliferative response 72 hr after PHA and IL-2 stimulation of nontransduced 14.1 T cells (o, no vector), SAX-transduced G418-selected 14.1 T cells (A, 14.1*SAX), splenocytes from control nude mice (r), and splenocytes recovered from two nude mice [recipients A (x) and B (a)] that had been given an intraperitoneal injection of 20 x 106 14.1*SAX T cells 37 days earlier.

strongly positive. In a normal mouse prepared for cell transfer with 450 rads of y-irradiation (1 rad = 0.01 Gy), NPTexpressing NeoR-transduced T cells were recovered 145 days after cell transfer. These NPT-containing recovered cells also retained their original immune specificity throughout their period in vivo. This was illustrated by an experiment in which these recovered cells were challenged with their specific antigen (SWM) in vitro and responded with vigorous proliferation (control, 6402 cpm of [3H]thymidine incorporated; SWM, 70,287 cpm). As further evidence of the stability of the immune helper function of these gene-modified T cells, nude mouse recipients of 14.1*SAX T cells produced antibody after SWM immunization (7.8 ± 2.3 ,g/ml), whereas control nude animals did not (