Vaccination Targeting Native Receptors to Enhance the Function and Proliferation of Chimeric Antigen Receptor (CAR)-Modified T Cells

Abstract

Purpose: The multiple mechanisms used by solid tumors to suppress tumor-specific immune responses are a major barrier to the success of adoptively transferred tumor-specific T cells. As viruses induce potent innate and adaptive immune responses, we hypothesized that the immunogenicity of viruses could be harnessed for the treatment of solid tumors if virus-specific T cells (VST) were modified with tumor-specific chimeric antigen receptors (CAR). We tested this hypothesis using VZV-specific T cells (VZVST) expressing a CAR for GD2, a disialoganglioside expressed on neuroblastoma and certain other tumors, so that the live-attenuated VZV vaccine could be used for in vivo stimulation.Experimental Design: We generated GMP-compliant, GD2.CAR-modified VZVSTs from healthy donors and cancer patients by stimulation of peripheral blood mononuclear cells with overlapping peptide libraries spanning selected VZV antigens, then tested their ability to recognize and kill GD2- and VZV antigen-expressing target cells.Results: Our choice of VZV antigens was validated by the observation that T cells specific for these antigens expanded in vivo after VZV vaccination. VZVSTs secreted cytokines in response to VZV antigens, killed VZV-infected target cells and limited infectious virus spread in autologous fibroblasts. However, while GD2.CAR-modified VZVSTs killed neuroblastoma cell lines on their first encounter, they failed to control tumor cells in subsequent cocultures. Despite this CAR-specific dysfunction, CAR-VZVSTs retained functional specificity for VZV antigens via their TCRs and GD2.CAR function was partially rescued by stimulation through the TCR or exposure to dendritic cell supernatants.Conclusions: Vaccination via the TCR may provide a means to reactivate CAR-T cells rendered dysfunctional by the tumor microenvironment (NCT01953900). Clin Cancer Res; 23(14); 3499-509. ©2017 AACR.

Figure 1. VZVSTs can be reactivated and expanded from the peripheral blood of healthy donors and cancer patients

(A) VZVSTs were generated from donors immunized by natural infection (n=9) or by VZV vaccination without history of natural infection (n=5). The frequency of antigen-specific T-cells in peripheral blood was determined by IFN-γ release ELISPOT assay after overnight stimulation of PBMCs with VZV antigen-spanning pepmixes (Day 0) and again on days 9 and 16. Closed circles (●) represent naturally infected donors and open circles (○) are vaccinated donors. (B) T cell expansion was evaluated by viable cell counting using trypan blue exclusion. Solid line represents naturally infected donors (n=9) and dashed line, vaccinated donors (n=5). Results are shown as mean cell numbers ± SEM. Each arrow indicates each stimulation. (C and D) The expression of surface markers as a percentage of the live cells was assessed on day 16 (n=6). Results are shown as mean ± SD. (E and F) PBMCs from 6 cancer patients including three pediatric patients were stimulated with VZV pepmix-pulsed mature DCs. Results from the five responders are shown. (E) Specificity on day 16 was measured by IFN-γ release ELISPOT assay (mean ± SEM, n=5). (F) Cell expansion was evaluated by counting using trypan blue exclusion. Results are shown as mean cell numbers ± SEM (n=5). Each arrow indicates each stimulation.

Figure 2. Circulating VZVSTs are increased in response to VZV booster vaccination

PBMCs drawn before and at intervals after immunization of seropositive adults with the VZV booster vaccine, Zostavax, were stimulated with VZV antigens. The frequency of antigen-specific T-cells in peripheral blood was determined by IFN-γ release ELISPOT assay. (A) Frequency of VZVSTs shown after 24 hours of stimulation and (B) after restimulation with pepmixes 4 days later in one donor and 6 days later in 2 donors. The background frequency of IFN-γ-secreting T-cells was subtracted from the specific responses. Results are shown as mean ± SD (n=3). (C) The response to each antigen of one of three vaccinated donors.

Figure 3. VZVSTs prevent infectious virus spread

(A) GFP-tagged, VZV-infected dermal fibroblasts were used as targets. Infected fibroblasts were GFP positive and a fraction show cytopathic effect. (B) Frequency of infected fibroblasts as measured by GFP expression. (C) GFP-tagged VZV-infected fibroblasts, fresh uninfected fibroblasts and effector cells were co-cultured at the ratio of 1:7:20. Three days later, cultures were collected and stained with CD45 antibodies, and analyzed by FACS. Indicated are the absolute numbers of events for GFP positive VZV-infected cells. (D) Infectious center assay. % inhibition was calculated as (number of infectious center in each test condition / number of infectious center in controls without effectors) x 100. VZVSTs prevent infectious spread, while PBMCs and ATCs had little protective capacity.

Figure 4. Chimeric antigen receptors (CARs) can be expressed in VZVSTs, maintained over three stimulations and have functional specificity for both VZV and GD2

(A) Diagram of GD2.CAR-VZVST generation. (B) GD2.CAR expression on transduced VZVSTs was measured using the 14g2a-specific anti-idiotype antibody (1A7) after the 1^st, 2^nd and 3^rd stimulations. T cell lines from donors either naturally infected with VZV ■ (n=3) or immunized by vaccination (n=3) were assessed. Results are shown as mean % of expression ± SD. (C) One representative dot blot showing expression of the GD2.CAR. (D) VZV pepmix-pulsed autologous activated T-cells (ATCs) were labeled with ⁵¹Cr and cultured with the transduced VZVSTs for 6 hours at the effector target ratios shown (solid lines). Control targets were autologous ATCs (dotted lines). Results are shown as mean of % specific lysis ± SEM (n=4). (E) GD2-positive NB cells (LAN-1 and JF) were labeled with ⁵¹Cr and cultured with GD2.CAR-transduced VZVSTs (solid lines) or non-transduced VZV-specific T cells (dotted lines) for 6 hours at the effector target ratios shown. Results are shown as mean of % specific lysis ± SEM (n=3). (F) GD2.CAR-VZVSTs and NT-VZVSTs were co-cultured with LAN-1 cells at the ratio of 1:1. Five days later, cultures were stained with CD3 and GD2 antibodies, and analyzed by flow cytometry.

Figure 5. Transduced T-cells are bi-specific

GD2.CAR-VZVSTs were stained with the 14.g2a-specific idiotype antibody 1A7 (mouse IgG1) and sorted with Anti-Mouse IgG1 MicroBeads. Sorted cells were rested 48 hours in incubator. (A) Purity of sorted cells was tested after 48 hours incubation. (B) The VZV antigen-specificity of the sorted cells from one of the two donors tested is shown and was determined using IFN-γ release ELISPOT assays. (C) Sorted cells were stimulated with VZV antigen pepmixes and IFN-γ or TNF-α production was evaluated by intracellular cytokine staining. One representative of two donors is shown.

Figure 6. CAR-VZVSTs became exhausted by stimulation through the CAR but the TCRs were still functional

(A) GFP-ffluc-LAN1 (GFP+) and T cell (CD3+) were analyzed quantitatively after the 1^st and 2^nd co-cultures. The fold change of T cell or tumor cell numbers between the end of 1^st coculture and the end of 2^nd coculture are shown. (n=6, mean ± SD) (B) Cytokine secretion from GD2.CAR-VZVST after encountering LAN1 cells. T cells and LAN1 cells were plated at 1:1 ratio and after 24 hours supernatants were collected for ELISA. On day 7, T cells were harvested and replated with LAN1 cells at the same E:T ratio. Supernatants were collected on Day 8 (24 hours after 2^nd co-culture) and then IFN-γ and IL-2 concentrations were measured. (n=6) (C) PD-1 and Tim-3 expression on non-transduced VZVST, GD2.CAR-VZVST before or 7 days after coculture with LAN1 cells are shown. GD2.CAR VZVST were gated on CAR+ and CD4+ or CD8+. Dot plots from one representative donor of two are shown. (D) The frequency of antigen-specific T-cells before and after coculture with LAN1 was determined by IFN-γ release ELISPOT assay after overnight stimulation with VZV antigen-spanning pepmixes. Data denote mean ± SD (n=3). (E) After the 2^nd co-culture, T cells were harvested and cultured with VZV pepmix loaded DCs, irrelevant pepmix loaded DCs, DCs alone, or no stimulation for three days and then co-cultured with fresh GFP-ffluc-LAN1 cells. The fold tumor cell increase on day 7 to day 0 are shown. Data denote mean ± SD (n=4, no stimulation, DC alone; n=3, DC+irrelevant pepmix, n=5, DC+VZV pepmix).