CHIMERIC ANTIGEN RECEPTOR AND CAR-T CELLS THAT BIND A HERPES VIRUS ANTIGEN

Abstract

An isolated chimeric antigen receptor (CAR) polypeptide, wherein the CAR includes an extracellular antigen-binding domain, including an antibody or antibody fragment that binds to a protein encoded by a herpes virus, or to a protein complex including the protein (herpes virus antigen), wherein the herpes virus antigen is present on the surface of a human cell that is latently infected with said herpes virus and supports the lytic phase of viral replication. The invention further relates to a nucleic acid molecule encoding the CAR of the invention, a genetically modified immune cell, preferably a T cell, expressing the CAR of the invention and the use of the cell in the treatment of a medical disorder associated with human herpesvirus, such as herpes virus-associated cancers, chronic active herpes virus infections or primary herpes virus infections. In preferred embodiments the herpes virus is Epstein-Barr virus (EBV) and a preferred herpes virus antigen target of the CAR is the EBV glycoprotein 350/220 (gp350/gp220).

Claims

1. A chimeric antigen receptor polypeptide (CAR), comprising: i. an extracellular antigen-binding domain, comprising an antibody or antibody fragment that binds to a protein encoded by a herpes virus or to a protein complex comprising said protein (herpes virus antigen), wherein said herpes virus antigen is present on the surface of a human cell that is latently infected with said herpes virus and supports the lytic phase of viral replication, ii. a transmembrane domain, and iii. an intracellular signaling domain.

2. The CAR polypeptide according to claim 1, wherein the herpes virus antigen is involved in the virus binding to a receptor on a target human cell (herpes virus receptor binding protein).

3. The CAR polypeptide according to claim 1, wherein the herpes virus antigen is an Epstein-Barr virus antigen (EBV antigen).

4. The CAR polypeptide according to claim 3, wherein the EBV antigen is present on the surface of EBV-infected cells.

5. The CAR polypeptide according to claim 3, wherein the EBV antigen is an EBV virion envelope protein or a protein of the EBV envelope complex.

6. The CAR polypeptide according to claim 1, wherein the herpes virus antigen is the EBV glycoprotein 350/220 (gp350/gp220).

7. The CAR polypeptide according to claims 1: wherein the CAR comprises a leader polypeptide positioned N-terminally of the VH and VL domains, and/or wherein the extracellular antigen-binding domain comprises a linker polypeptide positioned between the VH and VL domains, and/or comprising additionally a spacer polypeptide positioned between the extracellular antigen-binding domain and the transmembrane domain, and/or wherein the transmembrane domain is a CD28 or a CD8 alpha transmembrane domain; and/or wherein the intracellular domain comprises a CD28 or a 4-1BB co-stimulatory domain; and/or wherein the intracellular domain comprises a CD3 zeta chain signaling domain; and/or wherein the CAR comprises one or more linker polypeptides positioned between the VH and VL domains and the spacer, and/or between the spacer and the transmembrane domain.

8. An isolated nucleic acid molecule, comprising a nucleotide sequence which encodes a CAR polypeptide according to claim 1.

9. A genetically modified immune cell comprising a nucleic acid molecule according to claim 8 and/or expressing a CAR according to claim 1.

10. A method for the treatment of a medical condition associated with herpes virus infection in a subject comprising administering to the subject a genetically modified immune cell according to claim 9.

11. The method according to claim 10, wherein the medical condition associated with herpes virus infection is a herpes virus-associated cancer in which herpes virus-antigens are present on the surface of cancer cells.

12. The method according to claim 10, wherein the herpes virus is EBV.

13. The method according to claim 12, wherein the medical condition is an EBV-associated cancer.

14. The method according to claim 12, wherein the medical condition associated with EBV infection is chronic active EBV infection (CAEBV) or primary EBV infection (e.g. mononucleosis).

15. The method according to claim 10 for use in the treatment of immune deficient or immune compromised patients after chemotherapy, radiation, immune suppression or transplantation.

16. The CAR polypeptide according to claim 4, wherein the EBV-infected cells are selected from the group consisting of EBV-infected cancer cells, EBV-infected B cells and EBV-infected epithelial cells.

17. The CAR polypeptide according to claim 7, wherein the leader polypeptide is an IgHL leader, wherein the linker is a G4S linker, and wherein the spacer is an IgG1 CH3 or a IgG1 CH2-CH3 spacer.

18. The genetically modified immune cell according to claim 9, wherein the immune cell is selected from the group consisting of a T lymphocyte, an NK cell, a macrophage, a dendritic cell, a cytotoxic T lymphocyte and a T helper cell.

19. The method according to claim 10, wherein the medical condition is selected from the group consisting of a herpes virus-associated cancer, a chronic active herpes virus infection and a primary herpes virus infection.

20. The method according to claim 13, wherein the medical condition is selected from the group consisting of a lymphoproliferative disorder (LPD), B-cell lymphoma, Burkitt lymphoma (BL), Hodgkin lymphoma (HL), a diffuse large B cell lymphoma (DLBCL), a post-transplant lymphoproliferative disorder (PTLD), an epithelial carcinoma, a lymphoepithelioma, a carcinoma with lymphoid stroma and a glioma.

Description

BRIEF DESCRIPTION OF THE FIGURES

[0221] FIG. 1: 7A1 and 6G4 (CD28z) CAR designs and generation of CAR-T Cells.

[0222] FIG. 2: Potency of 7A1 and 6G4 gp350-CAR-T cells (CD28z) to recognize and kill a 293T cell line engineered to express gp350.

[0223] FIG. 3: Potency of 7A1-gp350-CAR-T cells (CD28z) to kill the EBV-latently infected B95.8 cell line.

[0224] FIG. 4: Design of containing 7A1-gp350CAR-T cells with 4-1BB signaling, potency to rec-ognize and kill a 293T/gp350 cell line and potency to recognize the B95.8 cell line.

[0225] FIG. 5: Demonstration of the potency of 7A1-gp350CAR-T and 6G4-gp350CAR-T cells (CD28z) to recognize and be activated by autologous LCL cell lines immortalized with M81-EBV-GFP.

[0226] FIG. 6: Humanized mice infected with B95.8-EBV-GFP and treated with 7A1-gp350CAR-T (CD28z).

[0227] FIG. 7: Humanized mice pre-treated with 7A1-gp350CAR-T (CD28z) and infected with EBV-B95.8-fLuc.

[0228] FIG. 8: Overview of CAR constructs of the present invention created and tested.

[0229] FIG. 9: Humanized NRG mice pre-treated with sorted CAR+CD8+ or CAR+CD4+/CD8+7A1-gp350CAR-T (CD28z) and infected with the EBV-M81/fLuc2 strain.

[0230] FIG. 10: Humanized mice infected with EBV-M81/fLuc2 and treated with sorted CD8+7A1-gp350CAR-T (CD28z).

DETAILED DESCRIPTION OF THE FIGURES

FIG. 1: 7A1 and 6G4 (CD28z) CAR Designs and Generation of CAR-T Cells.

[0231] A 7A1 and 6G4 monoclonal antibodies were diluted to different doses (ng) and incubated with EBV-GFP virus prior to infection of Raji cells. Neutralization activity is shown as reduction of detection of GFP.sup.+ infected cells. B Schematic representation of CAR constructs containing the CD28z signaling domain and recognizing EBV-gp350 (7A1 or 6G4). As a cognate negative control, we used a CAR construct recognizing HCMV-gB (gB). Restriction digestion with Pml1, Cla1 and BamHI confirming structure of retroviral constructs containing CARs are shown on an additional sheet (FIG. 1 (cont.). C Overview of the commonly used methods here employed for retroviral gene transfer of CARs into adult peripheral blood or from cord blood T cells. D Left panel: representative example of flow cytometry analyses for detection of CARs on the cell surface. CARs were detected with AF647- or AF488-conjugated goat anti-human IgG-Fc Fab-fragments (Jackson ImmunoResearch, West Grove, Pa.) directed against the IgF4 hinge incorporated in the constructs. Graphs on the right: Detection of gp350CAR (7A1, n=7, and 6G4, n=6) relative to gBCAR (n=4) expression on CD4.sup.+- and CD8.sup.+-CAR-T cells produced with adult peripheral blood T cells. E Detection of gp350CAR (7A1) expression on CD4.sup.+- and CD8.sup.+-CAR-T cells produced with cord blood T cells from 5 different donors.

FIG. 2: Potency of 7A1 and 6G4 gp350-CAR-T Cells (CD28z) to Recognize and Kill a 293T Cell Line Engineered to Express Gp350.

[0232] A Flow cytometry analysis of gp350.sup.+ expression on 293T w.t. cells (grey) compared with analysis of 293T cells transduced with a lentiviral vector expressing gp350 (black). Surface gp350 was detected with the mouse antibody 72A1 and goat anti-mouse IgG AF647. B Experimental scheme: 293T or 293T/gp350 cells were seeded on plates and a day later CAR-T cells were added. After 24 and 48 h of co-culture, ELISA and FACS analyses were performed. C IFN-γ detection in cell supernatants showing high specificity of both 7A1 and 6G4 gp350CAR-T cells against 293T/gp350 at both lower (1:1) and higher (3:1) effector to target ratios, whereas the negative control gBCAR-T cells did not recognize the 293T/gp350 target. These assays were performed one time with triplicate cultures run in parallel. Statistical analyses were performed with two-way ANOVA and Bonferroni post-test. *** indicate p<0.001. D For detection of dead 293T/gp350 target cells, the fixable viability dye (FVD) e450 (eBioscience™) was added to the cells prior to flow cytometry analyses, allowing the distinction between viable and dead cells. The panels on the left show representative examples of 7A1 and 6G4 gp350CAR-T cells resulting into high frequencies of dead 293T-gp350 cells, whereas gBCAR-T cells did not promote cell death. The graphs on the right depict the quantified cell death after 24 h and 48 h of co-culture, showing higher killing effects at the higher effector to target ratios (3:1). E Frequencies of T cells relative to the targets was quantified by flow cytometry showing a preferential expansion of 7A1 and 6G4 gp350 CAR-T cells at E:T 3:1 in comparison with gB-CAR-T cells. F Analyses of gp350 expression on 293T/gp350 target cells showed that co-cultures with of 7A1 and 6G4 gp350 CAR-T cells at E:T 3:1 resulted into antigenic loss, i.e. negative selection of gp350.sup.+ targets.

FIG. 3: Potency of 7A1-Gp350-CAR-T Cells (CD28z) to Kill the EBV-Latently Infected B95.8 Cell Line.

[0233] A Flow cytometry analysis of gp350.sup.+ expression on a sub-population (around 6%) of latently infected B95.8 cotton top tamarin cell line. Surface gp350 was detected with the mouse antibody 72A1 and goat anti-mouse IgG AF647. B Experimental scheme: B95.8 cells were seeded on plates and shortly after CAR-T cells were added ad different E:T ratios. After 36-48 and 86-96 h of co-culture, ELISA and FACS analyses were performed. These analyses represent merged data from four independent experiments using one donor. C IFN-γ detection in cell supernatants showing high reactivity of 7A1-gp350CAR-T cells against B95.8 cells, whereas no to low reactivity of the gBCAR-T control cells at higher (1:1 and 10:1) effector to target ratios, whereas the negative control gBCAR-T cells did not recognize the B95.8 targets. These assays were performed 12 times for ratios 10:1 and 1:1 and 9 times for ratio 0.1:1. Statistical analyses were performed with two-way ANOVA and Bonferroni post-test. * indicates p<0.05 and *** indicate p<0.001. D CAR-T cells were marked with the proliferation dye CellTrace (Thermo Fisher) prior to co-culture. Analysis by flow cytometry showed that after 86 h of co-culture (E:T ratio of 10:1), 7A1-gp350CAR-T cells showed loss of the dye and three waves of proliferation whereas gB-CAR-T cells showed little loss of the marking dye. Representative results from four experiments. E For detection of dead B95.8 cells, the fixable viability dye (FVD) e450 (eBioscience™) was added to the cells prior to flow cytometry analyses, allowing the distinction between viable and dead cells. The panels on the left show representative examples of 7A1-gp350CAR-T cells resulting into high frequencies of dead B95.8 cells, but gBCAR-T cells did not promote cell death. The graphs on the right depict the quantified cell death after shorter and longer periods of co-culture, showing higher killing effects at the higher effector to target ratios (10:1). F Quantification of EBV DNA in cells obtained after co-culture was performed by an in house PCR method by detection of DNA sequences encoding for EBV BALF5 by qRT-PCR. Reduction of EBV copy numbers were observed after 86 h of co-culture with 7A1 CAR T cells. G Analyses of gp350 expression on B95.8 cells was measured by flow cytometry after 36-48 h and 86-96 h of co-culture showing loss of gp350 expression after co-culture with 7A1-gp350CAR-T compared to gBCAR-T cells. These assays were performed four times for ratios 10:1 and 1:1 and 3 times for ratio 0.1:1. Statistical analyses were performed with two-way ANOVA and Bonferroni post-test. * indicates p<0.05.

FIG. 4: Design of Containing 7A1-gp350CAR-T Cells with 4-1BB Signaling, Potency to Recognize and Kill a 293T/Gp350 Cell Line and Potency to Recognize the B95.8 Cell Line.

[0234] A Schematic representation of CAR constructs containing the 4-1BB signaling domain and recognizing EBV-gp350 (7A1). As a cognate negative control we used a CAR construct recognizing HCMV-gB (gB). B Flow cytometry analyses for detection of 4-1BB CARs on the cell surface of CD4.sup.+- and CD8.sup.+-CAR-T cells produced with adult peripheral blood T cells. CARs were detected with goat anti-human IgG-Fc Fab-fragments directed against the IgF4 hinge incorporated in the constructs. C Experimental scheme: 293T Vs. 293T/gp350 or B95.8 cells were seeded on plates and CAR-T cells were added. After 48 h of co-culture, ELISA and FACS analyses were performed. D IFN-γ detection in cell supernatants showing specificity of 7A1-gp350CAR-T-41BB cells against 293T/gp350 at higher (3:1, 10:1) effector to target ratios, whereas the negative control gBCAR-T cells was unspecific to 293T/gp350 target. D Quantified cell death after 48 h of co-culture, showing specific killing effects of 7A1-gp350CAR-T-41BB only at the intermediate effector to target ratios (3:1). F IFN-y detection in cell supernatants showing specificity of 7A1-gp350CAR-T-41BB cells against B95.8 cells at lower (1:1, 3:1) effector to target ratios.

FIG. 5: Demonstration of the Potency of 7A1-gp350CAR-T and 6G4-gp350CAR-T Cells (CD28z) to Recognize and be Activated by Autologous LCL Cell Lines Immortalized with M81-EBV-GFP.

[0235] A Flow cytometry analysis of gp350.sup.+ expression on sub-populations (around 26-30%) of latently infected LCLs obtained after immortalization of B cells with M81 EBV. Surface gp350 could be detected with the antibody 7A1 or 6G4 used as primary antibodies for staining. B Experimental scheme: LCLs cells were seeded on plates and shortly after autologous CAR-T cells were added ad different E:T ratios. After 86 h of co-culture, ELISA and FACS analyses were performed. These analyses represent merged data from two independent experiments. C IFN-γ release detected in supernatants after 86 h of co-culture showing T cell activation of 7A1 and 6G4-CAR-T in contrast to none to poor activation of gBCAR-T cells. These assays were performed six times for ratios 10:1 and 1:1 and three times for ratio 3:1 and 0.1:1. Statistical analyses were performed with two-way ANOVA and Bonferroni post-test. * indicates p<0.05 and *** indicate p<0.001. D CAR-T cells were marked with the proliferation dye CellTrace (Thermo Fisher) prior to co-culture. Analysis by flow cytometry showed that after 86 h of co-culture (E:T ratio of 1:1 or 10:1), CD4+ and CD8+7A1-gp350CAR-T cells showed the highest levels of loss of the dye compared with 6G4-gp350CAR-T and gB-CAR-T cells. E Analyses of gp350 expression on LCLs was measured by flow cytometry after 86 h of co-culture showing loss of gp350 expression after co-culture with 7A1-gp350CAR-T and 6G4-gp350CAR-T compared to gB-CAR-T cells.

FIG. 6: Humanized Mice Infected with B95.8-EBV-GFP and Treated with 7A1-gp350CAR-T (CD28z).

[0236] A Experimental scheme: Nod-Rag.sup.−/−Il2γcR.sup.−/− (NRG) mice were irradiated and transplanted with CD34.sup.+ cells isolated from cord blood. 16 weeks later, mice were infected i.v. with EBV-B95.8/GFP (10.sup.5 GRU i.v.). Five weeks after infection, 5×10.sup.6 7 A1-gp350CAR-T cells generated from the same cord blood donor were infused into 2 mice, and 3 mice were maintained as non-treatment controls. Four weeks after CAR-T cell administration, mice were sacrificed for analyses. Tumors could be detected in spleens of one CAR-T and one control animal. B Flow cytometry analyses for detection of CAR.sup.+ CD45.sup.+/CD3.sup.+ and CD45.sup.+/CD3.sup.+/CD8 cells in peripheral blood was analyzed weekly. Detection of CAR-T cells was highest 2 weeks after infusion, reducing until 4 weeks. C Flow cytometry analyses for detection of CAR.sup.+ CD45.sup.+/CD3.sup.+ and CD45.sup.+/CD3.sup.+/CD8 cells in spleen and tumors was analyzed after euthanasia. CAR-T cells (mostly CD8.sup.+ T cells) could be detected in the range of 1.5%-2.5% T cells in spleen and tumor tissues. D DNA isolated from spleen and tumor tissues was analyzed by RT-qPCR. High levels of EBV DNA could be only detected in the tumor obtained from the control mouse, but not from the tumor obtained from the CAR-T treated mouse. E Tumor sections were analyzed by EBER in situ hybridization, demonstrating sparse to rare EBER staining in the tumor obtained from the CAR-T treated mouse and frequent EBER staining in the tumor obtained from the control mouse.

FIG. 7: Humanized Mice Pre-Treated with 7A1-gp350CAR-T (CD28z) and Infected with EBV-B95.8-fLuc.

[0237] A Experimental scheme: Nod-Rag.sup.−/−Il2γcR.sup.−/− (NRG) mice were irradiated and transplanted with CD34.sup.+ cells isolated from cord blood. Twenty-five weeks later, 5×10.sup.6 7 A1-gp350CAR-T cells generated from the same cord blood donor were infused in one mouse. One day later, one CAR-T and one control mouse were infected i.v. with EBV-B95.8/fLUC (10.sup.5 GRU i.v.). Six weeks after infection, mice were sacrificed for analyses. B Sequential bioluminescence optical imaging analyses showing lower EBV-fLuc infection and spread in CAR-T pre-treated mouse relative to control mouse at 4 and 6 weeks after EBV challenge. C Longitudinal analyses of bioluminescence signal in spleen area from 2 to 6 weeks after EBV infection. D Bioluminescence signal detected in the regions of liver and salivary gland regions at endpoint analysis. E Longitudinal analyses of detection CAR-T cells in peripheral blood showing 3% CD45.sup.+/CD3.sup.+/CAR.sup.+ and 1.5% CD45.sup.+/CD3.sup.+/CD8.sup.+/CAR.sup.+ cells one week after administration, but reducing over time. F CD4.sup.+ CAR-T cells could be detected in bone marrow and CD8.sup.+ CAR-T cells could be detected in spleen at endpoint analysis.

FIG. 8: Overview of CAR Constructs of the Present Invention Created and Tested.

[0238] Provided is an overview of the various constructs generated by the inventors. The majority of all CAR constructs has been generated and expressed in both lentiviral and retroviral vectors. CAR-T production has been completed for constructs #1, 3, 9, 11. Experimentation in 293T and B95.8 co-culture settings (as described above) has been completed for constructs #1 and 3, and the above mentioned in vivo experiments conducted with construct #3. Additional experimentation is ongoing demonstrating the desired functional efficacy for the remaining constructs.

FIG. 9: Humanized NRG Mice Pre-Treated with Sorted CAR.sup.+CD8.sup.+ or CAR.sup.+CD4.sup.+/CD8.sup.+ 7A1-gp350CAR-T (CD28z) and Infected with the EBV-M81/fLuc2 Strain.

[0239] A Experimental scheme: Nod-Rag.sup.−/−IL2γcR.sup.−/− (NRG) mice were irradiated and transplanted with CD34.sup.+ cells isolated from cord blood. 17 weeks later, humanized mice were administered with PBS and served as controls (CTR, n=3) or infused i.v. with 2×10.sup.6 FACS-sorted CAR.sup.+/CD8.sup.+ (n=3) or with CAR.sup.+/CD4.sup.+/CD8.sup.+7A1-gp350CAR-T cells (n=4). The flow cytometry dot-plot graphs show exemplary results for CAR.sup.+ and CD8.sup.+ or CD4.sup.+ T cell populations before and after sorting. A day after T cell administrations, all mice were infected with the EBV-M81/fLuc2 strain (10.sup.6 GRU, i.v.). Mice were regularly bled to monitor the dynamics of reconstitution of human lymphocytes and monitored by optical imaging analyses for bio-distribution of EBV infection. Mice were euthanized 5 weeks after EBV infection. B, C Flow cytometry analyses of peripheral blood for detection of human CD45.sup.+ and human CD8.sup.+ cells, respectively. All mice maintained long-term reconstitution with human leukocytes. D Upper panels: Luciferase signal of EBV-M81/fLuc2 in the left side view of infected humanized mice is displayed in a 2D bioluminescence analysis using IVIS SpectralCT and LiveImage and performed in weeks 2, 3 4 and 5 after infection. Lower graphs: Quantification of the intensity of bioluminescence (Flux, represented as photons per second) in the full body of the mice over the course of the experiment. The graphs depict the quantified values for optical imaging analyses obtained for each mouse in the separate cohorts and the calculated merged values comparing the three cohorts (the standard deviation for each analysis is indicated). The data show higher levels of EBV infection in the CTR cohort compared with mice pre-treated with CD8.sup.+ or with CD4.sup.+/CD8.sup.+7A1-gp350CAR-T cells. E DNA isolated from spleen and bone marrow samples was analyzed by qRT-PCR. EBV DNA was detected more frequently in samples obtained from CTR (5/6) in comparison with CD8.sup.+ (2/6) and CD4+/CD8.sup.+ (4/6) CAR-T treated mice. For the remaining mice, the PCR showed non-detectable (n.d.) results. F Panels show the correlations for each individual mouse between values obtained by optical imaging analyses (full body) and values obtained for qRT-PCR for detection of EBV DNA in spleen and bone marrow. The squares drawn around the dots in the lower left corner of the graphs are to indicate that mice treated with CAR-T cells clustered together in the lower values, whereas values for control mice were more scattered.

FIG. 10: Humanized Mice Infected with EBV-M81/fLuc2 and Treated with Sorted CD8+7A1-gp350CAR-T (CD28z)

[0240] A Experimental scheme: Nod-Rag.sup.−/−IL2γcR.sup.−/− (NRG) mice were irradiated and transplanted with CD34.sup.+ cells isolated from cord blood. 17 weeks later, mice were infected i.v. with the EBV-M81/fLuc2 strain (10.sup.6 GRU, i.v.). 3 and 5 weeks after EBV infections, FACS-sorted CAR.sup.+ CD8.sup.+7A1-gp350CAR-T cells generated from the same cord blood donor were infused (CD8.sup.+ CAR, 2×10.sup.6 cells i.v., n=7). The flow cytometry dot-plot graphs show exemplary results for CAR.sup.+ and CD8.sup.+ populations before and after sorting. EBV-infected humanized mice administered i.v. with PBS served as controls (CTR, n=6). Mice were regularly bled to monitor the dynamics of reconstitution of human lymphocytes and monitored by optical imaging analyses for bio-distribution of EBV infection. The arrows in figures and graphs depict the time-points when CAR-T cells were administered. Mice were euthanized 8 weeks after EBV infection. B, C Flow cytometry analyses for detection of human CD45.sup.+ and human CD8.sup.+ cells in peripheral blood, respectively. All mice maintained long-term reconstitution with human leukocytes. D Upper panels: Luciferase signal of EBV-M81/fLuc2 infected humanized mice is displayed in the left side view or frontal view in a 2D bioluminescence analysis using IVIS SpectralCT and LiveImage and performed in weeks 2, 3 4 and 5 after infection. Lower graphs: Quantification of the intensity of bioluminescence (Flux, defined as photons per second) in the full body of the mice over the course of the experiment. The graphs depict the quantified values for optical imaging analyses obtained for each mouse in the separate cohorts and the calculated merged values comparing the three cohorts (the standard deviation for each analysis is indicated). The data obtained for weeks 6 and 8 show higher levels of EBV infection in the CTR cohort compared with mice treated with CD8.sup.+ CAR-T cells. E DNA isolated from spleen, liver and bone marrow of the mice was analyzed by qRT-PCR. A 50% reduction in the qRT-PCR signal can be observed for spleen and bone marrow of CD8.sup.+ CAR-treated mice compared with CTR. F Panels show the correlations for each individual mouse between values obtained by optical imaging analyses (region of spleen, liver or full body) and values obtained for qRT-PCR for detection of EBV DNA in spleen, liver and bone marrow. The squares drawn around the dots in the lower left corner of the graphs are to indicate that mice treated with CAR-T cells clustered together in the lower values, whereas values of control mice were more scattered. G Bioluminescence optical imaging analyses were performed with isolated organs. Mice in CTR cohort showed higher incidence of EBV infections (kidney (66%), brain (33%) and lungs (50%)) than mice treated with CD8.sup.+ CAR-T cells (kidney (29%), brain (14%) and lungs (29%)).

EXAMPLES

[0241] The invention is demonstrated by way of the examples disclosed below. The examples provide technical support for and a more detailed description of potentially preferred, non-limiting embodiments of the invention.

[0242] Preclinical in vitro proof-of-concept experimentation is presented below demonstrating efficacy in appropriate in vitro and in vivo models. FIG. 8 presents an overview of the various CAR constructs generated.

Example 1: Engineering of Functional CARs Based on the Amino-Acid Sequences of the Heavy and Light Chains of 7A1 and 6G4

[0243] The invention is based on the engineering of functional CARs based on the amino-acid sequences of the heavy and light chains of 7A1 and 6G4 both antibodies known to have a high neutralization capacity against EBV. The corresponding DNA was synthesized commercially using a codon-optimization approach. A retroviral vector expressing CARs and signaling through the CD28zeta chain was employed. The DNA fragments encoding for VH and VL were subcloned into the retroviral vector backbone and clones were selected by restriction digestion and confirmed by DNA sequencing analysis. FIG. 1A shows the neutralization capacity of both antibodies. FIG. 1B provides an overview of the retroviral vector map of both 6G4 and 7A1 gp350-CAR constructs and corresponding restriction digests. The plasmid constructs were used for transfection of 293T cells using the Calcium Phosphate DNA precipitation technique to generate retroviral vectors. Subsequently, T cells obtained from peripheral blood mononuclear cells (PBMCs) or the CD34.sup.− fraction of cord blood mononuclear cells (CBMCs) were stimulated in vitro with cytokines and activating immune beads and transduced with retroviral vectors, as shown in FIG. 1C. FIG. 1D shows CAR expression on T cells from PBMCs analyzed by flow cytometry using an antibody recognizing an epitope in the IgG4 hinge. Expression of the 7A1-gp350-CAR was higher than the 6G4-gp350-CAR. FIG. 1E presents data on expression levels reached on cord blood derived T cells.

Example 2: Demonstration of the Potency of Gp350-CAR-T Cells to Kill a 293T Cell Line Engineered to Express Gp350

[0244] FIG. 2 presents data supporting the potency of 7A1 and 6G4-gp350-CAR-T cells to kill 293T cells engineered to express gp350. To test cytotoxic effects of gp350-CAR-T, CAR-T cells were co-cultured with 293T lentivirally transduced to stably express gp350 (FIG. 2A). Analysis of co-cultures was performed 24 and 48 h after co-culture (FIG. 2B). 6G4 and even more pronounced 7A1-CAR-T secreted high amounts (>15 ng/ml) of IFN-γ in the presence of 293T/gp350 compared to control CAR-T. An accumulative effect over time could be observed with increasing levels after 48 h compared to 24 h (FIG. 2C). Cytotoxic effects were detectable as early as 24 h after co-culture especially in the 3:1 effector to target ratio and with 7A1-CAR-T cells. After 48 h, also lower ratios showed increased amounts of dead targets compared to co-cultures without gp350 (FIG. 2D). Relative amounts of T cells were increased in co-cultures with gp350-CAR-T in 3:1 effector to target ratio assumingly caused by a synergistic effect of target killing and proliferation of T cells (FIG. 2E). Further, another phenomenon could be observed. 24 h after co-culture we saw an out-proliferation of 293T negative for gp350 in cultures with 7A1 and 6G4-CAR-T cells reflecting the selective pressure forced upon the target cells. This effect was even more pronounced after 48 h of co-culture with 7A1-CAR-T cells (FIG. 2F).

Example 3: Demonstration of the Potency of 7A1-Gp350-CAR-T Cells to Kill the EBV-Latently Infected B95.8 Cotton Top Tamarin Cell Line

[0245] FIG. 3 presents data supporting the potency of 7A1-gp350-CAR-T cells to kill EBV-latently infected cell lines. B95.8 is a lymphoblastoid cell line derived from the cotton-top tamarin infected with EBV and also producing EBV viral particles. Surface staining for gp350 revealed expression of gp350 (FIG. 3A). After co-culture of B95.8 cells with CAR-T cells, analyses were performed 36 to 48 and 86 to 96 hours after (FIG. 3B). Analysis of the supernatant of the co-cultures revealed a dose-dependent IFN-γ release as well as an increasing accumulation of such to later time points (FIG. 3C). Labeling of the CAR-T cells prior to co-culture with a proliferation dye showed increased proliferation of gp350-CAR-T upon co-culture with target cells whereas no proliferation could be tracked in the control CAR-T group (FIG. 3D). Cytotoxic effects could be detected in high effector to target ratios already after 36-48 hours. 7A1-CAR-T cells showed increased cytotoxic effects than gB-CAR-T only in 10:1 effector to target ratios (FIG. 3E). Isolation of DNA followed by a BALF4 PCR to detect EBV copy numbers revealed decreased levels of EBV copies in gp350-CAR-T co-cultures compared to control CAR-T groups (FIG. 3F). However, whether this effect is caused by out-proliferation of T cells or killing of B95.8 cells remains unclear. Surface staining of gp350 on target cells after co-culture showed a dose-dependent loss of gp350 on target cells suggesting killing of gp350.sup.+ targets (FIG. 3G).

Example 4: Demonstration that 7A1-Gp350-CAR-T Containing a 4-1BB Signaling Recognize 293T/Gp350 and B95.8 Targets

[0246] FIG. 4 presents data supporting the potency of 7A1-gp350-CAR-T with a 4-1BB signaling domain (FIG. 4A) that can be expressed on CD4.sup.+ and CD8.sup.+ T cells (FIG. 4B) in in vitro experiments using 293T/gp350 and B95.8 targets (FIG. 4C). At 3:1 and 10:1 E:T ratios higher levels of IFN-γ were evident in the gp350CAR-T cell cultures with the 293T/gp350 target (100-400 pg/ml) than in the gBCAR-T co-cultures (10-50 pg/ml) (FIG. 4D). Preferential killing of 293T/gp350 by gp350CAR-T was seen at the 3:1 E:T ratio (FIG. 4E). Co-culture of B95.8 cells with gp350CAR-T cells resulted in higher IFN-γ detection at all E:T ratios tested (FIG. 4F). Overall, these results showed that gp350CAR-T with the 4-1BB domain were functional, but produced weaker effects than gp350CAR-T with the CD28z domain.

Example 5: Demonstration of the Potency of 7A1-Gp350-CAR-T and 6G4-Gp350-CAR-T Cells to Recognize and be Activated by a Same Donor-Derived LCL Cell Line Immortalized with M81 EBV

[0247] FIG. 5 presents data supporting the potency of 7A1-gp350-CAR-T cells and 6G4-gp350-CAR-T cells with the CD28z domain to recognize a donor-derived lymphoblastoid cell line (LCL) immortalized with the EBV viral strain M81. CD19.sup.+ cells were infected with EBV-M81 virus and after complete transformation checked for surface gp350 expression. Both antibody epitopes, 7A1 and 6G4, were detectable at high levels on the surface (FIG. 5A). CAR-T cells were co-cultured with LCLs and analysis was performed after 86 hours of co-culture (FIG. 5B). When co-cultured with gp350-CAR-T cells high levels of IFN-γ could be detected after 86 h. Highest levels of IFN-γ were observed in 3:1 and 1:1 ratios not in the highest ratio of 10:1 suggesting either a use of IFN-γ or an inhibiting effect of such high amounts of CAR-T cells (FIG. 5C). T cell proliferation of 7A1-CAR-T cells could be detected in comparison to control CAR-T (FIG. 5D). Loss of gp350 on target cells could also be detected especially in higher ratios of 7A1-CAR-T after 86 h indicating that gp350.sup.+ targets were killed however we could not detect increased amounts of dead targets with our FACS-based viability assay (FIG. 5E).

Example 6: 7A1-Gp350-CAR-T Cells were Tested In Vivo in an EBV Infection Model Using Humanized Mice Reconstituted with the Human Immune System

[0248] FIG. 6 shows 7A1-gp350-CAR-T cell efficacy in vivo in an EBV infection model using humanized mice. NRG mice were irradiated and hCD34.sup.+ cells were transplanted i.v. Sixteen weeks later, mice were infected with EBV-B95.8/GFP. Five weeks after EBV infection, mice were infused with 5×10.sup.6 7 A1-gp350-CAR-T cells (same donor matched to the hCD34 cells). Peripheral blood was taken weekly and mice were sacrificed 4 weeks following CAR-T cell injection (FIG. 6A). CAR-T cells were detected at high levels in peripheral blood and reached peak values two weeks post injection (FIG. 6B). At endpoint analysis, autopsy revealed tumor formation in spleen in one mouse with and one without CAR-T cells. Other mice did not show macroscopic tumor formation. Splenic and tumor tissues were then analyzed for presence of CAR-T cells. In spleen CAR-T cells were detected in both mice that received CAR-T cells in the beginning. In tumor tissue of the mouse that initially received CAR-T cells high amounts of CAR-T cells were detected. Further, levels were exceeding the ones measured in healthy splenic tissue suggesting detection of the tumor by CAR-T cells (FIG. 6C). Further, DNA was isolated from spleen and tumor tissues and analyzed by RT-qPCR. Viral DNA copies were found in splenocytes of both mice developing a tumor. While high amounts of viral DNA copies were found in the tumor of the mouse without CAR-T cells, only baseline PCR signal was detectable in the tumor of the mouse with CAR-T cells indicating killing of EBV.sup.+ tumor cells by CAR-T cells (FIG. 6D). Histopathological analysis of tumor sections also revealed lower levels of EBER staining in the tumor of the mice treated with CAR-T cells (FIG. 6E).

Example 7: In Vivo Testing Gp350-CAR-T in EBV Humanized Mouse Model EBV/B95.8fLuc

[0249] More extensive testing of 7A1-CAR-T cells in vivo to control and eradicate local and systemic EBV lymphoma and malignancies has been carried out, making use of a system in which EBV-B95.8/fLUC virus spread and tumor formation can be followed dynamically by non-invasive optical imaging. Results of a pilot experiment are shown in FIG. 6. NRG mice were irradiated and hCD34+ cells were transplanted i.v. 26 weeks later, one mouse was infused with 5×10.sup.6 7 A1-CAR-T cells (same donor matched to the hCD34 cells). One day later, both mice were infected with EBV/B95.8-fLuc. Peripheral blood was taken every second week and spread of infection was monitored every second week by IVIS Imaging. Mice were sacrificed 6 weeks following CAR-T cell injection (FIG. 7A). 4 weeks after infection, bioluminescence detection by optical imaging analyses revealed drastic differences in spread of infection. Whereas in the mice that received CAR-T cells infection was limited to the spleen, spread of infection to the liver was observable in the one without CAR-T cells. However, two weeks later although signal in the CAR-T cells treated mouse was still lower than in the control mouse also in the CAR-T cells treated mouse spread of infection had occurred (FIG. 7B). This suggests control of infection in the mouse with CAR-T cells until week 4 after infection. This is also underlined by quantification of spleen signal over the course of the experiment. Until week 4 signal in spleen in CAR-T treated mouse stayed stable whereas control mice showed rapid increase in signal from early on (FIG. 7C). At endpoint analysis, bioluminescence signal in liver and salivary glands regions were also measured and revealed lower signal in the mouse with CAR-T cells compared to control mouse (FIG. 7D). CAR-T cells were initially detected in blood but then levels decreased indicating a migration out of the blood stream into the tissues (FIG. 7E). This hypothesis is supported by the detection of CD4.sup.+ and CD8.sup.+ CAR-T cells in spleen and bone marrow at endpoint analysis (FIG. 7F).

Example 8: Humanized NRG Mice Pre-Treated with Sorted CAR.SUP.+ CD8.SUP.+ or CAR.SUP.+ CD4.SUP.+./CD8^{+ Derived from 7A1-gp350CAR-T (CD28z) Cells and Infected with the EBV-M81/fLuc2 Strain}

[0250] FIG. 9 shows the effects of 7A1-gp350-CAR-T cells that were sorted as CAR+CD8+ cells or sorted and then recombined 1:1 as CAR.sup.+ CD4.sup.+ plus CD8.sup.+ T cells and administered into humanized mice. The CAR construct employed was #3 according to FIG. 8, namely “gp350CAR7B(S.28.z)” derived from the 7A1 antibody, comprising the an IgHL-VH-G4S-VL scFV, an IgG Fc CH3 spacer and CD28-CD3ζ signaling domain. The EBV-M81fLuc2 infection in humanized mice was followed non-invasively and dynamically by optical imaging analyses. Seventeen weeks after stem cell transplantation, mice were administered with the CAR-T cells (CAR.sup.+ CD8.sup.+ or CAR.sup.+ CD4.sup.+ plus CD8.sup.+) or with PBS as control group and a day later they were infected with EBV. We observed a very consistent pattern of viral distribution, i.e., initially in spleen and then spreading systemically. Although immune reconstitution with endogenous leukocytes progressed normally for the mice in the different cohorts (FIG. 9B, C), we were unsuccessful in detecting CAR-T cells in peripheral blood or in lymphatic tissues of the mice. Nevertheless, a single administration with both CAR.sup.+ CD8.sup.+ or CAR.sup.+ CD4.sup.+ plus CD8.sup.+ cells prior to infection resulted in a diminution of EBV infection and spread monitored by optical imaging analyses (FIG. 9D). Further, DNA was isolated from spleen and bone marrow and analyzed by qRT-PCR. Viral DNA copies were found in splenocytes of all control mice, but were not detectable in spleen of several of the mice pre-treated with CAR-T cells (FIG. 9E). A similar trend was observed for bone marrow analyses (FIG. 9E). We observed a correlation between the data obtained by optical imaging and PCR, indicating that data of mice treated with CAR-T cells clustered apart from data of control mice (FIG. 9E). For these proof-of-concept experiments, data acquisition was not blinded and sample sizes were not statistically determined prior to experiments. Statistical analysis was performed using the GraphPad Prism software (Graphpad Software Inc., La Jolle, Calif., USA, Version: 6 and 7). T-test was used to calculate statistical significance.

Example 9: Humanized NRG Mice Infected with the EBV-M81/fLuc2 Strain and Treated with Sorted CAR.SUP.+ CD8.SUP.+ Derived from 7A1-gp350CAR-T (CD28z) Cells

[0251] In this experiment, CAR.sup.+ CD8.sup.+ cells were selected as a single population by sorting and used therapeutically after pre-established EBV-M81/fLuc2 infection. The CAR construct employed was #3 according to FIG. 8, namely “gp350CAR7B(S.28.z)” derived from the 7A1 antibody, comprising the an IgHL-VH-G4S-VL scFV, an IgG Fc CH3 spacer and CD28-CD3ζ signaling domain. All mice used in the study showed long-term reconstitution with human CD45.sup.+ or CD8.sup.+ T cells (FIG. 10B, C). EBV infection and viral distribution was detectable in all mice initially in spleen and then spreading systemically (particularly towards anatomical regions corresponding to liver and salivary glands) (FIG. 10D). Compared with control mice, treated mice showed lower EBV infection and spread at week 6 and 8 of the experiment, i.e., after the second administration with CAR.sup.+ CD8.sup.+ cells. Analyses of EBV genomic copies in spleen and bone marrow by PCR showed a reduction in the average EBV infection load for CAR.sup.+ CD8.sup.+ treated mice compared with control mice (approximately 50%), but the differences for liver were marginal (FIG. 10E). A correlation between the data obtained by optical imaging and PCR was observed, showing that the values of mice treated with CAR-T cells clustered apart from the values of control mice (FIG. 10F). Statistical analysis was performed using the GraphPad Prism software (Graphpad Software Inc., La Jolle, Calif., USA, Version: 6 and 7). T-test was used to calculate statistical significance. Corroborating the data obtained by non-invasive full-body optical imaging, explanted tissues analyzed by optical imaging showed an incidence EBV infection in control mice compared with CAR-T-treated mice in kidneys, brains and lungs (approximately doubled, FIG. 10G).

REFERENCES

[0252] Adhikary, D., U. Behrends, A. Moosmann, K. Witter, G. W. Bornkamm and J. Mautner (2006). “Control of Epstein-Barr virus infection in vitro by T helper cells specific for virion glycoproteins.” J Exp Med 203(4): 995-1006. [0253] Antsiferova, O., A. Muller, P. C. Ramer, O. Chijioke, B. Chatterjee, A. Raykova, R. Planas, M. Sospedra, A. Shumilov, M. H. Tsai, H. J. Delecluse and C. Munz (2014). “Adoptive transfer of EBV specific CD8+ T cell clones can transiently control EBV infection in humanized mice.” PLoS Pathog 10(8): e1004333. [0254] Daenthanasanmak, A., G. Salguero, B. S. Sundarasetty, C. Waskow, K. N. Cosgun, C. A. Guzman, P. Riese, L. Gerasch, A. Schneider, A. Ingendoh, M. Messerle, I. Gabaev, B. Woelk, E. Ruggiero, M. Schmidt, C. von Kalle, C. Figueiredo, B. Eiz-Vesper, C. von Kaisenberg, A. Ganser and R. Stripecke (2015). “Engineered dendritic cells from cord blood and adult blood accelerate effector T cell immune reconstitution against HCMV.” Mol Ther Methods Clin Dev 1: 14060. [0255] Dotti, G., S. Gottschalk, B. Savoldo and M. K. Brenner (2014). “Design and development of therapies using chimeric antigen receptor-expressing T cells.” Immunol Rev 257(1): 107-126. [0256] Eyquem, J., J. Mansilla-Soto, T. Giavridis, S. J. van der Stegen, M. Hamieh, K. M. Cunanan, A. Odak, M. Gonen and M. Sadelain (2017). “Targeting a CAR to the TRAC locus with CRISPR/Cas9 enhances tumour rejection.” Nature 543(7643): 113-117. [0257] Mamonkin, M., R. H. Rouce, H. Tashiro and M. K. Brenner (2015). “A T-cell-directed chimeric antigen receptor for the selective treatment of T-cell malignancies.” Blood 126(8): 983-992. [0258] Papadopoulou, A., U. Gerdemann, U. L. Katari, I. Tzannou, H. Liu, C. Martinez, K. Leung, G. Carrum, A. P. Gee, J. F. Vera, R. A. Krance, M. K. Brenner, C. M. Rooney, H. E. Heslop and A. M. Leen (2014). “Activity of broad-spectrum T cells as treatment for AdV, EBV, CMV, BKV, and HHV6 infections after HSCT.” Sci Transl Med 6(242): 242ra283. [0259] Salguero, G., A. Daenthanasanmak, C. Munz, A. Raykova, C. A. Guzman, P. Riese, C. Figueiredo, F. Langer, A. Schneider, L. Macke, B. S. Sundarasetty, T. Witte, A. Ganser and R. Stripecke (2014). “Dendritic cell-mediated immune humanization of mice: implications for allogeneic and xenogeneic stem cell transplantation.” J Immunol 192(10): 4636-4647. [0260] Xiaojun Tang, Yan Zhou, Wenjie Li, Qi Tang, Renjie Chen, Jin Zhu, Zhenqing Feng, (2014), “T cells expressing a LMP1-specific chimeric antigen receptor mediate antitumor effects against LMP1-positive nasopharyngeal carcinoma cells in vitro and in vivo”, The Journal of Biomedical Research, 2014, 28(6):468-475. [0261] Michael P. Weekes, Peter Tomasec, Edward L. Huttlin, Ceri A. Fielding, David Nusinow, Richard J. Stanton, Eddie C. Y. Wang, Rebecca Aicheler, Isa Murrell, Gavin W. G. Wilkinson, Paul J. Lehner, and Steven P. Gygi, “Quantitative Temporal Viromics: An Approach to Investigate Host-Pathogen Interaction” Cell 157, 1460-1472, Jun. 5, 2014. [0262] Lindsey M Hutt-Fletcher (2015), “EBV glycoproteins: where are we now?” Future Virol. 2015; 10(10): 1155-1162. [0263] Arvin A, Campadelli-Fiume G, Mocarski E, et al., editors. “Human Herpesviruses: Biology, Therapy, and Immunoprophylaxis”. Cambridge: Cambridge University Press; 2007