Genetic Manipulation of Tumor-specific Cytotoxic T Lymphocytes to Restore Responsiveness to IL-7

Introduction

Adoptive transfer of antigen-specific cytotoxic T lymphocytes (CTLs) and tumor infiltrating T lymphocytes can provide objective clinical responses in patients with malignant diseases.^12,3,4,5,6,7 These clinical responses appear to require the in vivo expansion and persistence of the infused CTLs^2,6,8,9 so that significant efforts have been made to maintain T cell survival in vivo.

Cytokines such as interleukin (IL)-2 promote both T cell expansion and survival,¹⁰ and IL-2 has frequently been used to support adoptive T cell therapies.^1,6,11 Although often effective, the prolonged administration of IL-2 is associated with significant toxicity that includes severe mucositis, nausea, diarrhea, edema, respiratory distress, liver and renal dysfunctions,^11,12,13 and expansion of regulatory T cells that impair the function of CTLs.¹⁴ Infusion of IL-7 may offer an alternative approach.

IL-7 is a common γ (γc) chain cytokine that is essential for homeostatic expansion of naive T cells^10,15 and for maintaining memory T cells.^16,17 Administration of IL-7 accelerates immunoreconstitution in murine models,¹⁸ and the cytokine has been well tolerated in early phase clinical trials.^19,20 In these human studies, IL-7 produced polyclonal expansion of circulating naive CD4⁺ and CD8⁺ T lymphocytes without measurable expansion of regulatory T cells.^19,20 Unfortunately, IL-7 administration can have minimal effects on adoptively transferred CTLs, as few of these cells express the IL-7Rα at the time of infusion. We have, therefore, determined whether forced expression of IL-7Rα in CTLs can restore their response to IL-7, or whether the downstream response to ligation of this receptor is lost at the same time as the receptor itself ceases to be expressed. Using our model of Epstein–Barr-Virus-specific CTLs (EBV-CTLs) that target post-transplant lymphomas, Hodgkin's lymphoma and nasopharyngeal carcinoma,^2,4,5,21 we found that transgenic expression of IL-7Rα consistently promotes CTL proliferation in response to IL-7 both in vitro and in vivo, and provides antitumor activity. These effects occur without modification of CTL antigen specificity or removal of their antigen dependency. Hence, IL-7 can induce equivalent antitumor activity to IL-2 when CTLs express a transgenic IL-7Rα.

Results

Discussion

Providing adoptively transferred antigen-specific CTLs with improved expansion and long-term survival in vivo should benefit their effectiveness as cancer immunotherapy.^11,13,32,33 At this point, we show that this effect can be achieved by engineering EBV-CTLs to efficiently respond to IL-7, a cytokine that physiologically maintains T lymphocyte homeostasis¹⁵ and can be safely administered as a recombinant protein.^19,20 Thus, the forced expression of the IL-7Rα subunit in EBV-CTLs restores their responsiveness to IL-7, and sustains their activity against a model human tumor in vitro and in vivo without affecting their antigen specificity or dependency.

T lymphocyte expansion and persistence requires, among other things, the cytokines IL-2, IL-7, and IL-15.¹⁰ Although the receptors for these cytokines share the γc chain component, each plays a distinct role at different phases of T cell development and each has a distinct signaling pathway.^10,16,29 IL-2Rα and IL-15Rα specific subunits and the shared IL-2/15Rβ subunit are highly expressed after T cell receptor engagement, so that signaling through both pathways supports activation induced expansion of T lymphocytes.¹⁰ Expression of IL-7Rα, by contrast, is limited to naive and memory T cells, and its dominant role is in sustaining or restoring lymphoid homeostasis.^15,18 Once T lymphocytes have encountered their cognate antigen and differentiated to effector T cells, IL-7Rα is downregulated in all but a small subset of effector CD8⁺ T lymphocytes that may form long-term memory T cells.¹⁶

These differentiation-associated changes in IL-7Rα expression have significant implications for the adoptive transfer of antigen-specific CTLs, as the in vitro stimulation received by peripheral blood T lymphocytes from antigens and cytokines can successfully expand the desired CTL populations, but would lead to a concomitant loss of IL-7Rα expression. It is presently evident that the in vivo expansion of CTLs after adoptive transfer is best accomplished in the lympho-depleted host, since in these individuals there is an increased production of IL-7 and IL-15 in an effort to restore the physiologic size of the lymphoid pool.^13,32,34,35 Although CTLs generated ex vivo respond to IL-15, their lack of IL-7Rα will render them unresponsive in vivo not only to the endogenous released IL-7 but also to the administered recombinant protein. Downregulation of IL-7Rα during effector T cell differentiation is a general phenomenon. Although it has been reported that tumor-infiltrating T lymphocytes isolated from melanoma re-express IL-7Rα after adoptive transfer in vivo and acquire a less differentiated phenotype,²⁶ these cells are more heterogeneous than highly selected antigen-specific CTLs, which do not physiologically re-express IL-7Rα in vivo. Our data showing that forced expression of IL-7Rα restores EBV-CTL response to IL-7 demonstrates that the signaling pathway downstream of the receptor remains functional in antigen-specific CTLs and that at this point the cells can expand and survive in response to IL-7 as well as to IL-2 and IL-15. We also found that IL-7Rα⁺ EBV-CTLs responding to IL-7 retain antitumor effects both in vitro and in vivo against EBV⁺ lymphomas and that the potency of the IL-7 induced effect was equivalent to IL-2, a cytokine that may induce significant toxicity and unwanted expansion of regulatory T cells. This restoration of IL-7 responsiveness is not exclusive to EBV-CTLs because we have found that re-expression of IL-7Rα by CTLs specific for other tumor-associated antigens has equivalent effects (Supplementary Figure S3).

In view of the fact that IL-7Rα is constitutively expressed predominantly by naive T cells, its forced re-expression in antigen-specific CTLs raises the concern that cytokine stimulation will not only promote cell survival and expansion, but will also produce unwanted alterations of function, such as diminished cytotoxic activity or shifts in the polarity of the cytokines released. Because CTLs are polyclonal, it is also possible that some subsets would respond with greater proliferation than others following cytokine stimulation, leading to selective outgrowth of CTLs with a more restricted range of specificities. IL-7 appears to have none of these undesired effects, unlike other γc chain cytokines such as IL-4, which drastically diminish cytolytic activity of mature CD8⁺ T lymphocytes.^36,37 Forced expression of IL-7Rα by EBV-CTLs did not alter their cytotoxicity, specificity, or capacity to respond to other γc cytokines that provide proliferative signals, such as IL-2 and IL-15. Forced expression of IL-7Rα and IL-7 exposure in vitro also had no effect on the phenotype of the EBV-CTLs, which retained the effector-memory profile that has been associated with long-term persistence in our previous studies.⁸

From a safety perspective, it is encouraging to know that the growth of IL-7Rα⁺EBV-CTLs remained strictly antigen dependent and that expansion ceased after antigen withdrawal, even in the presence of cytokines. Although γ-retroviral gene transfer of the γc chain to the hematopoietic stem cells of X-SCID patients has led to five cases of T cell leukemia,^38,39 the fact that we introduced the α-chain of one cytokine receptor rather than the γc chain shared by the receptors for multiple cytokines, and the fact that our target cells are fully differentiated T cells rather than hematopoietic stem cells reinforce the safety of this approach for a clinical application. However, whether the inclusion of a suicide gene is necessary, IL-7Rα could be coexpressed with one of them as we previously described for antigen-specific CTL genetically modified to express cytokines.³⁰

We conclude that forced expression of the IL-7Rα by CTLs can be used to recapitulate the response of these cells to this cytokine and thereby promote their in vivo antitumor activity after an adoptive transfer, avoiding the detrimental effects that can be encountered using the administration of recombinant IL-2 to sustain the expansion of adoptively transferred tumor-specific CTLs.

Materials and Methods

Plasmid construction and retrovirus production. Full-length human IL-7Rα (NCBI NM_002185) was cloned into the SFG γ-retroviral vector (SFG.IL-7Rα).³⁰ The γ-retroviral vectors encoding the fusion protein (eGFP-FFLuc) and FFLuc were previously described.³⁰ γ-Retroviral supernatants were prepared using transient transfection of 293T cells, as previously described.⁴⁰

Generation and transduction of antigen specific CTLs. EBV-CTLs were prepared using peripheral blood mononuclear cells, obtained from healthy donors (protocol approved by the Baylor College of Medicine's IRB), EBV-transformed lymphoblastoid cells (EBV-LCLs), and IL-2 (50 U/ml) (Proleukin; Chiron, Emeryville, CA) as previously described.² EBV-CTL lines obtained after three stimulations were transduced with γ-retroviral supernatants, as previously described.³⁰ In brief, 3 days after the stimulation with EBV-LCLs, CTLs were plated in 24-well plates precoated with a recombinant fibronectin fragment (FN CH-296; Retronectin; Takara Shuzo, Otsu, Japan). After the addition of γ-retroviral supernatant and IL-2 (100 U/ml), CTLs were spun and incubated at 37 °C in 5% CO₂. Three days later EBV-CTLs were collected and then maintained in culture by weekly stimulation with EBV-LCLs and IL-2 (50 U/ml) or IL-7 (5 ng/ml) (PeproTech; Rocky Hill, NJ). The procedures to generate CTLs specific for Cytomegalovirus, adenovirus, and cancer-testis antigens were previously described.^24,25

Immunophenotyping. Cells were stained with CD3, CD4, CD8, CD56, CD127, CD62L, CD28, CD27, CCR7, CD45RO, CD45RA monoclonal antibodies from Becton-Dickinson (Mountain View, CA) and with monoclonal antibodies specific for the T cell receptor -Vβ repertoire (IOTest βMark kit; Immunotech, Emeryville, CA). To detect the phosphorylation of Stat5, we used a monoclonal antibody binding to Tyr-694 and conjugated with Alexa Fluor 647 (from Becton-Dickinson). Human leukocyte antigen–tetramers and pentamers were provided by Baylor College of Medicine core facility and by Proimmune (Bradenton, FL), respectively. Cells were analyzed by a FACScan (Becton-Dickinson).

Chromium release assay. We evaluated the cytotoxic activity of EBV-CTLs using a standard 4-hour ⁵¹Cr release assay, as previously described.² As target cells we used autologous and mismatched EBV-LCLs and K562 cell line to, respectively, measure major histocompatibility complex restricted, major histocompatibility complex unrestricted, and natural killer activity.

Real-time Q-PCR. IL-7Rα transcripts were detected by Q-PCR. Total RNA extraction and amplification were performed as previously described.²⁵ Primer and probe sequences were provided by Applied Biosystems (Foster City, CA).

Enzyme-linked immunospot assay. IFN-γ enzyme-linked immunospot assay was performed, as previously described.²⁸ T cells were plated in triplicate and serially diluted from 1 × 10⁵ to 1 × 10⁴ cells/well and then peptide (5 µmol/l) was added to each. T cells incubated with an irrelevant peptide and T cells stimulated with 25 ng/ml phorbol myristate acetate and 1 µg/ml ionomycin (Sigma, St Louis, MO) were used as negative and positive control, respectively.

Expansion and antitumor activity in a xenogenic SCID mouse model. To assess the expansion, persistence, and antitumor effect of EBV-CTLs in vivo, we used a SCID mouse model and an in vivo imaging system as previously described.^30,31 CB17 SCID mice 8–10 week old were purchased from Harland Sprague Dawley (Indianapolis, IN). Mouse experiments were performed in accordance with Baylor College of Medicine's Animal Husbandry guidelines.

To evaluate the in vivo expansion, control and IL-7Rα⁺EBV-CTLs were transduced with a second γ-retroviral vector encoding for eGFP-FFLuc.³⁰ SCID mice were sublethally irradiated (250 rad) and injected subcutaneously with 10 × 10⁶ EBV-LCLs suspended in Matrigel (Becton-Dickinson). When the tumor was palpable (15–20 days after inoculation), one single dose of EBV-CTLs (10 × 10⁶) was injected intravenously. After CTL transfer, mice received IL-2 (1,000 U) or IL-7 (200 ng) intraperitoneally three times per week. Control groups received no cytokines after CTL transfer. To in vivo image EBV-CTLs expressing eGFP-FFLuc, mice were injected intraperitoneally with D-luciferin (150 mg/kg), and analyzed using the Xenogen-In Vivo Imaging System, as previously described.³⁰ The intensity of the signal was measured as total photon/s/cm²/sr (p/s/cm²/sr).⁴¹

To evaluate the antitumor effects, EBV-LCLs were transduced with the γ-retroviral vector encoding for FFluc and then selected in puromicin for two weeks. SCID mice were engrafted intraperitoneally with EBV-LCLs (3 × 10⁶ cells/mouse), and tumor growth was monitored using the bioluminescence system. When the signal was consistently increasing, usually by day 7–10, mice were treated intraperitoneally with three doses of control or IL-7Rα⁺EBV-CTLs (10 × 10⁶ CTLs/mouse). The animals were divided in three groups that received either IL-2 (1,000 U) or IL-7 (1,000 ng) intraperitoneally three times a week or no cytokines. One additional control group received only tumor cells.

Statistical analysis. All in vitro data are presented as mean ± 1 SD. After log transformation, ANOVA and Student's t test were used to compare the cell count data between treatment groups with multiple comparisons adjusted by Tukey method. When P-value < 0.05, a mean difference was accepted as statistically significant. To compare the growth and expansion trend over time, cell count data for every time points were analyzed by the generalized linear model. For the bioluminescence experiments, intensity signals were log-transformed and summarized using mean ± SD at baseline and multiple subsequent time points for each group of mice. Changes in intensity of signal from baseline at each time point were summarized. The response profiles over time were analyzed by the robust generalized estimating equations method for the repeated measurements.

Supplementary Material Figure S1. IL-7Rα is down regulated in CTLs specific for cytomegalovirus, adenovirus, and cancer testis antigens. Multivirus-specific CTLs were generated as previously described(25). Briefly PBMC obtained from healthy seropositive donors were stimulated with autologous monocytes and EBV-LCLs transduced with an adenovirus vectors encoding the CMV pp65 protein (AdF35pp65-I-GFP)(29) as a source of CMV and Adenovirus antigens. IL-2 (50 U/mL) was added to the culture at day 14. The expression of IL-7Rα was monitored by FACS analysis during culture. Panel a shows that IL-7Rα expression was down regulated in T lymphocytes during the generation of virus-specific CTLs. IL-7Rα was undetectable in CMV- and Ad-specific CTLs quantified using NLV- and TDL-specific tetramer staining, respectively (Panel b). Panels c-f illustrate the functional characterization of multivirus-specific CTL lines using IFNγ Elispot assay (panels c and e) and cytotoxicity assay (panels d and f). Figure S2. Growth in IL-7 does not impair the antigen specificity of IL-7Rα+EBV-CTLs. CTL clones specific for tumor-associated antigens such as PRAME were generated as previously described(24). Briefly, polyclonal T-cell lines obtained after 3 stimulations with artificial antigen presenting cells loaded with peptides derived from PRAME antigen were cloned by limiting dilution assay in 96 well plates, in the presence of IL-2 (100 U/mL), CD3 antibody (50 ng/mL OKT3; Ortho Biotech, Bridgewater, NJ) and irradiated (4,000 rad) allogeneic PBMC as feeder cells. Panel a show that these clones lack the expression of IL-7Rα by FACS analysis. Panels b and c illustrate that these clones are specific for the ALY-PRAME peptides using IFNγ Elispot assay (panel b) and cytotoxicity assay (panels c) against PHA-blasts loaded with ALY-peptide or an irrelevant peptide as target cells. Figure S3. Expression of IL-7Rα in MART-1 specific CTLs restores responsiveness to IL-7. Panel a. MART-1 specific CTLs were generated as previously described(24). Control MART-1 CTLs lacking the expression of IL-7Rα were transduced either with the SFG.IL-7Rα gamma retroviral vector or with the control vector ΔCD34. After transduction, IL-7Rα was re-expressed by CTLs transduced with SFG.IL7Rα but not by CTLs transduced with the control vector. Panel b illustrates that, after transduction, IL-7Rα was expressed by MART-1 tetramer+ CTLs. Panel c illustrates that transgenic IL-7Rα in MART-1 CTLs is functional as Stat5 is phosphorylated after incubation with IL-7. Panel d illustrates that after transduction the percentage of IL-7Rα+ MART-1 CTLs increased when the cells were stimulated once a week with antigen-presenting cells loaded with the MART-1 peptide and in presence of IL-7. Panel e shows that IL-7Rα+ MART-1 CTLs proliferate in response to IL-2 (50 U/ml) and IL-7 (5 ng/mL), while control MART-1 CTLs proliferate only in response to IL-2. Data represent two different MART-1 CTL lines. Cells were stimulated with antigen-presenting cells loaded with MART-1 peptide. CTL proliferation was quantified using the H³Thymidine incorporation assay. Table S1. Phenotypic analysis for memory markers of EBV-CTL lines. Table S2. Tetramer analysis of EBV-CTL lines. Table S3. Repertoire analysis of EBV-CTL lines generated from 4 representative donors.

Supplementary Material