AU2014262282B2

AU2014262282B2 - HLA-restricted, peptide-specific antigen binding proteins

Info

Publication number: AU2014262282B2
Application number: AU2014262282A
Authority: AU
Inventors: Nai-Kong Cheung; Jian Hu; Dimiter TASSEV
Original assignee: Memorial Sloan Kettering Cancer Center
Current assignee: Memorial Sloan Kettering Cancer Center
Priority date: 2011-02-11
Filing date: 2014-11-17
Publication date: 2016-10-27
Anticipated expiration: 2032-02-13
Also published as: AU2014262282A1

Abstract

Antigen binding proteins with TCR-like paratopes, that is, with an antigen binding region specific for an HLA-A2 restricted peptide are disclosed. The antigen binding proteins encompass antibodies in a variety of forms, including 5 full-length antibodies, substantially intact antibodies, Fab fragments, F(ab')2 fragments, and single chain Fv fragments. Fusion proteins, such as scFv fusions with immunoglobulin or T-cell receptor domains, incorporating the specificity of the antigen binding region for each peptide are also contemplated by the invention. Furthermore, immunoconjugates may include 10 antibodies to which is linked a radioisotope, fluorescent or other detectable marker, cytotoxin, or other molecule are also encompassed by the invention. Among other things, immunoconjugates can be used for delivery of an agent to elicit a therapeutic effect or to facilitate an immune effector function. 5956397_1 (GHMatters) P94551.AU.1 WENDYB 0.s 0J HlIt:A-A2-RFf'PNAPYL 0.7 -R sq.:~Ht.A-A2.-SL GEQQYSV S0.6 HL ~ KA-A2-CMTiWNQMNL SVH&AA2-VISNDVCAQV' ' 50- HL KA-A2-FLTPKKLQCV 0.4 2 HLA-A2-RLTRF SRV G HLA--A2-Et MLGEFLK.L a None Clone 42 Clone 43 Clone 45 I g GHLA-A2-RMFPNAPYL 1.!*.-u~~. m HLA-A2-CMT WNQMNL 1 HLA-A%2-TLPPAWQPFL 1. 4 HLA-A2-ALLMAGLAL ~ 1 2' ~- / ~ILA-A2-FLT.PKKQCV HLA-A2-V1SNDVCAQV H LA-A2-ALFDIESKV 6 HLA--A2--VILKKATEYV 0.6 N':HL A-A2-LMLGEFLKL R MIPNAPYL Peptide 0.2 f None WT1 Clone 45 Anti-HLA-A2 Figure 2

Description

-1- 2014262282 05 Oct 2016

HLA-RESTRICTED, PEPTIDE-SPECIFIC ANTIGEN BINDING

PROTEINS

Cross-Reference to Related Applications [0001] This application claims priority from U.S. Provisional Application No.

5 61/463,082 , filed February 11, 2011, entitled GENERATION AND USE OF HLA-A2 RESTRICTED, PEPTIDE-SPECIFIC MONOCLONAL ANTIBODIES AND CHIMERIC ANTIGEN RECEPTORS. This Provisional Application is hereby incorporated by reference in its entirety into the present disclosure.

SEQUENCE LISTING 10 [0002] The instant application contains a Sequence Listing, created on

February 2, 2012; the file, in ASCII format, is designated 3314019AWO_seqlisting_ST25.txt and is 47.8 kilobytes in size. The file is hereby incorporated by reference in its entirety into the instant application.

BACKGROUND OF THE INVENTION 15 Technical Field [0003] The present invention relates generally to antigen-binding protein molecules involved in immune function. More particularly, the present invention relates to recombinant antibodies, chimeric antigen receptors and fragments thereof with specificity for an HLA-restricted peptide, where the 20 peptide is derived from a cellular or viral protein of interest.

Background Information [0004] Advances in adoptive T cell immunotherapy have led to several promising options for cancer patients in the past decade. T-cell based immunotherapy for cancer stemmed from studies which showed a correlation 25 of increased numbers of tumor infiltrating lymphocytes (TILs) in surgical specimens and patient outcome. It is generally believed that this infiltration of TILs represents activation of an anti-tumor mechanism and that the infiltration 8270468J (GHMatters) P94551.AU.1 5Oclober2016 -2- 2014262282 05 Oct 2016 was mediated through the expression of tumor associated antigens in the context of MHC. These findings eventually led researchers to try and take advantage of antigen-specific T cells for the treatment of cancer.

[0005] For induction of cytotoxic T- cell (CTL) responses, intracellular 5 proteins are usually degraded by the proteasome or endo/lysosomes, and the resulting peptide fragments bind to MHC class I or II molecules. These peptide-MHC complexes are displayed at the cell surface where they provide targets for T cell recognition via a peptide-MHC (pMHC)-T cell receptor (TCR) interaction. Vaccinations with peptides derived from cellular and viral protein 10 can induce HLA-A0201-restricted cytotoxic CD8 T cells, which are capable of killing tumor cells or virally-infected cells.

[0006] Antibodies are increasingly being used as therapeutic agents to fight cancer, autoimmune disease and infection. Therapeutic antibodies have been exploited based on their multiple mechanisms of action, which include the 15 following: 1) naked antibodies killing tumor cells directly by ADCC or CDC (e.g. trastuzumab), 2) blocking or stimulating a cell membrane molecule to induce cell death (e.g. cetuximab), 3) neutralizing a secreted moiety (e.g. bevacizumab), 4) killing via an attached moiety such as a drug, toxin, radioisotope and 5) modulating the immune system via T cell effector 20 functions.

[0007] In almost all cases, to generate a therapeutic benefit, antibodies have to possess critical properties including high affinity for their targeted antigen, minimal acute and long-term side effects, and in specific applications, high affinity for human Fc receptors (4). In addition, the targeted antigen has to be 25 expressed at high levels on tumors but not on normal tissues (specificity or selectivity), consistently expressed in the specific tumor among patients and within patients (low heterogeneity), and should either be essential for the survival of the cancer cell or unlikely to be down regulated.

[0008] To achieve these attributes, researchers can now reengineer existing 30 antibodies to make them less immunogenic, modifying both protein and 8270468_1 (GHMatters) P94551 .AU.1 5Oclober2016 -3- 2014262282 05 Oct 2016 carbohydrate residues in the Fc regions to enhance ADCC and CDC, shrinking their sizes for potentially better tumor penetration, mutating the variable regions to improve affinity, increasing avidity by changing antibody valency, and constructing novel antibody-fusion proteins such as those for 5 multi-step targeting (5) and for redirecting immune cells by way of a chimeric antigen receptor (CAR). Furthermore, researchers continue to define the structural attributes and the host characteristics responsible for success among currently approved antibodies (6).

[0009] With the objective of eliminating or neutralizing the pathogenic agent 10 or disease target, including bacterial, viral or tumor targets, antigen-specific, antibody-based treatments are particularly attractive because of the antibody's exquisite specificity.

SUMMARY OF THE INVENTION

[0010] The present invention, therefore, is based on the identification of 15 antigen-specific binding sequences from which a variety of antigen-binding proteins can be produced, for example, an antibody specific for an antigen that represents a complex of a protein fragment (peptide) and an HLA molecule similar to that typically recognized by a T-cell receptor following antigen processing and presentation of the protein to the T-cell. Phage 20 display is used to select an initial antigen-binding molecule that can be used to engineer the antigen-binding proteins of the invention, which include antibodies and chimeric antigen receptors (CARs).

[0011] In one aspect, therefore, the invention relates to an isolated antigenbinding protein or antigen-binding fragment thereof comprising one of: 25 (A) an antigen binding region having the amino acid sequence of one of SEQ ID NOS: 5, 8, 10, 13, 14, 17, 20; (B) an antigen binding region comprising a VH and VL, respectively, with amino acid sequences selected from SEQ ID NOs: 24 and 25; 26 and 27; 28 and 29; 30 and 31; 32 and 33; 34 and 35; and 36 and 37; or 8270468_1 (GHMatters) P94551 .AU.1 5 October 201$ -4- 2014262282 05 Oct 2016 (C) (i) the following three light chain (LC) complementarity determining regions (CDRs): (a) a LC CDR1 comprising the amino acid sequence of SEQ ID NO: 56; and 5 (b) a LC CDR2 and CDR3 comprising respectively, the amino acid sequence of SEQ ID NOs: 58 and 65, 59 and 66, 60 and 67, 61 and 68, 61 and 69, 62 and 70 and 63 and 71; and (ii) the following three heavy chain (HC) CDRs:

(a) a HC CDR1 comprising the amino acid sequence of SEQ ID 10 NO: 38; and (b) a HC CDR2 and CDR3 comprising respectively the amino acid sequence of one of SEQ ID NOs: 41 and 49, 42 and 50, 43 and 51, 44 and 52, 45 and 53, 46 and 54 and 47 and 55.

[0012] In a related aspect, the invention relates to an isolated antigen-15 binding protein or antigen-binding fragment thereof, wherein the isolated antigen-binding protein is an antibody or a chimeric antigen receptor. The antibody is a full-length antibody, a substantially intact antibody, a Fab fragment, a F(ab')2 fragment or a single chain variable fragment (scFv).

[0013] In the isolated antigen-binding protein, whether an antibody or CAR, 20 the antigen-binding region specifically binds to an epitope of an HLA-peptide complex.

[0014] Peptides that are recognized by the antigen-binding proteins of the invention as part of an HLA-peptide complex include, but are not limited to, a peptide with the amino acid sequence LLDFVRFMGV (SEQ ID NO:4); a 25 peptide with the amino acid sequence RLTRFLSRV (SEQ ID NO: 7); a peptide with the amino acid sequence RIITSTILV (SEQ ID NO: 12); and a peptide with the amino acid sequence LLEEMFLTV (SEQ ID NO: 19). In some embodiments, the peptide is recognized in associate with an HLA-A2 antigen. 8270468_1 (GHMatters) P94551 .AU.1 50dOber2016 -5- 2014262282 05 Oct 2016 [0015] In yet another aspect, the isolated antigen-binding protein of the invention is a scFv comprising an amino acid sequence selected from the group consisting of SEQ ID NOS: 5,8,10,13,14,17 and 20.

[0016] In a related aspect, the isolated antigen-binding protein is a fusion 5 protein comprising an antigen-binding region as disclosed in any of Tables 1 -8.

[0017] In another aspect, the invention relates to an immunoconjugate comprising a first component which is an antigen-binding protein, or antigenbinding fragment thereof as disclosed herein. The immunoconjugate 10 comprises a second component that is a cytotoxin, a detectable label, a radioisotope, a therapeutic agent, a binding protein or a molecule having a second amino acid sequence. Where the second component is a binding protein or second antibody, the binding protein or second antibody has binding specificity for a target that is different from the HLA-peptide complex. 15 [0018] In a related aspect, therefore, the present invention relates to bispecific antibody comprising an antigen-binding protein or functional fragment thereof as described herein.

BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 shows the binding of bacterial supernatant from individual 20 EBNA3C scFv clones 315, 335, 327 and 345 (A) and purified EBNA clone 315 scFv (B) to various HLA-A2-peptide complexes demonstrating that clone 315 is highly specific for the HLA-A2-LLDFVRFMGV complex.

[0020] Figure 2 the binding of bacterial supernatant from individual WT-1 scFv clones 42, 43 and 45 (A) and purified WT-1 clone 45 scFv (B) to various 25 HLA-A2-peptide complexes demonstrating that WT-1 clones 42, 43 and 45 are highly specific for the recombinant HLA-A2-RMFPNAPYL complex.

[0021 ] Figure 3 shows that HLA-A2 can be detected on TAP-deficient (TAP-) T2 cells that were either pulsed or unpulsed with LLDFVRFMGV or another (irrelevant) peptide (A) but that EBNA clone 315 scFv recognizes T2 82704$8_1 (GHMatters) P94551 AU.1 5 October 201$ -6- 2014262282 05 Oct 2016 cells that have been pulsed with LLDFVRFMGV but not unpulsed cells or cells pulsed with irrelevant peptide (B) with a lower limit of detection at about 78nM (C).

[0022] Figure 4 shows that HLA-A2 can be detected on TAP-deficient 5 (TAP-) T2 cells that were either pulsed or unpulsed with RMFPNAPYL or LLDFVRFMGV (A) but that WT-1 clone 45 scFv recognizes T2 cells that have been pulsed with RMFPNAPYL but not unpulsed cells or cells pulsed with LLDFVRFMGV (B) [0023] Figure 5 shows that when DIMT (A) and 6268A (B) BLCLs are 10 incubated with LLDFVRFMGV (middle panel) or KLQCVDLHV peptides (right panel) and stained with EBNA clone 315 scFv, only HLA-A2+ DIMT peptide-pulsed with LLDFVRFMGV could be stained, showing that EBNA clone 315 and LLDFVRFMGV are HLA-A2 restricted; a time course (bottom panel) shows that the HLA-A2-LLDFVRFMGV complex is stable on the cell surface. 15 [0024] Figure 6 shows the Tomlinson library vector used in PCR to add appropriate restriction enzyme sites to either side of the WT1 Clone 45 and EBNA Clone 315 scFv sequences (Figure 6A). Figure 6B shows the digested PCR products as they appeared on a 1% agarose gel following digestion with Nhel and Apal. 20 [0025] Figure 7 shows the full IgG expression vector (Figure 7A) that was used to generate an expression vector for scFv-Fc fusion proteins (Figure 7B). A. The structure of the proprietary IgG expression vector (11381 bp). The vector expresses the heavy and light chains under two separate CMV promoters. The variable heavy chain (VH) is fused to the first, second and 25 third constant heavy chains (CH-i, 2,3) and expressed under one promoter while the variable light chain (VL) is fused to the constant light chain (Cl) and expressed under a different promoter. This vector was further modified to lack the first constant region of the heavy chain (CFL), and this vector was used for the construction of scFv-Fc fusion proteins. B. After excision of the VH from

30 the IgG vector using Nhel and Apal, the pre-digested, purified scFv PCR 8270468J (GHMatters) P94551.AU.1 5October2016 -7- 2014262282 05 Oct 2016 products were ligated to the IgG vector to allow for the expression of the scFv fused to the CH2,3 domains (Fc).

[0026] Figure 8 shows the results of binding studies using EBNA Clone 315 scFv-Fc in which purified EBNA Clone 315 scFv-Fc was shown to maintain its 5 binding ability towards the recombinant complex when tested for binding on an ELISA plate coated with or without HLA-A2-LLDFVRFMGV (Figure 8A) and when tested for binding on T2 cells pulsed with or without the LLDFVRFMGV peptide (Figure 8B). When T2 cells were incubated with decreasing concentrations of the LLDFVRFMGV peptide and subsequently 10 stained with EBNA Clone 315 scFv-Fc, a lower limit of detection was demonstrated to be in the same range as the scFv (200 nM- 20 nM) (Figure 8C).

[0027] Figure 9 shows the results of kinetics determination of EBNA Clone 315 to HLA-A2-LLDFVRFMGV using surface plasmon resonance. 15 [0028] Figure 10 shows the results of HLA-A2-LLDFVRFMGV complex quantitation on T2 cells using fluorescently-conjugated EBNA Clone 315 scFv-Fc (Figure 10A). Fluorescently-conjugated EBNA Clone 315 scFv-Fc was tested for binding on T2 cells pulsed with (20, 10, or 5 μΜ) or without (0 μΜ) the LLDFVRFMGV peptide in serum-free IMDM media at 37°C for 5 hours. 20 The cells, in addition to beads containing known amounts of anti-human IgGi antibodies, were stained with the scFv-Fc and the cell’s fluorescence intensity was correlated to that of the beads and their number of binding sites. Using these four peptide concentrations and corresponding number of complexes, a standard curve was created with an R2 value of 0.9948. Figure B shows a 25 close-up view of the lower end of the peptide and complex spectrum.

[0029] Figure 11 shows the results of binding and specificity studies when purified WT1 Clone 45 scFv-Fc was tested for binding on an ELISA plate coated with or without HLA-A2-RMFPNAPYL (Figure 11 A). Figure 11B shows that when purified WT1 Clone 45 scFv-Fc was tested for binding on T2 cells 8270468_1 (GHMatters) P94551 .AU.1 5 October 2016 -8- 2014262282 05 Oct 2016 pulsed with the RMFPNAPYL or RLTRFLSRV peptide (40 μΜ), the scFv-Fc (unfilled lines) was only able to recognize RMFPNAPYL-pulsed T2 cells.

[0030] Figure12 shows that when DIMT (top) and 6268A (bottom) BLCLs were incubated with RMFPNAPYL (right panel) and the peptide-pulsed BLCLs 5 were stained with WT1 Clone 315 scFv-Fc (unfilled lines) or a control scFv (filled lines), only the FILA-A2-positive RMFPNAPYL peptide-pulsed DIMT BLCLs could be stained.

[0031] Figure 13 shows EBNA Clone 315 scFv-Fc mediated ADCC (measured using 51 Cr release) of LLDFVRFMGV peptide-pulsed cells. 10 [0032] Figure 14 shows the MSCV-based vector (left panel) containing an IRES-GFP sequence along with ampicilin-resistance used for transduction and expression of anti-EBNA CAR in NK92MI cells. A The EBNA Clone 315 scFv sequence was cloned into the CAR gene (EBNA CAR) and further cloned into an MSCV-based vector (left panel) which contained an IRES-GFP 15 sequence along with ampicilin-resistance. The resulting CAR (right panel) is composed of the scFv and hinge region on the extracellular surface, a transmembrane domain, along with 4-1BB and the ΰϋ3ζ chain present within the cell. B After retroviral packaging using 293T GP2 cells and transduction into NK92MI cells, approximately 24% of the NK92MI cells contained the 20 construct based on GFP expression (left panel; unfilled lines) when compared to mock transduced (empty retrovirus) NK92MI cells (left panel; filled lines).

Of the GFP-positive cells, the top 20% were flow cytometrically sorted and expanded to yield a population of stably transduced cells which were greater than 90% GFP positive (right panel). Retroviral transduction was done on 25 three separate occasions, with 24% being the highest efficiency.

[0033] Figure 15 shows EcoRI and Xhol digestion validation of the WT1 Clone 45 CAR vector. A. Along with sequence validation, plasmids isolated from 8 different bacterial colonies, after ligation, transformation and EcoRI and Xhol digestion, were run on a 1% agarose gel. Based on the lambda Hindlll 30 and 100 bp markers, it was determined that the bands were the correct size 8270468_1 (GHMatters) P94551 .AU.1 5October2016 -9- 2014262282 05 Oct 2016 (-1500 bp and -6000 bp). B. The structure of the resulting WT1 Clone 45 CAR vector has the same components as the original St. Jude CAR vector with the only difference being the scFv sequence.

[0034] Figure 16 shows that Clone 315 CAR-expressing NK92MI cells can 5 specifically detect the HLA-A2-EBNA3C complex on peptide-pulsed T2 cells via CD107a expression. T2 cells were pulsed with or without LLDFVRFMGV or YMFPNAPYL peptides at 20 μΜ. CAR-equipped NK92MI cells were then cultured in media containing an anti-CD107a-PE conjugated antibody at 37°C for 5 hours with or without peptide pulsed or unpulsed cells. A. CAR-10 equipped NK92MI cells were gated based on GFP fluorescence and analyzed for CD107a expression. NK92MI cells which were cultured without any T2 cells or those which were cocultured with unpulsed and YMFPNAPYL-pulsed T2 cells were unreactive while NK92MI cells which were cocultured with LLDFVRFMGV-pulsed T2 cells led to a 27% increase in CD107a expression 15 above background levels. B. T2 cells were pulsed with decreasing concentrations of LLDFVRFMGV and subsequently cocultured with CAR-equipped NK92MI cells. NK92MI cells presented noticeable amounts of CD107a on their cell surface even when T2 cells were pulsed with only 10 nM of peptide. 20 [0035] Figure 17 shows the results of flow cytometry in which FILA-A2+ (DIMT) and HLA-A2’ (6268A) BLCLs were pulsed with LLDFVRFMGV and CAR-equipped NK92MI cells were then cultured in media containing an anti-CD107a-PE conjugated antibody with or without peptide pulsed or unpulsed cells. CAR-equipped NK92MI cells were gated based on GFP fluorescence 25 and analyzed for CD107a expression. NK92MI cells which were cultured without any BLCL or those which were cocultured with LLDFVRFMGV-pulsed 6268A BLCL were unreactive while NK92MI cells which were cocultured with unpulsed DIMT BLCL or LLDFVRFMGV-pulsed DIMT BLCL led to a 0.5% and 25% increase in CD107a expression above background levels (pulsed 6268A) 30 showing that EBNA Clone 315 CAR-expressing NK92MI cells can specifically detect the HLA-A2-EBNA3C complex on peptide-pulsed BLCLs via CD107a expression. 82704$8_1 (GHMatters) P94551 .AU.1 5 October 2016 -10- 2014262282 05 Oct 2016

[0036] Figure 18 shows the results of a 51 Cr release assay in whichT2 cells were pulsed with or without decreasing concentrations of LLDFVRFMGV. CAR-equipped NK92MI cells were cocultured with 51Cr-labeled T2 cells at a 3:1 E:T ratio demonstrating that EBNA Clone 315 CAR-expressing NK92MI 5 cells can specifically detect the HLA-A2-EBNA3C complex on peptide-pulsed T2 cells. Even with 2 nM of peptide, peptide-specific cytotoxicity could be observed when compared to unpulsed T2 cells.

[0037] Figure19 shows EBNA Clone 315 CAR-expressing NK92MI cells can specifically detect the HLA-A2-EBNA3C complex on peptide-pulsed BLCLs 10 via 51 Cr release. BLCLs were pulsed with LLDFVRFMGV. CAR-equipped NK92MI cells were then cultured with 51 Cr labeled target cells. A. CAR equipped NK92MI cells were able to specifically differentiate between peptide pulsed DIMT and 6268A BLCL, with a clear difference in cytotoxicity between the two different targets. B. CAR-mediated killing of peptide-pulsed DIMT 15 BLCL could be blocked using the EBNA Clone 315 scFv-Fc fusion protein, but not by an irrelevant, isotype-matched scFv-Fc, at a 20:1 E:T ratio.

[0038] Figure 20 shows the results of a 51 Cr release assay in which CAR-equipped NK92MI cells were cultured with 51Cr labeled, unpulsed BLCLs. A. CAR-equipped NK92MI cells were more reactive towards the HLA-A2+ DIMT 20 and JG19 BLCL versus the HLA-A2' 6268A and GKO BLCL when cocultured in the absence of any exogenous peptide. B. CAR-mediated killing of unpulsed DIMT BLCL could be blocked using the EBNA Clone 315 scFv-Fc fusion protein but not by an irrelevant, isotype-matched scFv-Fc, at a 10:1 E:T ratio demonstrating that EBNA Clone 315 CAR-expressing NK92MI cells can 25 specifically detect the HLA-A2-EBNA3C complex on HLA-A2+ BLCLs.

[0039] Figures 21 shows the results of a 51 Cr release assay of EBNA in which CD16(V)-expressing NK92MI cells were cultured with 51Cr labeled, LLDFVRFMGV-pulsed DIMT BLCL and either EBNA Clone 315 or an irrelevant scFv-Fc. At an E:T ratio of 15:1, EBNA Clone 315 scFv-Fc was 30 able to kill 30-35% of target cells. 8270468_1 (GHMatters) P94551 .AU.1 50dober2016 -11- 2014262282 05 Oct 2016 [0040] Figure 22 shows the results of a 51 Cr release assay of EBNA in which Clone 315 CAR-expressing NK92MI cells were cultured with LLDFVRFMGV-pulsed DIMT BLCL as above and either EBNA Clone 315 or an irrelevant scFv-Fc. At the same E:T ratio as in Figure 21, the CAR-5 equipped cells were able to kill 80-90% of target cells, with specific inhibition using EBNA Clone scFv-Fc demonstrating that CAR-mediated killing is more potent than scFv-Fc-mediated ADCC on peptide-pulsed DIMT BLCL.

[0041 ] Figure 23 shows the MSCV-based vector (left panel) containing an IRES-GFP sequence along with ampicilin-resistance used for transduction 10 and expression of anti-WT1 CAR in NK92MI cells. A. The WT1 Clone 45 scFv sequence was cloned into the CAR gene (anti-WT1 CAR) and further cloned into the MSCV-based vector. The resulting CAR (right panel) is composed of the scFv and hinge region on the extracellular surface, a transmembrane domain, along with 4-1BB and the ΰϋ3ζ chain present within 15 the cell. B After retroviral packaging using 293T GP2 cells and transduction into NK92MI cells, approximately 27.5% of the NK92MI cells contained the construct based on GFP expression (left panel; unfilled lines) when compared to mock transduced (empty retrovirus) NK92MI cells (left panel; filled lines).

Of the GFP-positive cells, the top 20% were flow cytometrically sorted and 20 expanded to yield a population of stably transduced cells which were greater than 98% GFP positive (right panel).

[0042] Figure 24 shows the results of a 51 Cr release assay in which CAR-equipped NK92MI cells were able to specifically differentiate between peptide pulsed DIMT () and 6268A BLCL (·), with a clear difference in cytotoxicity 25 between the two different targets demonstrating that NK92MI cells expressing WT1 Clone 45 CAR can specifically detect the HLA-A2-RMFPNAPYL complex on peptide-pulsed BLCLs.

[0043] Figure 25 shows that CAR-mediated killing of peptide-pulsed DIMT BLCL could be blocked using a commercial anti-HLA-A2 antibody (5 pg/ml), 30 but not by an irrelevant, isotype-matched antibody (5 pg/ml), at a 9:1 E:T ratio. 8270468_1 (GHMatters) P94551 AU.1 5 October 2016 -12- 2014262282 05 Oct 2016 [0044] Figure 26 shows that NK92MI cells expressing WT1 Clone 45 CAR can specifically detect the HLA-A2-RMFPNAPYL complex on DIMT BLCL via 51 Cr release. A. CAR equipped NK92MI cells were able to specifically differentiate between DIMT and 6268A BLCL, with a clear difference in 5 cytotoxicity between the two different targets. B. CAR-mediated killing of DIMT BLCL could be blocked using the WT1 Clone 45 scFv-Fc fusion protein (20 pg/ml), but not by an irrelavent, isotype-matched scFv-Fc, at a 2:1 E:T ratio.

[0045] Figure 27 shows that NK92MI cells expressing WT1 Clone 45 CAR 10 can specifically detect the HLA-A2-RMFPNAPYL complex on peptide-pulsed BLCLs via 51 Cr release. A CAR-equipped NK92MI cells were able to specifically differentiate between peptide pulsed DIMT and 6268A BLCL, with a clear difference in cytotoxicity between the two different targets.

[0046] Figure 28 shows that NK92MI cells expressing WT1 Clone 45 CAR 15 can specifically detect the HLA-A2-RMFPNAPYL complex on 697 and OVCAR-3 cells via 51 Cr release.

DETAILED DESCRIPTION OF THE INVENTION

[0047] All patents, publications, applications and other references cited herein are hereby incorporated in their entirety into the present application. It 20 is to be understood that, if any prior art publication is referred to herein, such reference does not constitute an admission that the publication forms a part of the common general knowledge in the art, in Australia or any other country.

[0048] In practicing the present invention, many conventional techniques in molecular biology, microbiology, cell biology, biochemistry, and immunology 25 are used, which are within the skill of the art. These techniques are described in greater detail in, for example, Molecular Cloning: a Laboratory Manual 3rd edition, J.F. Sambrook and D.W. Russell, ed. Cold Spring Harbor Laboratory Press 2001; Recombinant Antibodies for Immunotherapy, Melvyn Little, ed. Cambridge University Press 2009; "Oligonucleotide Synthesis" (M. J. Gait, 30 ed., 1984); "Animal Cell Culture" (R. I. Freshney, ed., 1987); "Methods in 8270468_1 (GHMatters) P94551 .AU.1 5October2016 -13- 2014262282 05 Oct 2016

Enzymology" (Academic Press, Inc.); "Current Protocols in Molecular Biology" (F. M. Ausubel et al., eds., 1987, and periodic updates); "PCR: The Polymerase Chain Reaction", (Mullis et al., ed., 1994); "A Practical Guide to Molecular Cloning" (Perbal Bernard V., 1988); "Phage Display: A Laboratory 5 Manual" (Barbas et al., 2001). The contents of these references and other references containing standard protocols, widely known to and relied upon by those of skill in the art, including manufacturers’ instructions are hereby incorporated by reference as part of the present disclosure.

[0049] The following abbreviations are used throughout the application: 10 [0050] ADCC: Antibody-dependent cellular cytotoxicity [0051] ALL: Acute lymphocytic leukemia [0052] AML: Acute myeloid leukemia [0053] APC: Antigen presenting cell [0054] β2Μ: Beta-2-microglobulin 15 [0055] BiTE: Bi-specific T cell engaging antibody [0056] BLCL: EBV-transformed B-cell lymphoblastic cell line [0057] CAR: Chimeric antigen receptor [0058] CDC: Complement dependent cytotoxicity [0059] CMC: Complement mediated cytotoxicity 20 [0060] CDR: Complementarity determining region (see also HVR below) [0061] Cl: Constant domain of the light chain [0062] CH-i: 1st constant domain of the heavy chain [0063] CHi, 2,3: 1st, 2nd and 3rd constant domains of the heavy chain 8270468_1 (GHMallers) P94551 .AU.1 5 October 201$ 2014262282 05 Oct 2016 -14- [0064] CH2,3: 2nd and 3rd constant domains of the heavy chain [0065] CHO: Chinese hamster ovary [0066] CTL: Cytotoxic T cell [0067] EBNA3C: Epstein-Barr nucleur antigen 3C 5 [0068] EBV: Epstein-Barr virus [0069] ECMV: Encephalomyocarditis virus [0070] ER: Endoplasmic reticulum [0071] E:T Ratio: EffectonTarget ratio [0072] Fab: Antibody binding fragment 10 [0073] FACS: Flow assisted cytometric cell sorting [0074] FBS: Fetal bovine serum [0075] GFP: Green fluorescence protein [0076] HC: Fleavy chain [0077] HEL: Hen egg lysozyme 15 [0078] HLA: Human leukocyte antigen [0079] HVR-H: Hypervariable region-heavy chain (see also CDR) [0080] HVR-L: Hypervariable region-light chain [0081] Ig: Immunoglobulin [0082] IPTG: isopropyl-1 -thio-p-D-galactopyranoside 20 [0083] IRES: Internal ribosome entry site [0084] Kd: Dissociation constant 8270468J (GHMatters) P94551.AU.1 5Oc1ober2016 -15- 2014262282 05 Oct 2016 [0085] koffi Dissociation rate [0086] kon: Association rate [0087] MHC: Major histocompatibility complex [0088] OPD: O-phenylenediamine 5 [0089] PEG: Polyethylene glycol [0090] scFv: Single-chain variable fragment [0091] SPR: Surface plasmon resonance [0092] TB: Terrific Broth [0093] TCE: T cell epitope 10 [0094] TCR: T cell receptor [0095] TIL: Tumor infiltrating lymphocyte [0096] VH: Variable heavy chain [0097] Vi_: Variable light chain [0098] WT1: Wilms tumor protein 1 15 [0099] In the description that follows, certain conventions will be followed as regards the usage of terminology. Generally, terms used herein are intended to be interpreted consistently with the meaning of those terms as they are known to those of skill in the art.

[00100] An “antigen-binding protein” is a protein or polypeptide that comprises 20 an antigen-binding region or antigen-binding portion, that is, has a strong affinity to another molecule to which it binds. Antigen-binding proteins encompass antibodies, antigen receptors and fusion proteins. 8270468_1 (GHMatters) P94551 .AU.1 5October2016 -16- 2014262282 05 Oct 2016 [00101] "Antibody" and "antibodies" as those terms are known in the art refer to antigen binding proteins that arise in the context of the immune system.

The term "antibody" as referred to herein includes whole, full length antibodies and any fragment thereof in which the "antigen-binding portion" or "antigen- 5 binding region" is retained or single chains thereof. A naturally occurring "antibody" is a glycoprotein comprising at least two heavy (H) chains and two light (L) chains inter-connected by disulfide bonds. Each heavy chain is comprised of a heavy chain variable region (abbreviated herein as VH) and a heavy chain constant region. The heavy chain constant region is comprised of 10 three domains, CH1, CH2 and CH3. Each light chain is comprised of a light chain variable region (abbreviated herein as VL) and a light chain constant region. The light chain constant region is comprised of one domain, Cl. The VH and VL regions can be further subdivided into regions of hypervariability, termed complementarity determining regions (CDR), interspersed with regions 15 that are more conserved, termed framework regions (FR). Each VH and VL is, composed of three CDRs and four FRs arranged from amino-terminus to carboxy-terminus in the following order: FR1, CDR1, FR2, CDR2, FR3, CDR3, FR4. The variable regions of the heavy and light chains contain a binding domain that interacts with an antigen. The constant regions of the antibodies 20 may mediate the binding of the immunoglobulin to host tissues or factors, including various cells of the immune system (e.g., effector cells) and the first component (C1q) of the classical complement system.

[00102] The term "antigen-binding portion" or "antigen-binding region" of an antibody (or simply "antigen portion"), as used herein, refers to that region or 25 portion of the antibody that confers antigen specificity; fragments of antigenbinding proteins, for example, antibodies therefore, includes one or more fragments of an antibody that retain the ability to specifically bind to an antigen (e.g., an HLA-peptide complex). It has been shown that the antigenbinding function of an antibody can be performed by fragments of a full-length 30 antibody. Examples of antigen-binding fragments encompassed within the term "antibody fragments" of an antibody include a Fab fragment, a monovalent fragment consisting of the VL, VH, CL and CH1 domains; a F(ab)2 8270468_1 (GHMatters) P94551 .AU.1 5 October 2016 -17- 2014262282 05 Oct 2016 fragment, a bivalent fragment comprising two Fab fragments linked by a disulfide bridge at the hinge region; a Fd fragment consisting of the VH and CH1 domains; a Fv fragment consisting of the VL and VH domains of a single arm of an antibody; a dAb fragment (Ward et al., 1989 Nature 341:544-546), 5 which consists of a Vh domain; and an isolated complementarity determining region (CDR).

[00103] Furthermore, although the two domains of the Fv fragment, VLand VH, are coded for by separate genes, they can be joined, using recombinant methods, by a synthetic linker that enables them to be made as a single 10 protein chain in which the Vl and Vh regions pair to form monovalent molecules (known as single chain Fv (scFv); see e.g., Bird et al., 1988 Science 242:423-426; and Fluston et al., 1988 Proc. Natl. Acad. Sci. 85:5879-5883). Such single chain antibodies are also intended to be encompassed within the term "antigen-binding portion" of an antibody. These antibody 15 fragments are obtained using conventional techniques known to those of skill in the art, and the fragments are screened for utility in the same manner as are intact antibodies.

[00104] An "isolated antibody" or “isolated antigen-binding protein” is one which has been identified and separated and/or recovered from a component 20 of its natural environment.

[00105] Traditionally, the MFIC-peptide complex could only be recognized by a T-cell receptor (TCR), limiting our ability to detect an epitope of interest to use of T cell-based readout assays. In the present disclosure, antigen binding proteins, including antibodies and chimeric antigen receptors, having an

25 antigen-binding region based on scFvs that are selected from human scFv phage display libraries using recombinant FILA-peptide complexes are described. These molecules demonstrated exquisite specificity, for example as shown with anti-EBNA and anti-WT1 antigen-binding proteins that recognize only the HLA-A2-LLDFVRFMGV and HLA-A2-RMFPNAPYL 30 complexes, respectively. In addition, along with their inability to bind to HLA-8270468_1 (GFIMatters) P94551 .AU.1 5October2016 -18- 2014262282 05 Oct 2016 complexes containing other peptides, the molecules were also unable to bind to the peptides themselves, further demonstrating their TCR-like specificity.

[00106] The scFvs of the disclosure selected by phage display were initially tested for their ability to bind to peptide presented on the surface of HLA-5 positive cells. After T2 cells and BLCLs were incubated in the presence of peptide, the cells could selectively recognize them using flow cytometry. In the case of one peptide, LLDFVRFMGV (SEQ ID NO: ), the complex which the peptide formed with FILA could be detected on the surface of a BLCL even 24 hours after pulsing, further demonstrating the utility of these antibodies. 10 [00107] In some embodiments, the antigen binding proteins of the invention include antibodies that have the scFv sequence fused to the 2nd and 3rd constant domains of the heavy chain (CH2,3), forming the bottom third of the Fc region of a human immunoglobulin to yield a bivalent protein and fragments thereof, increasing the overall avidity and stability of the antibody. 15 In addition, the Fc portion allows the direct conjugation of other molecules, including but not limited to fluorescent dyes, cytotoxins, radioisotopes etc. to the antibody for example, for use in antigen quantitation studies, to immobilize the antibody for affinity measurements using surface plasmon resonance (SPR), for targeted delivery of a therapeutic agent, to test for Fc-mediated 20 cytotoxicity using CD16-expressing immune effector cells and many other applications.

[00108] The purified scFv-Fc fusion proteins were tested for binding to their targeted T-cell epitopes (TCEs) by way of ELISA and peptide-pulsed APCs. Once they were validated to maintain their specificity, one molecule, EBNA 25 Clone 315 was used for affinity determination. That this molecule was able to bind bound to its targeted TCE through a 1:1 interaction with 10-100 fold greater affinity compared to a typical TCR-MHC-peptide complex interaction was demonstrated.

[00109] Correlation of peptide pulsing of APCs with antigen density was 30 demonstrated. Fluorescently-conjugated scFv-Fc, combined with quantitation 82704$8_1 (GHMallers) P94551 .AU.1 5 October 2016 -19- 2014262282 05 Oct 2016 beads, allowed the approximation of the number of complexes that are formed when cells are incubated with different concentrations of peptide. Using this information, it was possible to approximate the sensitivity of an scFv and scFv-Fc fusion protein to be around 100 complexes, using flow cytometry. 5 [00110] Lastly, whether the Fc portion of the fusion protein maintained its effector function was tested. Using a scFv embodiment of the invention, CD16(V)-transduced NK92MI cells, and peptide-pulsed target cells, it was demonstrated that the antibody maintained its Fc-mediated effector functions by way of ADCC. 10 [00111 ] The results presented here highlight the specificity, sensitivity and utility of the antigen binding proteins of the invention in targeting MHC-peptide complexes.

[00112] In one embodiment, therefore, the present invention relates to recombinant antigen-binding molecules and portions thereof that recognize a 15 complex of a peptide/protein fragment derived from an intracellular or viral protein, and an MHC class I molecule, for example, as the complex might be appear at the cell surface for recognition by a T-cell.

[00113] The molecules of the invention are based on the identification and selection of a single chain variable fragment (scFv) using phage display, the 20 amino acid sequence of which confers the molecules' specificity for the MHC restricted peptide of interest and forms the basis of all antigen binding proteins of the disclosure. The scFv, therefore, can be used to design a diverse array of "antibody" molecules, including, for example, full length antibodies, fragments thereof, such as Fab and F(ab')2, minibodies, fusion proteins, 25 including scFv-Fc fusions, multivalent antibodies, that is, antibodies that have more than one specificity for the same antigen or different antigens, for example, bispecific T-cell engaging antibodies (BiTe), tribodies, etc. (see Cuesta et al., Multivalent antibodies: when design surpasses evolution.

Trends in Biotechnology 28:355-362 2010). 8270468_1 (GHMatters) P94551 .AU.1 5October2016 -20- 2014262282 05 Oct 2016 tOOH 4] In an embodiment in which the antigen-binding protein is a full length antibody, the heavy and light chains of an antibody of the invention may be full-length (e.g., an antibody can include at least one, and preferably two, complete heavy chains, and at least one, and preferably two, complete light 5 chains) or may include an antigen-binding portion (a Fab, F(ab')2, Fv or a single chain Fv fragment ("scFv")). In other embodiments, the antibody heavy chain constant region is chosen from, e.g., lgG1, lgG2, lgG3, lgG4, IgM, lgA1, lgA2, IgD, and IgE. In some embodiments, the immunoglobulin isotype is selected from lgG1, lgG2, lgG3, and lgG4, more particularly, lgG1 (e.g., 10 human lgG1). The choice of antibody type will depend on the immune effector function that the antibody is designed to elicit.

[00115] In constructing a recombinant immunoglobulin, appropriate amino acid sequences for constant regions of various immunoglobulin isotypes and methods for the production of a wide array of antibodies are well known to 15 those of skill in the art.

[00116] In some embodiments, the constant region of the antibody is altered, e.g., mutated, to modify the properties of the antibody (e.g., to increase or decrease one or more of: Fc receptor binding, antibody carbohydrate, for example glycosylation or fucosylation, the number of cysteine residues, 20 effector cell function, or complement function).

[00117] In one embodiment, the antigen binding protein is an anti-EBNA3C antibody or fragment thereof that has an antigen binding region that comprises the amino acid sequence of SEQ ID NO: 5 and specifically binds to 25 a peptide with the amino acid sequence LLDFVRFMGV (SEQ ID NO: 4) in conjunction with HLA-A2. In other embodiments, the anti-EBNA3C antibody is a scFv-Fc fusion protein or full length human IgG with VH and VL regions or CDRs selected from Table 1. 8270468_1 (GHMatters) P94551 .AU.1 5 October 201$ -21- 2014262282 05 Oct 2016

Table 1 Antigen EBNA3C Peptide LLDFVRFMGV (SEQ ID NO: 4 CDRs 1 2 3 VH GYAMS (SEQ ID NO: 39) EIAPPGLNTRYADSVKG (SEQ ID NO: 41) SDTAFDY (SEQ ID NO: 49) VL RASQSISSYLN (SEQ ID NO: 56) LASNLQS (SEQ ID NO: 58) QQAEYMPLT (SEQ ID NO: 65) Full VH EVQLLESGGGLVQPGGSLRLSCAASG FTFSGYAM SWVRQAPG KG LEWVSEIAPPG LNTRYADSVKG RFTIS RDNSKNTLYLQMNSLRAEDTAVYYCAKSDTAFDYWGQGTLVTVS (SEQ ID NO: 24) Full VL STDIQMTQSPSSLSASVGDRVTITCRASQSISSYLNWYQQKPGKAPKLLIYLASNLQSGVPSRFSGSGSGTDF TLTISSLQPEDFATYYCQQAEYMPLTFGQGTKVEIKRA (SEQ ID NO: 25 ) scFv clone 315 EVQLLESGGGLVQPGGSLRLSCAASG FTFSGYAM SWVRQAPG KG LEWVSEIAPPG LNTRYADSVKG RFTIS RDNSKNTLYLQMNSLRAEDTAVYYCAKSDTAFDYWGQGTLVTVSSGGGGSGGGGSGGGGSTDIQMTQS PSSLSASVGDRVTITCRASQSISSYLNWYQQKPGKAPKLLIYLASNLQSGVPSRFSGSGSGTDFTLTISSLQPE DFATYYCQQAEYMPLTFGQGTKVEIKRA (SEQ ID NO: 5) DNA (5' -3') GAGGTGCAGCTGTTGGAGTCTGGGGGAGGCTTGGTACAGCCTGGGGGGTCCCTGAGACTCTCCTGTG CAGCCTCTGGATTCACCTTTAGCGGCTATGCCATGAGCTGGGTCCGCCAGGCTCCAGGGAAGGGGCTG GAGTGGGTCTCAGAGATTGCGCCGCCTGGTTTGAATACACGTTACGCAGACTCCGTGAAGGGCCGGTT CACTAT CT CCAG AG ACAATTCCAAG AACACGCT GTAT CTGCAAATG AACAGCCT G AG AGCCG AGG ACA CGGCCGTATATTACTGTGCGAAATCGGATACTGCTTTTGACTACTGGGGCCAGGGAACCCTGGTCACC GTCTCG AGCGGTGG AGGCGGTTCAGGCGG AGGTGGCAGCGGCGGT GGCGGGT CG ACGG ACAT CCA GAT G ACCCAGT CT CCAT CCTCCCTGTCTGCAT CT GT AGG AG ACAG AGT CACCATCACTT GCCGGGCAAG TCAGAGCATTAGCAGCTATTTAAATTGGTATCAGCAGAAACCAGGGAAAGCCCCTAAGCTCCTGATCT ATCTGGCATCCAATTTGCAAAGTGGGGTCCCATCAAGGTTCAGTGGCAGTGGATCTGGGACAGATTTC ACTCTCACCATCAGCAGT CTGCAACCT G AAG ATTTT GCAACTTACT ACT GTCAACAGGCGG AGT AT AT G CCTCTGACGTTCGGCCAAGGGACCAAGGTGGAAATCAAACGGGCC (SEQ ID NO: 6) [00118] In yet another embodiment, the antigen binding protein is an anti-CCND1 antibody or fragment thereof that comprises the amino acid sequence 5 of one of SEQ ID NOs: 8 or 10 and specifically binds to a peptide with the amino acids sequence RLTRFLSRV (SEQ ID NO: 7) in conjunction with HLA-A2. In other embodiments, the anti-CCND1 antibody is a scFv-Fc fusion or full length human IgG with VFI and VL regions or CDRs selected from Tables 2 and 3. 10 82704$8_1 (GHMatters) P94551 .AU.1 5 October 201$ -22- 2014262282 05 Oct 2016

Table 2 Antigen CCND1 Peptide RLTRFLSRV (SEQ ID NO. 7) CDRs 1 2 3 VH SYAMS(38) TISDSDATDYADSVKG(42) TTDYFDY(50) VL RASQSISSYLN(56) YASYLQS(59) QQSSSSPDT(66) Full VH EVQ.LLESGGGLVQ.PGGSLRLSCATSGFTFSSYAMSWVRQAPGKGLEWVSTISDSDATDYADSVKGRFTISR DNSKNTLYLQMNSLRAEDTAVYYCAKTTDYFDYWGQGTLVTVS(26) Full VL STDIQMTQSPSSLSASVGDRVTITCRASQSISSYLNWYQQKPGKAPKLLIYYASYLQSGVPSRFSGSGSGTDF TLTISSLQPEDFATYYCQQSSSSPDTFGQGTKVEIKRAA(27) scFv EVQLLESGGGLVQ.PGGSLRLSCATSGFTFSSYAMSWVRQAPGKGLEWVSTISDSDATDYADSVKGRFTISR clone 5, DNSKNTLYLQMNSLRAEDTAVYYCAKTTDYFDYWGQGTLVTVSSGGGGSGGGGSGGGGSTDIQMTQSP 17 SSLSASVGDRVTITCRASQSISSYLNWYQQKPGKAPKLLIYYASYLQSGVPSRFSGSGSGTDFTLTISSLQPED FATYYCQQSSSSPDTFGQGTKVEIKRAA(8) DNA GAGGTGCAGCTGTTGGAGTCTGGGGGAGGCTTGGTACAGCCTGGGGGGTCCCTGAGACTCTCCTGTG (5' -3') CAACCTCTGGATTCACCTTTAGCAGCTATGCCATGAGCTGGGTCCGCCAGGCTCCAGGGAAGGGGCTG GAGTGGGTCTCAACTATTTCTGATAGTGATGCTACAGATTACGCAGACTCCGTGAAGGGCAGGTTCAC CATCTCCAGAGACAATTCCAAGAACACGCTGTATCTGCAAATGAACAGCCTGAGAGCCGAGGACACG GCCGTATATTACTGTGCGAAAACTACTGATTATTTTGACTACTGGGGCCAGGGAACCCTGGTCACCGTC TCGAGCGGTGGAGGCGGTTCAGGCGGAGGTGGCAGCGGCGGTGGCGGGTCGACGGACATCCAGAT GACCCAGTCTCCATCCTCCCTGTCTGCATCTGTAGGAGACAGAGTCACCATCACTTGCCGGGCAAGTCA GAGCATTAGCAGCTATTTAAATTGGTATCAGCAGAAACCAGGGAAAGCCCCTAAGCTCCTGATCTATT ATGCATCCTATTTGCAAAGTGGGGTCCCATCAAGGTTCAGTGGCAGTGGATCTGGGACAGATTTCACT CT C ACC AT C AG C AGTCTG C AACCTG A AG ATTTT G C AACTTACT ACT GT CAAC AGT CTT CTAGTT CTCCTG ATACGTTCGGCCAAGGGACCAAGGTGGAAATCAAACGGGCGGCC(9) 8270468_1 (GHMallers) P94551 .AU.1 5 October 2016 -23- 2014262282 05 Oct 2016

Table 3 Antigen CCND1 Peptide RLTRFLSRV(SEQ ID NO: 7) CDRs: 1 2 3 VH SYAMS(38) DISDDGDATYYADSVKG(43) SSTTFDY(51) VL RASQSISSYLN(56) AASALQS(60) QQGTDSPAT(67) Full VH EVQLLESGGGLVQPGGSLRLSCAASGFTFSSYAMSWVRQAPGKGLEWVSDISDDGDATYYADSVKGRFTI SRDNSKNTLYLQMNSLRAEDTAVYYCAKSSTTFDYWGQGTLVTVS(28) Full VL STDIQMTQSPSSLSASVGDRVTITCRASQSISSYLNWYQQKPGKAPKLLIYAASALQSGVPSRFSGSGSGTDF TLTISSLQPEDFATYYCQQGTDSPATFGQGTKVEIKRAA(29) scFv clone 43 EVQLLESGGGLVQPGGSLRLSCAASGFTFSSYAMSWVRQAPGKGLEWVSDISDDGDATYYADSVKGRFTI SRDNSKNTLYLQMNSLRAEDTAVYYCAKSSTTFDYWGQGTLVTVSSGGGGSGGGGSGGGGSTDIQMTQ SPSSLSASVGDRVTITCRASQSISSYLNWYQQKPGKAPKLUYAASALQSGVPSRFSGSGSGTDFTLTISSLQP EDFATYYCQQGTDSPATFGQGTKVEIKRAA(IO) DNA (5' -3') GAGGTGCAGCTGTTGGAGTCTGGGGGAGGCTTGGTACAGCCTGGGGGGTCCCTGAGACTCTCCTGTG CAGCCTCTGGATTCACCTTTAGCAGCTATGCCATGAGCTGGGTCCGCCAGGCTCCAGGGAAGGGGCTG GAGTGGGTCTCAGATATTTCTGATGATGGTGATGCTACATATTACGCAGACTCCGTGAAGGGCAGGTT CACCAT CTCCAG AG ACAATT CCAAG AACACGCT GTAT CTGCAAAT G AACAGCCT G AG AGCCG AGG ACA CGGCCGTATATTACTGTGCGAAATCTTCTACTACTTTTGACTACTGGGGCCAGGGAACCCTGGTCACCG TCTCGAGCGGTGGAGGCGGTTCAGGCGGAGGTGGCAGCGGCGGTGGCGGGTCGACGGACATCCAGA TGACCCAGTCTCCATCCTCCCTGTCTGCATCTGTAGGAGACAGAGTCACCATCACTTGCCGGGCAAGTC AGAGCATTAGCAGCTATTTAAATTGGTATCAGCAGAAACCAGGGAAAGCCCCTAAGCTCCTGATCTAT GCTGCATCCGCCTTGCAAAGTGGGGTCCCATCAAGGTTCAGTGGCAGTGGATCTGGGACAGATTTCAC TCTC ACC AT C AG C AGTCTG C A ACCTG AAG ΑΤΠΤ GCAACTT ACTACT GT CAACAGGGT ACT GAT AGT CC TGCTACGTTCGGCCAAGGGACCAAGGTGGAAATCAAACGGGCGGCC(ll) [00119] In yet another embodiment, the antigen binding protein is an anti-HUD antibody or fragment thereof that comprises the amino acid sequence of 5 one of SEQ ID NOs: 13, 14 and 17 and has an antigen-binding region that specifically binds to a peptide with the amino acid sequence RIITSTILV (SEQ ID NO: 12) in conjunction with HLA-A2. In other embodiments, the anti-HUD antibody is a scFv-Fc fusion protein or full length human IgG with VFI and VL regions or CDRs selected from Tables 4-6. 10 8270468_1 (GHMatters) P94551 .AU.1 5October2016 -24- 2014262282 05 Oct 2016

Table 4 Antigen HUD Peptide RIITSTILV(SEQ ID NO: 12) CDRs: 1 2 3 VH SYAMS(38) DIASTGYYTDYADSVKG(44) NNASFDY(52) VL RASQSISSYLN(56) DASTLQS(61) QQTDSYPTT(68) Full VH EVQLLESGGGLVQPGGSLRLSCAASG FTFSSYAM SWVRQAPG KGLEWVSD1 AST GYYTDYADSVKG RFTIS RDNSKNTLYLQMNSLRAEDTAVYYCAKNNASFDYWGQGTLVTVS(30) Full VL STDIQMTQSPSSLSASVGDRVTITCRASQSISSYLNWYQQKPGKAPKLLIYDASTLQSGVPSRFSGSGSGTDF TLTISSLQPEDFATYYCQQTDSYPTTFGQGTKVEIKR(31) scFv EVQLLESGGGLVQPGGSLRLSCAASG FTFSSYAM SWVRQAPG KG LE WVSDIASTGYYTDYADSVKG RFTIS clone RDNSKNTLYLQMNSLRAEDTAVYYCAKNNASFDYWGQGTLVTVSSGGGGSGGGGSGGGGSTDIQMTQ H128 SPSSLSASVGDRVTITCRASQSISSYLNWYQQKPGKAPKLLIYDASTLQSGVPSRFSGSGSGTDFTLTISSLQP EDFATYYCQQTDSYPTTFGQGTKVEIKR(13) DNA GAGGTGCAGCTGTTGGAGTCTGGGGGAGGCTTGGTACAGCCTGGGGGGTCCCTGAGACTCTCCTGTG (5' -3') CAGCCTCTGGATTCACCTTTAGCAGCTATGCCATGAGCTGGGTCCGCCAGGCTCCAGGGAAGGGGCTG GAGTGGGTCTCGGATATTGCTTCTACTGGTTATTATACAGATTACGCAGACTCCGTGAAGGGCCGGTT CACCATCTCCAGAGACAATTCCAAGAACACGCTGTATCTGCAAATGAACAGCCTGAGAGCCGAGGACA CGGCCGTATATTACTGTGCGAAAAATAATGCTAGTTTTGACTACTGGGGCCAGGGAACCCTGGTCACC GTCTCGAGCGGTGGAGGCGGTTCAGGCGGAGGTGGCAGCGGCGGTGGCGGGTCGACGGACATCCA GATGACCCAGTCTCCATCCTCCCTGTCTGCATCTGTAGGAGACAGAGTCACCATCACTTGCCGGGCAAG TC AG AG C ATT AG C AG CT ATTT A A ATT G GTATC AG C AG AAACCAGGG AAAGCCCCTAAGCT CCTG ATCT ATGATGCATCCACTTTGCAAAGTGGGGTCCCATCAAGGTTCAGTGGCAGTGGATCTGGGACAGATTTC ACT CTC ACC ATC AG C AGTCTG C A ACCTG AAG ATTTTGC AACTTACT ACT GT C AAC AG ACT GATT CTT AT C CTACTACGTTCGGCCAAGGGACCAAGGTGGAAATCAAACGG(15) 8270468_1 (GHMatters) P94551 .AU.1 5Ocl0ber2016 -25- 2014262282 05 Oct 2016

Table 5 Antigen HUD Peptide RIITSTILV(SEQ ID NO: 12) CDRs: 1 2 3 VH SYAMS(38) SISSSGYYTDYADSVKG(45) SASSFDY(53) VL RASQSISSYLN(56) DASTLQS(61) QQDDAYPTT(69) Full VH EVQLLESGGGLVQPGGSLRLSCAASGFTFSSYAMSWVRQAPGKGLEWVSSISSSGSYTDYADSVKGRFTIS RDNSKNTLYLQMNSLRAEDTAVYYCAKSASSFDYWGQGTLVTVS(32) Full VL STDIQMTQSPSSLSASVGDRVTITCRASQSISSYLNWYQQKPGKAPKLLIYDASTLQSGVPSRFSGSGSGTDF TLTISSLQPEDFATYYCQQDDAYPTTFGQGTKVEIKR(33) scFv EVQLLESGGGLVQPGGSLRLSCAASGFTFSSYAMSWVRQAPGKGLEWVSSISSSGSYTDYADSVKGRFTIS clone RDNSKNTLYLQ.MNSLRAEDTAVYYCAKSASSFDYWGQGTLVTVSSGGGGSGGGGSGGGGSTDIQMTQS H78 PSSLSASVGDRVTITCRASQSISSYLNWYQQKPGKAPKLLIYDASTLQSGVPSRFSGSGSGTDFTLTISSLQPE DFATYYCQQDDAYPTTFGQGTKVEIKR(14) DNA GAGGTGCAGCTGTTGGAGTCTGGGGGAGGCTTGGTACAGCCTGGGGGGTCCCTGAGACTCTCCTGTG (5' -3') CAGCCTCTGGATTCACCTTTAGCAGCTATGCCATGAGCTGGGTCCGCCAGGCTCCAGGGAAGGGGCTG GAGTGGGTCTCATCTATTAGTAGTTCTGGTAGTTATACAGATTACGCAGACTCCGTGAAGGGCCGGTT CACCATCTCCAGAGACAATTCCAAGAACACGCTGTATCTGCAAATGAACAGCCTGAGAGCCGAGGACA CGGCCGTATATTACTGTGCGAAATCTGCTTCTTCTTTTGACTACTGGGGCCAGGGAACCCTGGTCACCG TCTCGAGCGGTGGAGGCGGTTCAGGCGGAGGTGGCAGCGGCGGTGGCGGGTCGACGGACATCCAGA TGACCCAGTCTCCATCCTCCCTGTCTGCATCTGTAGGAGACAGAGTCACCATCACTTGCCGGGCAAGTC AGAGCATTAGCAGCTATTTAAATTGGTATCAGCAGAAACCAGGGAAAGCCCCTAAGCTCCTGATCTAT GATGCATCCACTTTGCAAAGTGGGGTCCCATCAAGGTTCAGTGGCAGTGGATCTGGGACAGATTTCAC TCTCACCATCAGCAGTCTGCAACCTGAAGATTTTGCAACTTACTACTGTCAACAGGATGATGCTTATCCT ACTACGTTCGGCCAAGGGACCAAGGTGGAAATCAAACGG(16) 82704$8_1 (GHMatters) P94551 AU.1 5 October 2016 -26- 2014262282 05 Oct 2016

Table 6 Antigen HUD Peptide RIITSTILV(SEQ ID NO: 12) CDRs: 1 2 3 VH SYAMS(38) SISSDGSYTDYADSVKG(46) STDAFDY(54) VL RASQSISSYLN(56) AASYLQS(62) QQDNNYPTT(70) Full VH EVQLLESGGGLVQPGGSLRLSCAASGFTFSSYAMSWVRQAPGKGLEWVSSISSDGSYTDYADSVKGRFTIS RDNSKNTLYLQMNSLRAEDTAVYYCAKSTDAFDYWGQGTLVTVS(34) Full VL STDIQMTQSPSSLSASVGDRVTITCRASQSISSYLNWYQQKPGKAPKLLIYAASYLQSGVPSRFSGSGSGTDF SLTISSLQPEDFATYYCQQDNNYPTTFGQGTKVEIKR(35) scFv EVQLLESGGGLVQPGGSLRLSCAASGFTFSSYAMSWVRQAPGKGLEWVSSISSDGSYTDYADSVKGRFTIS clone RDNSKNTLYLQMNSLRAEDTAVYYCAKSTDAFDYWGQGTLVTVSSGGGGSGGGGSGGGGSTDIQMTQS H110 PSSLSASVGDRVTITCRASQSISSYLNWYQQKPGKAPKLLIYAASYLQSGVPSRFSGSGSGTDFSLTISSLQPE DFATYYCQQDNNYPTTFGQGTKVEIKR(17) DNA GAGGTGCAGCTGTTGGAGTCTGGGGGAGGCTTGGTACAGCCTGGGGGGTCCCTGAGACTCTCCTGTG (5' -3') CAGCCTCTGGATTCACCTTTAGCAGCTATGCCATGAGCTGGGTCCGCCAGGCTCCAGGGAAGGGGCTG GAGTGGGTCTCAAGTATTTCTTCTGATGGTAGTTATACAGATTACGCAGACTCCGTGAAGGGCCGGTT CACCATCTCCAG AG ACAATT CCAAG AACACGCT GTAT CTGCAAAT G AACAGCCT G AG AGCCG AGG ACA CGGCCGTATATTACTGTGCGAAATCTACTGATGCTTTTGACTACTGGGGCCAGGGAACCCTGGTCACC GTCTCGAGCGGTGGAGGCGGTTCAGGCGGAGGTGGCAGCGGCGGTGGCGGGTCGACGGACATCCA G ATG ACCCAGT CT CC AT CCTCCCTGT CTGCAT CT GT AGG AG ACAG AGT CACCATCACTT GCCGGGCAAG TCAGAGCATTAGCAGCTATTTAAATTGGTATCAGCAGAAACCAGGGAAAGCCCCTAAGCTCCTGATCT ATGCTGCATCCTATTTGCAAAGTGGGGTCCCATCAAGGTTCAGTGGCAGTGGATCTGGGACAGATTTC T CT CT C ACC AT C AG C AGTCTG C A ACCT G AAG ATTTT GC AACTTACT ACT GT C AAC AGG ATAAT AATT AT C CTACTACGTTCGGCCAAGGGACCAAGGTGGAAATCAAACGG(18) [00120] In yet another embodiment, the antigen binding protein is an anti-cdr2 antibody or fragment thereof that comprises the amino acid sequence of SEQ 5 ID NO: 20 and specifically binds to a peptide with amino acids LLEEMFLTV (SEQ ID NO: 19) in conjunction with HLA-A2. In other embodiments, the anti-cdr2 antibody is a scFv-Fc fusion protein or full length human IgG with VFI and VL regions or CDRs selected from Table 7. 10 8270468_1 (GHMallers) P94551 .AU.1 5 October 201$ -27- 2014262282 05 Oct 2016

Table 7 Antigen cdr2 Peptide LLEEMFLTV(SEQ ID NO: 19) CDRs: 1 2 3 VH SYAMS(38) Tl N YSGSGTTYADSVKG(47) NAAYFDY(55) VL RASQSISSYLN(56) GASGLQS(63) QQSANAPTT(71) Full VH EVQLLESGGGLVQPGGSLRLSCAASG FTFSSYAM S WVRQAPG KG LE WVSTIN YSGSGTTYADSVKG RFTIS RDNSKNTLYLQMNSLRAEDTAVYYCAKNAAYFDYWGQGTLVTVS(36) Full VL STDIQMTQSPSSLSASVGDRVTITCRASQSISSYLNWYQQKPGKAPKLLIYGASGLQSGVPSRFSGSGSGTD FTLTISSLQPEDFATYYCQQSANAPTTFGQGTKVEIKR(37) scFv EVQLLESGGGLVQPGGSLRLSCAASG FTFSSYAM S WVRQAPG KG LE WVSTI N YSGSGTTYADSVKG RFTIS clone RDNSKNTLYLQMNSLRAEDTAVYYCAKNAAYFDYWGQGTLVTVSSGGGGSGGGGSGGGGSTDIQMTQS L9 PSSLSASVGDRVTITCRASQSISSYLNWYQQKPGKAPKLLIYGASGLQSGVPSRFSGSGSGTDFTLTISSLQPE DFATYYCQQSANAPTTFGQGTKVEIKR(20) DNA GAGGTGCAGCTGTTGGAGTCTGGGGGAGGCTTGGTACAGCCTGGGGGGTCCCTGAGACTCTCCTGTG (5' -3') CAGCCTCTGGATTCACCTTTAGCAGCTATGCCATGAGCTGGGTCCGCCAGGCTCCAGGGAAGGGGCTG GAGTGGGTCTCAACTATTAATTATTCTGGTTCTGGTACAACTTACGCAGACTCCGTGAAGGGCAGGTTC ACCATCTCCAG AG ACAATT CCAAG AACACGCT GTAT CT GCAAAT G AACAGCCT G AG AGCCG AGG ACAC GGCCGTATATTACTGTGCGAAAAATGCTGCTTATTTTGACTACTGGGGCCAGGGAACCCTGGTCACCG TCTCGAGCGGTGGAGGCGGTTCAGGCGGAGGTGGCAGCGGCGGTGGCGGGTCGACGGACATCCAGA TGACCCAGTCTCCATCCTCCCTGTCTGCATCTGTAGGAGACAGAGTCACCATCACTTGCCGGGCAAGTC AGAGCATTAGCAGCTATTTAAATTGGTATCAGCAGAAACCAGGGAAAGCCCCTAAGCTCCTGATCTAT GGTGCATCCGGTTTGCAAAGTGGGGTCCCATCAAGGTTCAGTGGCAGTGGATCTGGGACAGATTTCAC TCTCACCATCAGCAGTCTGCAACCTGAAGATTTTGCAACTTACTACTGTCAACAGTCTGCTAATGCTCCT ACTACGTTCGGCCAAGGGACCAAGGTGGAAATCAAACGG(21) [00121] Embodiments of the antigen-binding proteins of the disclosure in accordance with Tables 1-7 include, but are not limited to the following: 5 [00122] an anti-EBNA3C antibody which binds to HLA-A2 restricted peptide LLDFVRFMGV (SEQ ID NO: 4) comprising: (i) an FIVR-L1 sequence of RASQSISSYLN (SEQ ID NO: 56) (ii) an HVR-L2 sequence of LASNLQS (SEQ ID NO: 58) (iii) an HVR-L3 sequence of QQAEYMPLT (SEQ ID NO: 65) (iv) an HVR-H1 sequence of GYAMS (SEQ ID NO: 39) (v) an HVR-H2 10 sequence of EIAPPGLNTRYADSVKG (SEQ ID NO: 41), and (vi) an HVR-H3 sequence of SDTAFDY (SEQ ID NO: 49); [00123] an anti-CCND1 antibody which binds to FILA-A2 restricted peptide RLTRFLSRV (SEQ ID NO: 7) comprising: (i) an FIVR-L1 sequence of RASQSISSYLN (SEQ ID NO: 56) (ii) an HVR-L2 sequence of YASYLQS 15 (SEQ ID NO: 59) (iii) an HVR-L3 sequence of QQSSSSPDT (SEQ ID NO: 66) 8270466.1 (GHMatters) P94551 AU.1 5 October 2016 -28- 2014262282 05 Oct 2016 (iv) an HVR-H1 sequence of SYAMS (SEQ ID NO: 38) (v) an HVR-H2 sequence of TISDSDATDYADSVKG (SEQ ID NO: 42), and (vi) an HVR-H3 sequence of TTDYFDY (SEQ ID NO: 50); [00124] an anti-CCND1 antibody which binds to HLA-A2 restricted peptide 5 RLTRFLSRV (SEQ ID NO: 7) comprising: (i) an HVR-L1 sequence of RASQSISSYLN (SEQ ID NO: 56) (ii) an HVR-L2 sequence of AASALQS (SEQ ID NO: 60) (iii) an HVR-L3 sequence of QQGTDSPAT (SEQ ID NO: 67) (iv) an HVR-H1 sequence of SYAMS (SEQ ID NO: 38) (v) an HVR-H2 sequence of DISDDGDATYYADSVKG (SEQ ID NO: 43), and (vi) an HVR-H3 10 sequence of SSTTFDY (SEQ ID NO: 51); [00125] an anti-HUD antibody which binds to FILA-A2 restricted peptide RIITSTILV (SEQ ID NO: 12) comprising: (i) an FIVR-L1 sequence of RASQSISSYLN (SEQ ID NO: 56) (ii) an HVR-L2 sequence of DASTLQS (SEQ ID NO: 61) (iii) an HVR-L3 sequence of QQTDSYPTT (SEQ ID NO: 68) 15 (iv) an HVR-H1 sequence of SYAMS (SEQ ID NO: 38) (v) an HVR-H2 sequence of DI AST GYYTDYADSVKG (SEQ ID NO: 44), and (vi) an HVR-H3 sequence of NNASFDY (SEQ ID NO: 52); [00126] an anti-HUD antibody which binds to HLA-A2 restricted peptide RIITSTILV (SEQ ID NO: 12) comprising: (i) an HVR-L1 sequence of

20 RASQSISSYLN (SEQ ID NO: 56) (ii) an HVR-L2 sequence of DASTLQS (SEQ ID NO: 61) (iii) an HVR-L3 sequence of QQDDAYPTT (SEQ ID NO: 69) (iv) an HVR-H1 sequence of SYAMS (SEQ ID NO: 38) (v) an HVR-H2 sequence of SISSSGYYTDYADSVKG (SEQ ID NO: 45), and (vi) an HVR-H3 sequence of SASSFDY (SEQ ID NO: 53); 25 [00127] an anti-HUD antibody which binds to HLA-A2 restricted peptide RIITSTILV (SEQ ID NO: 12) comprising: (i) an HVR-L1 sequence of RASQSISSYLN (SEQ ID NO: 56) (ii) an HVR-L2 sequence of AASYLQS (SEQ ID NO: 62) (iii) an HVR-L3 sequence of QQDNNYPTT (SEQ ID NO: 70) (iv) an HVR-H1 sequence of SYAMS (SEQ ID NO: 38) (v) an HVR-H2 8270468_1 (GHMatters) P94551 .AU.1 5 October 2016 -29- 2014262282 05 Oct 2016 sequence of SISSDGSYTDYADSVKG (SEQ ID NO: 46), and (vi) an HVR-H3 sequence of STDAFDY (SEQ ID NO: 54); and [00128] an anti-cdr2 antibody which binds to HLA-A2 restricted peptide LLEEMFLTV (SEQ ID NO: 19) comprising: (i) an FIVR-L1 sequence of 5 RASQSISSYLN (SEQ ID NO: 56) (ii) an HVR-L2 sequence of GASGLQS (SEQ ID NO: 63) (iii) an HVR-L3 sequence of QQSANAPTT (SEQ ID NO: 71) (iv) an HVR-H1 sequence of SYAMS (SEQ ID NO: 38) (v) an FIVR-FI2 sequence of TINYSGSGTTYADSVKG (SEQ ID NO: 47), and (vi) an HVR-H3 sequence of NAAYFDY (SEQ ID NO: 55). 10 [00129] EXAMPLES - General Procedures

Example 1: Production of biotinylated MHC-peptide complexes [00130] Soluble MFIC class l/peptide complexes were generated by overexpression of the FILA-A2 heavy chain (HC) and β2 microglobulin (β2Μ) 15 as recombinant proteins in E. coli and subsequent in vitro refolding and assembly in the presence of high concentrations of specific peptide (35, 36). To obtain soluble MFIC/peptide complexes the FIC sequence was mutagenized to remove the cytosolic and transmembrane regions. In order to specifically biotinylate refolded, monomeric MFIC/peptide complexes, the FIC 20 was expressed as a fusion protein containing a specific biotinylation site at the C-terminus (37, 38). These short sequences are sufficient for enzymatic in vitro biotinylation of a single lysine residue within this sequence using the biotin protein ligase BirA (39). This procedure was carried out by the MSKCC Tetramer Core Facility. 25

Example 2: Selection of phage on biotinylated MHC-peptide complexes Ex. 2.1: Selection of phage on HLA-A2/EBNA3C (EBNA) complex [00131] The Tomlinson I + J human scFv phage display libraries (40), containing approximately 2.85 x 108 independent scFv clones, were used for 30 selection according to previously published methods (22) with modifications. 8270468_1 (GHMatters) P94551 .AU.1 5 October 2016 -30- 2014262282 05 Oct 2016 7.5 x 1012 Phage, from the combination of both libraries, were first preincubated with streptavidin paramagnetic Dynabeads (30 pi; Dynal, Oslo, Norway) and 150 pg unbiotinylated HLA-A2-YVDPVITSI (SEQ ID NO: ) (irrelevant complex) in 1 ml of PBS to remove any phage which expressed an 5 antibody that binds to streptavidin or the general framework of HLA-A2.

[00132] The dynabeads were subsequently captured using a magnet and the supernatant (phage and irrelevant complex mixture) transferred to a separate tube containing 7.5 pg of biotinylated HLA-A2-LLDFVRFMGV (Epstein-Barr virus EBNA3C-derived) and 7.5 pg of biotinylated FILA-A2-NLVPMVATV 10 (Cytomedullovirus pp65-derived) and incubated at RT for 1 hour. The final mixture (1 ml) was then added to 200 μΙ of Dynabeads (preincubated with 2% Milk and washed with PBS) and the contents were mixed for 15 min. at RT with continuous rotation. The beads were then washed 10 times with PBS/0.1% Tween and 3 times with PBS and the bound phage were eluted 15 from the Dynabeads using 1 mg/ml trypsin in PBS (0.5 ml) for 15 min. at RT.

[00133] The phage were then used to infect TG1 E. coli (growing in log phase) at 37°C in 20 ml of LB for 1 hour. 1012 KM13 helper phage was subsequently added to the mixture, further incubated for an additional 30 minutes, and the cells pelleted using centrifugation (3000 rpm for 10 min.). 20 The resulting cell pellet was resuspended in 200 ml LB + Ampicillin (100 pg/ml) + Kanamycin (50 pg/ml) and incubated overnight at 30°C.

[00134] The following morning, the overnight cultures were centrifuged at 3000 rpm for 15 min. and the supernatant (180 ml) was mixed with polyethylene glycol (PEG) on ice for 1 hour so as to precipitate the amplified 25 phage from the previous round of selection. The PEG/phage mixture was then centrifuged at 3000 rpm for 20 min., and some of the resulting phage pellet used for subsequent rounds of panning while the rest was frozen down in 15% glycerol at -80°C. Subsequent rounds of panning were done using the same protocol as above with an increase in Dynabead washing steps and a 30 decrease in the amount of biotinylated complexes used for selection.

827046$_1 (GHMatters) P94551 AU.1 5 October 201S -31- 2014262282 05 Oct 2016 [00135] After the final round of antibody selection (3rd or 4th round), the eluted phage were used to infect both TG1 and HB2151 E. coir, TG1 cells were cultured overnight as mentioned above while the HB2151 cells were plated on TYE + Ampicillin (100 pg/ml) agar plates. The next morning, individual 5 colonies from the agar plate were picked and used to inoculate individual wells of a 48-well plate containing 400 pi LB + Ampicillin (100 pg/ml)/well.

After incubation for 3-6 hours at 37°C, 200 μΙ of 50% glycerol solution was added to each well and the plates stored at -80°C as monoclonal stock cultures. 10 Ex. 2.2: Selection of phage on HLA-A2-RMFPNAPYL (WT-1) complex [00136] Selection was done similarly to the method above with slight modifications. 3.7 x 1012 Phage from the combination of both libraries, were first preincubated with streptavidin paramagnetic Dynabeads (50 μΙ; Dynal, Oslo, Norway) and 20 μg unbiotinylated HLA-A2-NLVPMVATV (irrelevant 15 complex) in 1 ml of PBS to deplete the streptavidin and HLA-A2 binders. The dynabeads were subsequently captured using a magnet and the supernatant (phage and irrelevant complex mixture) transferred to a separate tube containing 5 pg of biotinylated HLA-A2-RMFPNAPYL (WT1-derived) and incubated at RT for 1 hour. The final mixture (1 ml) was then added to 100 μΙ 20 of Dynabeads (preincubated with 2% Milk and washed with PBS) and the contents were mixed for 30 min. at RT with continuous rotation. The beads were then washed 10 times with PBS/0.1% Tween and 3 times with PBS and the bound phage were eluted from the Dynabeads using 1 mg/ml trypsin in PBS (0.5 ml) for 20 min. at RT. All subsequent steps were performed as 25 above.

Example 3: Expression and purification of soluble scFv from HB2151 [00137] Using the monoclonal glycerol stocks containing individual HB2151 clones, separate 48-well plates containing 400 μΙ LB + Ampicillin (100 pg/ml)/well were inoculated in a replica-plate type format using sterile pipette 30 tips. The 48-well culture plates were subsequently incubated at 37°C until the 8270468_1 (GHMatters) P94551 .AU.1 5Oclober2016 -32- 2014262282 05 Oct 2016 majority of the wells reached an OD600 of 0.4. 200 μΙ LB + Ampicillin (100 pg/ml) + isopropyl-1-thio^-D-galactopyranoside (IPTG; 1mM final concentration) was subsequently added to each well to induce scFv production and the plates were further incubated overnight at 28°C. The next 5 morning, the plates were centrifuged at 3000 rpm for 15 min. and the supernatant used for scFv screening.

[00138] For large scale expression and purification, monoclonal glycerol stocks were used to inoculate 3 ml of Terrific Broth (TB) and incubated at 37°C until an OD600 of 0.8 was reached. Each 3 ml culture was 10 subsequently divided amongst four flasks, each containing 250 ml TB + Ampicillin (100 pg/ml). The cultures were then incubated at 37°C until an OD600 of 0.4-0.5 was reach, after which IPTG was added to a final concentration of 0.5 mM and the cultures incubated overnight at 30°C. The next morning, the overnight cultures were centrifuged at 4000 rpm for 25 min. 15 The supernatant was discarded and the pellets dissolved in 50 ml PBS + 10 mM imidazole. The cell suspensions were passed through a cell homogenizer (5000 pounds per square inch) and the resulting cell lysates were centrifuged at 12,000 rpm for 15 min. The supernatants were then passed over a 0.22 pm filter pre-layered with diatomaceous earth and the resulting filtrates loaded 20 over Vivapure maxiprepMC Nickel affinity columns (Sartorius Stedim Biotech, Aubagne, France) using centrifugation (100 rpm for 5 min.). The columns were then washed 4 times using 10ml PBS + 30 mM imidazole (500 rpm for 3 min.) and the scFvs eluted using 20 ml PBS + 300 mM imidazole (500 rpm for 3 min.). The eluted scFvs were concentrated using 10,000 molecular weight 25 cut-off membrane Vivaspin centrifuge tubes at 3000 rpm for 30 min. (Sartorius Stedim Biotech) and dialyzed back into regular PBS. The final scFv products were subsequently stored at -80°C.

Example 4: Construction of scFv-Fc fusion protein and expression in DG44 CHO cells 30 [00139] Using a proprietary antibody expression vector (referred to herein as

IgG Vector), similar to that of pFUSE—hlgG1-Fc1 (InvivoGen; San Diego, 8270468_1 (GHMatters) P94551 .AU.1 5Oclober2016 -33- 2014262282 05 Oct 2016

California), the construct was first modified to contain the CH2, and CH3 domains of a human IgGi (scFv-Fc Vector). Subsequently, the EBNA Clone 315 and WT1 Clone 45 scFv sequences were PCR amplified to contain the required Nhel and Apal restriction sites which would be compatible with the 5 scFv-Fc vector. The resulting scFv PCR products and antibody expression plasmid were digested using the above enzymes (Nhel at 37°C for 2 hours and Apal at 25°C for 2 hours) and then ligated together. The ligation products were then transformed into E.coli, plated on TYE + Amplicilin (100 pg/ml), colonies were picked and their plasmids sequenced at the MSKCC 10 Sequencing Core Facility. Once the sequences were validated to have the correct scFv sequences upstream of the human IgGi CH2 and CH3 domains, the DNA (5-6 pg was electroporated (Amaxa Nucleofactor; Lonza,

Switzerland) into 5 x 106 DG44 Chinese Hamster Ovary (CHO) Cells (Invitrogen) using Program U-030 and 100 pi Solution V. The cells were then 15 cultured in OptiCHO media (Invitrogen) containing G418 (500 pg/ml; added 7 days post-electroporation) at a cell density of 1-5 x 106 DG44 per ml of media. The cells were then expanded to approximately 700 ml of culture media, which was centrifuged to remove the cells and supernatant used for antibody purification. 20 Example 5: Expression and purification of soluble scFv-Fc fusion protein [00140] DG44 supernatant containing the soluble scFv-Fc fusion protein was purified using the KappaSelect affinity chromatograph medium (GE Healthcare). First, 1.5 ml of KappaSelect resin was loaded onto a column and activated with 20 ml of PBS. The supernatant was loaded onto the column 25 using a peristaltic pump at a flow rate of approximately 1 ml/min. The column was subsequently washed using 40 ml of PBS until the flow-thru registered an OD280 of less than 0.05. The scFv-Fc fusion protein was then eluted from the resin using 10 ml citrate buffer (pH 2.0) and directly into 10ml of 1 M Tris for neutralization. The eluted scFv-Fc was subsequently concentrated using a 30 50,000 MWCO Vivaspin centrifuge tube (Sartorius Stedim) and tested for its

ability to bind to recombinant antigen using ELISA and the Biacore T100 (GE 8270468_1 (GHMatters) P94551 .AU.1 50ctober2016 -34- 2014262282 05 Oct 2016

Healthcare) as well as natively presented peptide on the surface of T2 cells using flow cytometry.

Example 6: Monoclonal ELISA with bacterial phage clones and purified scFv and scFv-Fc 5 [00141] Vinyl flat bottom microtiter plates (Thermo Fisher) were used for ELISA assays. Plates were initially coated overnight at 4°C with BSA-biotin (10 pg/ml; 50 μΙ/well). The next morning, the contents were discarded and the plates incubated at RT with streptavidin (10 pg/ml; 50 pl/well) for 1 hour. The contents were discarded and the plates incubated with recombinant 10 biotinylated HLA-A2-peptide complexes (5 pg/ml; 50 pl/well) at RT for 1 hour. The plates were then incubated with 2% Milk (150 pl/well) at RT for 1 hour. After blocking, the plates were washed 2 times with PBS and then incubated with bacterial supernatant from their respective HB2151 culture plate wells, purified scFv, or purified scFv-Fc at RT for 1 hour. The contents were 15 discarded, the plates washed 5 times with PBS, and then incubated at RT for 1 hour with either a mouse-anti-myc tag antibody (Clone 9E10; Sigma Aldrich. 0.5 pg/ml; 100 pl/well in 0.5% Milk) to detect the scFv or a goat-anti-human-HRP (Jackson Immunoresearch Laboratories. 0.5 pg/ml; 100 pl/well in 0.5% Milk) to detect the scFv-Fc. The contents were discarded, the plates washed 20 5 times with PBS, and those receiving the scFv were further incubated with a goat-anti-mouse-HRP (Jackson Immunoresearch Laboratories. 0.5 pg/ml; 100 μΙ/well in 0.5% Milk) at RT for 1 hour while the plates receiving the scFv-Fc were developed using o-phenylenediamine (OPD) buffer (150 pl/well), which was made by combining 20 mg of OPD tablets in 40 ml of citrate phosphate 25 buffer with 40 pi 30% hydrogen peroxide. The color reaction was stopped by adding 30 pi of 5N sulfuric acid to each well and the plates read using the Dynex MRX ELISA plate reader at 490 nm. Lastly, the contents of the scFv plates were discarded, the plates washed 5 times with PBS, and developed according to the method above. 30 Example 7: Cell lines and peptides 8270468_1 (GHMatters) PS4551 .AU.1 5October2016 -35- 2014262282 05 Oct 2016 [00142] Tap-deficient HLA-A2+ T2 cells, 6268A, GKO (both HLA-A2'), DIMT and JG19 (both HLA-A2+) B-cell lymphoblastic cell lines (BLCLs) were used for antigen presentation studies. Cells were normally cultured in RPMI 1640 + 10% Fetal Bovine Serum (FBS). For antigen presentation, T2 cells were 5 harvested and transferred to serum-free IMDM + 10 pg/ml p2-microglobulin (β2Μ). The T2 cells would then be incubated with 20 μΜ or less of either LLDFVRFMGV-peptide (derived from EBNA3C) or any number of irrelavent peptides at 37°C for 5 hours. Studies with BLCLs were done in the same manner as with T2 cells with the occlusion of β2Μ in the media. Pulse-Chase 10 experiments with DIMT BLCLs were done by first pulsing the BLCLs in serum-free IMDM with 20 μΜ LLDFVRFMGV for 5 hours at 37°C. The cells were then washed with fresh RPMI 1640 + 10% FBS, transferred back into this culture medium and cultured further at 37°C for 5 and 24 hours, followed by flow cytometric analysis at each time point using EBNA Clone 315 scFv. 15 Example 8: Binding Kinetics Analysis

[00143] Kinetic measurements were performed by surface plasmon resonance using the BIAcore T100 (GE Biosciences). Briefly, the first two flow cells of a CM5 chip (GE Biosciences) were activated using the standard amine coupling reagents in FIBS-EP running buffer (0.01 M FIEPES, 0.15 M 20 NaCI, 3 mM EDTA, 0.005% Tween 20) with flow cell 2 immobilized with the purified EBNA Clone 315 scFv-Fc fusion protein using 10 mM Acetate (pH 5). Subsequently, the target HLA-A2-peptide monomer (222 nM-13.875 nM) was injected over both the 1st (reference) and 2nd flow cells at 20 pl/min. for 120 sec., followed by the addition of running buffer for an extra 180 sec. Kinetics 25 values were determined using the BIAcore T100 Evaluation Software 2.0 and 1:1 binding model (local Rmax).

Example 9: Flow Cytometry [00144] Peptide-pulsed T2 cells and BLCLs were transferred to plastic polystyrene round-bottom tubes (Becton Dickinson Labware) and washed with 30 PBS. The cells were subsequently incubated with 5 pg of either targeted or 8270468_1 (GHMatters) P94551 .AU.1 5 October 2016 -36- 2014262282 05 Oct 2016 non-specific purified scFv or scFv-Fc on ice for 40 min. The cells were washed with PBS and then incubated with 1 pg of biotinylated mouse-anti-myc antibody (Clone 9E10; Sigma Aldrich) or biotinylated mouse-anti-human IgG Fc-specific (Jackson Immunoresearch Laboratories) on ice for 30 min. 5 The cells were washed with PBS and then incubated with streptavidin-PE (BD Biosciences). Lastly, the cells were washed once more with PBS and analyzed on the BD FACS Calibur.

[00145] For CD107a cytotoxicity assays, transduced-NK92MI cells and target T2 cells were cocultured in a 1:1 effector:target (E:T) ratio (2.5-5.0 x 105 cells 10 each) in 200 pi complete Alpha Essential medium (12.5% horse serum and 12.5% FBS) (Invitrogen) + 10-15 μΙ anti-CD107a-PE at 37°C for 5 Hours. The cell mixture was then washed with PBS and analyzed on the BD FACS Caliber.

[00146] For FACS sorting experiments, retrovirally transduced NK92MI cells 15 were sorted based on GFP intensity using the BD Aria Flow Cytometer under the guidence of the MSKCC Flow Cytometry facility.

Example 10: Quantitation of HLA-A2-LLDFVRFMGV complexes on peptide-pulsed T2 cells [00147] For MHC-peptide complex quantitation, the EBNA Clone 315 scFv-Fc 20 was first directly conjugated to Alexa Fluor 647 using the APEX Alexa Fluor 647 Antibody Labeling Kit (Invitrogen). The kit yields about 10-20 pg of labeled antibody.

[00148] For quantitation, the Quantum Simply Cellular anti-Human IgG kit was used (Bangs Laboratories) along with the technical assistance of Hong- 25 fen Guo in our laboratory. Briefly, the kit is comprised of five microsphere populations; one blank and four labeled with increasing amounts of antihuman IgG. The beads and the peptide pulsed T2 cells (37°C for 5 hours) were then labeled with the same fluorescently conjugated EBNA Clone 315 scFv-Fc on ice for 30 minutes. The cells were then washed with PBS and 30 analyzed on the BD FACS Calibur along with the labeled beads. The Excel- 8270468_1 (GHMatters) P94551 .AU.1 5 October 2016 -37- 2014262282 05 Oct 2016 based QuickCal analysis template that’s provided with each kit aids in correlating fluorescence intensity with antigen density on the T2 cells. Each of the 4 data points are the average of duplicates.

Example 11: Construction of the WT-1 clone 45 chimeric antigen receptor 5 [00149] The original chimeric antigen receptor was obtained from Dr. Dario

Campana from St. Jude Children’s Hospital and previously described (41).

For future compatibility purposes, a 5θΡν-Οϋ3ζ-4-1 BB DNA construct (similar to that seen in the original chimeric immune receptor, with EcoRI and Xhol flanking the 5’ and 3’ ends) was purchased (pUC57 vector from Genescript; 10 Piscataway, NJ) and contained an irrelevant scFv flanked by Sfil and Notl.

The plasmid containing the EBNA Clone 315 scFv sequence (plT2 vector from the Tomlinson library) was purified (Qiagen miniprep DNA isolation kit) from overnight culture of the bacterial stock in LB + Amplicilin (100 pg/ml).

The scFv sequence was excised from the plT2 vector using Sfil (50°C for 2 15 hours) and Notl (37°C for 2 hours) inserted into the purchased and predigested (Sfil and Notl) pUC57 vector. After ligation, the productwas transformed into NEB 5-alpha competent E. coli (New England Biolabs), the cells plated on TYE + Amplicilin (100 pg/ml), colonies were picked and cultured in LB + Amplicilin (100 pg/ml), their plasmids purified and the product 20 sizes were verified by gel electrophoresis. Plasmids which were found to have the correct ligation products were subsequently excised from the pUC57 vector using EcoRI (37°C for 2 hours) and Xhol (37°C for 2 hours) and used for insertion into the vector provided to us by the Campana laboratory. The ligation products were then transformed into E.coli as above, plated on TYE + 25 Amplicilin (100 pg/ml), colonies were picked and their plasmids sequenced using the reverse primer 788A (5'-CCCTTGAACCTCCTCGTTCGACC-3') (SEQ ID NO: 72) at the MSKCC Sequencing Core Facility. Once the sequences were validated, the DNA was packaged into retrovirus and used to infect NK92MI cells. 30

Example 12: Construction of the EBNA Clone 315 chimeric antigen receptor 8270468_1 (GHMatters) P94551 .AU.1 5Oclober2016 -38- 2014262282 05 Oct 2016 [00150] Due to compatibility issues, the pUC57 5θΡν-Οϋ3ζ-4-1 BB DNA construct purchased from Genescript and mentioned above was used to replace the WT1 Clone 45 scFv with the EBNA Clone 315 scFv. First, the plasmid containing the EBNA Clone 315 scFv sequence (plT2 vector from the 5 Tomlinson library) was purified (Qiagen miniprep DNA isolation kit) from overnight culture of the bacterial stock in LB + Amplicilin (100 pg/ml). The scFv sequence was excised from the plT2 vector using Sfil (50°C for 2 hours) and Notl (37°C for 2 hours) and ligated to the predigested (Sfil and Notl) pUC57 vector. After ligation, the product was transformed into E. coli, 10 colonies were picked, cultured overnight, their plasmids purified and the product sizes verified by gel electrophoresis. Plasmids which were found to have the correct ligation products were subsequently excised from the pUC57 vector using EcoRI (37°C for 2 hours) and Xhol (37°C for 1 minute). Due to the presence of a Xhol site inside of the EBNA Clone 315 scFv sequence, the 15 DNA was partially digested with Xhol and then completely digested using EcoRI. This allowed for the isolation of the correct DNA fragment which kept the integrity of the scFv sequence while removing the entire CAR sequence from the pUC57 vector. After insertion into the vector provided to us by the Campana laboratory, the ligation products were then transformed into E.coli 20 as above, plated on TYE + Amplicilin (100 pg/ml), colonies were picked and their plasmids sequenced using the reverse primer 788A (5'-CCCTTGAACCTCCTCGTTCGACC-3') (SEQ ID NO: 72) at the MSKCC Sequencing Core Facility. Once the sequences were validated, the DNA was packaged into retrovirus and used to infect NK92MI cells. 25 Example 13: Retroviral production, DNA packaging, and infection of NK92MI cells [00151] To produce CAR-containing retrovirus, the following procedure was employed which used a 293T-based retroviral production cell line (GP2). Briefly, 7 pg of CAR DNA was combined with 3.5 pg of PCLAmpho helper 30 construct and 3.5 pg pVSVg in 1 ml of serum-free DMEM. This mixture was then combined with 1 ml serum-free DMEM containing 36 pi of Lipofectamine 2000 (Invitrogen) and incubated at RT for 20 min. Afterwards, the DNA-8270468_1 (GHMatters) P94551 .AU.1 50ctober2016 -39- 2014262282 05 Oct 2016

Lipofectamine complex (2 ml) was mixed with GP2 cells (3-5 x 106) in 10 ml of DMEM + 10% FBS and cultured at 37°C for 72 hours. Subsequently, the supernatant (12 ml) was depleted of GP2 cells during recovery and incubated with 3 ml Lenti-X Concentrator solution (Clontech) at 4°C for 12-16 hours. 5 Afterwards, the solution was centrifuged at 3000 rpm for 15 min., the supernatant discarded, and the pellet dissolved in 1 ml complete Alpha Essential medium containing 5 x 105 NK92MI cells. The cells were then incubated for 72 hours and checked by flow cytometry for CAR expression via GFP (the CAR gene is expressed under a CMV promoter which is followed by 10 IRES-GFP).

Exam pie 14: Construction of scFv-Fc fusion protein and expression in DG44 CHO cells [00152] Using a proprietary antibody expression vector similar to that of pFUSE—hlgG1-Fc1 (Invivogen; San Diego, California), the Clone 315 scFv 15 sequence was first PCR amplified to contain the required Nhel and Apal restriction sites. The resulting PCR product and expression plasmid were digested using the above enzymes (Nhel at 37°C for 2 hours and Apal at 25°C for 2 hours) and ligated together. The ligation products were then transformed into E.coli, plated on TYE + Amplicilin (100 pg/ml), colonies were 20 picked and their plasmids sequenced at the MSKCC Sequencing Core Facility. Once the sequences were validated to have the Clone 315 scFv sequence upstream of the human IgGi CH2 and CH3 domains, the DNA was electroporated (Amaxa Nucleofactor; Lonza, Switzerland) into 5x 106 DG44 Chinese Hamster Ovary (CHO) Cells (Invitrogen) using Program U-030 and 25 100 pi Solution V. The cells were then cultured in OptiCHO media (Invitrogen) containing G418 (500 pg/ml; added 7 days post-electroporation) at a cell density of 1-5 x 106 DG44 per ml of media.

Example 15: Retroviral production, DNA packaging, and infection of N«92Ml cells 8270468J (GHMatters) P94551.AU.1 5October2016 -40- 2014262282 05 Oct 2016 [00153] To produce CAR-containing retrovirus, the following procedure was employed which used a 293T-based retroviral production cell line (GP2). Briefly, 7 pg of CAR DNA was combined with 3.5 pg of PCLAmpho helper construct and 3.5 pg pVSVg in 1 ml of serum-free DMEM. This mixture was 5 then combined with 1 ml serum-free DMEM containing 36 pi of Lipofectamine 2000 (Invitrogen) and incubated at RT for 20 min. Afterwards, the DNA-Lipofectamine complex (2 ml) was mixed with GP2 cells (3-5 x 106) in 10 ml of DMEM + 10% FBS and cultured at 37°C for 72 hours. Subsequently, the supernatant (12 ml) was depleted of GP2 cells during recovery and incubated 10 with 3 ml Lenti-X Concentrator solution (Clontech) at 4°C for 12-16 hours. Afterwards, the solution was centrifuged at 3000 rpm for 15 min., the supernatant discarded, and the pellet dissolved in 1 ml complete Alpha Essential medium containing 5 x 105 NK92MI cells. The cells were then incubated for 72 hours and checked by flow cytometry for CAR expression via 15 GFP (the CAR gene is expressed under a CMV promoter which is followed by IRES-GFP).

Example 16: 51 Cr Release Cytotoxicity Assay [00154] The capacity of CAR equipped NK92MI cells to lyse BLCLs was evaluated using a 51 Chromium release assay. Briefly, peptide pulsed or 20 unpulsed 51Cr-labeled BLCLs were plated in round-bottom 96-well plates (5 X 103 cells/well) in RPM I 1640 with 10% FBS. Subsequently, CAR equipped NK92MI cells were added to the BLCL containing wells at different effector (E)/target (T) ratios and incubated for 4 hours at 37°C, after which the cultures were depleted of cells and 51Cr-release was measured in the supernatants. 25 All E:T ratios were done in triplicate, with the average plotted on the graphs. % 51 Cr Release was determined using the following formula: ((Sample Release - Spontaneous Release) / (Total Release - Spontaneous Release)) X 100.

Example 17: Affinity selection of phage on virally-derived recombinant HLA-30 A2-peptide complexes 8270468_1 (GHMatters) P94551 .AU.1 5October2016 -41- 2014262282 05 Oct 2016

[00155] Biotinylated and non-biotinylated recombinant HLA-A2-peptide complexes presenting various different peptides previously shown to bind to HLA-A2 were obtained from the MSKCC Tetramer Core Facility. For selection purposes, the Tomlinson I and J phage display libraries were combined and 5 first preincubated with non-biotinylated, irrelevant FILA-A2-YVDPVITSI complex along with streptavidin paramagnetic beads so that any phage which expresses an antibody that may bind to the general framework of FILA-A2, or the streptavidin beads themselves, are eventually discarded during the washing steps. Subsequently, the contents (phage and irrelevant complex) 10 were incubated with biotinylated FILA-A2-LLDFVRFMGV (EBNA3C) and biotinylated FILA-A2-NLVPMVATV (pp65) simultaneously in equimolar ratios and captured using streptavidin paramagnetic beads. Once the beads were bound to the biotinylated complexes, the beads were washed with PBS containing Tween 20 and the bound phage were eluted from the beads using 15 trypsin. After two additional rounds of selection, the recovered phage were used to infect HB2151 E. coli and plated on ampicillin-containing agar. The next morning, individual colonies were picked, cultured overnight in 48-well culture plates, and their supernatants tested for the presence of scFv on 96-well ELISA plates pre-coated with recombinant FILA-A2-peptide complexes. 20 [00156] The first three rounds of selection resulted in a 55-fold increase in phage recovery, based on output/input ratio, and scFvs which only bound to the FILA-A2-EBNA3C complex. Phage display selection results on recombinant HLA-A2-LLDFVRFMGV and HLA-A2-NLVPMVATV complexes are shown in Table 8. 25 8270468_1 (GHMatters) P94551.AU.1 5Odober2016 -42- 2014262282 05 Oct 2016

Table 8 Round 1* Round 2* Round 3* Round 4** Input 7.5 x 1012 4.9 x 1012 2.4 x 1012 1.8 x 1012 Output 4.7 x 106 6 x 106 8.4 x 107 1.25x10® Output/lnput 6.3 x 10 7 1.22 x 10'6 3.5x1 O'5 6.9x1 O'5 Fold Enrichment (From Rd 1) — 2 55.5 109.5 HLA-A2- EBNA3C Peptide-Specific Clones*** — — 40/48 (83%) 37/48 (77%) HLA-A2-pp65 — — 0/48 (0%) 0/48 (0%) Peptide-Specific Clones*** * Rd 1-3: Panning against Biotinylated-HLA-A2-pp65 Peptide + Biotinylated-HLA-A2-EBNA Peptide ** Rd 4: Panning against Biotinylated HLA-A2-EBNA3C Peptide Only *** Relative signal at least 2-fold greater than background.

[00157] These results were somewhat surprising since both of the peptides on HLA-A2 were derived from viral-related proteins, which are not seen in the 5 human protein repertoire. To confirm these findings, an additional round of selection was undertaken on just the HLA-A2-EBNA3C complex alone which resulted in a further amplification of recovered phage (109-fold) and a similar percentage of clones which bound to the HLA-A2-EBNA3C complex (83% positive after Round 3 and 77% after Round 4). 10 [00158] Bacterial supernatant from individual clones after 3 rounds of phage selection were tested for binding to recombinant, biotinylated-HLA-A2-peptide complexes on vinyl microtiter plates. While several clones resulted in crossreactivity to more than just the targeted HLA-A2-LLDFVRFMGV complex (Clones 335 and 345), Clones 315 and 327 were found to have the desired 15 specificity. 8270468_1 (GHMallers) P94551 AU.1 5 October 201$ -43- 2014262282 05 Oct 2016 [00159] Purified EBNA Clone 315 scFv was retested against a similar panel of recombinant, biotinylated HLA-A2-peptide complexes. Purified EBNA Clone 315 scFv maintained its specificity over a panel of FILA-A2-peptide complex in addition to its inability to bind to the native peptide by itself. The 5 anti-FILA-A2 antibody BB7.2 was included to demonstrate that all FILA-A2-peptide complexes are adherent and presented properly on the plate.

[00160] During the screening processes, therefore, several different scFv were found to bind to the targeted FILA-A2-EBNA3C complex, however only a few scFv sequences resulted in absolute specificity and did not bind to FILA- 10 A2-peptide complexes of different origins (Figure 1 A). Of those clones which were tested, EBNA Clones 315 and 327 had the same peptide sequence and were further characterized. After scFv purification, a subsequent validation ELISA demonstrates that EBNA Clone 315 maintained its specificity towards the targeted FILA-A2-EBNA complex, in addition to failing to bind to the 15 LLDFVRFMGV peptide by itself (Figure 1B). These initial binding assays demonstrate the TCR-like binding ability of this antibody.

Example 18: Affinity selection of phage on WT1 -derived recombinant HLA-A2-peptide complex [00161] Antibody selection using phage against biotinylated FILA-A2- 20 RMFPNAPYL (WT1 -derived) was done in a similar manner to that which was described above. Briefly, the Tomlinson I and J phage display libraries were first combined and preincubated with non-biotinylated, irrelevant FILA-A2-NLVPMVATV complex and streptavidin paramagnetic beads. Subsequently, the contents (phage and irrelevant complex) were incubated with biotinylated 25 FILA-A2-RMFPNAPYL and captured using fresh streptavidin paramagnetic beads. Once bound to the biotinylated complex, the beads were washed with PBS containing Tween 20 and the bound phage were eluted from the beads using trypsin. After two additional rounds of selection, the recovered phage were used to infect HB2151 E. coli and plated on ampicillin-containing agar. 30 The next morning, individual colonies were picked, cultured overnight in 48-well culture plates, and their supernatants tested for the presence of scFv on 8270468J (GHMatters) P94551.AU.1 50ctober2016 -44- 2014262282 05 Oct 2016 96-well ELISA plates pre-coated with recombinant HLA-A2-peptide complexes.

[00162] The first three rounds of selection resulted in a 90-fold enrichment in phage when comparing the output/input ratios. Phage display selection 5 results on recombinant HLA-A2-RMFPNAPYL complex are shown in Table 10.

Table 9

Round 1* Round 2** Round 3 Input 3.7 x 1012 5.6 x 1011 1.55 x 10' Output 4.0 x 10s 3.2 x 10s 1.52 x 10‘ Output/input 1.08 x 10'6 5.7 x 10 6 9.8x1 O’5 Fold Enrichment (From Rd 1) — 5.3 90.7 HLA-A2- 3/48 RMFPNAPYL-

Specific Clones*** * Rd 1: Panning against 5 pg Complex. ** Rd 2-3: Panning against 2.5 pg Complex. *** Relative signal at least 3-fold greater than background (Irrelevant HLA-A2-Complex).

[00163] Bacterial supernatant from three individual clones after three rounds 10 of phage selection were tested for binding to recombinant, biotinylated-HLA-A2-peptide complexes on vinyl microtiter plates. All three clones which were tested had the necessary specificity to only recognize the HLA-A2-RMFPNAPYL complex. It was discovered that all three clones had the same DNA sequence. Purified WT1 Clone 45 scFv was retested against a similar 15 panel of recombinant, biotinylated HLA-A2-peptide complexes. Purified WT1 8270468_1 (GHMatters) P94551 .AU.1 5Oclober2016 -45- 2014262282 05 Oct 2016

Clone 45 scFv maintained its specificity over a panel of HLA-A2-peptide complex in addition to its inability to bind to the native peptide outside of the context of MHC. The anti-HLA-A2 antibody BB7.2 was included to demonstrate that all HLA-A2-peptide complexes are adherent and presented 5 properly on the plate.

[00164] After screening 48 clones for binding to the specific HLA-A2-RMFPNAPYL complex, therefore, three clones were found to bind specifically to their targeted complex but failed to bind to complexes which displayed an irrelevant peptide (Figure 2A). Of the clones which were tested, all of them 10 were found to have the same peptide sequence and WT1 Clone 45 was chosen for further characterization. After scFv purification, a subsequent validation ELISA demonstrates that WT1 Clone 45 maintained its specificity towards the targeted HLA-A2-WT1 complex, in addition to failing to bind to the RMFPNAPYL peptide by itself (Figure 2B). These initial binding assays 15 demonstrate the TCR-like binding ability of this antibody.

Example 19: Binding and specificity studies with purified EBNA Clone 315 and WT1 Clone 45 scFvs on peptide-pulsed T2 cells [00165] To demonstrate that the isolated EBNA Clone 315 and WT1 Clone 45 scFvs are able to recognize and bind to their native complexes on the surface

20 of peptide-pulsed antigen presenting cells (APCs), the TAP-deficient T2 cell line was used. T2 cells were first incubated for 5 hours at 37°C with either LLDFVRFMGV (EBNA3C-derived), RMFPNAPYL (WT1-derived) or irrelevant peptide KLQCVDLHV in serum-free medium containing β2Μ. The cells were subsequently washed and stained with the purified WT1 Clone 45, EBNA 25 Clone 315 or an irrelevant scFv. In addition, peptide pulsed and unpulsed T2 cells were also incubated with an anti-HLA-A2-FITC (BB7.2) antibody. This BB7.2 staining control was included due to previous studies which demonstrate that if a peptide is able to bind HLA-A2 on the T2 cell surface, the HLA-A2 molecule is stabilized, and the stabilization can be visualized by 30 an increase in fluorescence intensity (81). As shown in Figure 3A, T2 cells which have been pulsed with either the LLDFVRFMGV or KLQCVDLHV 8270468_1 (GHMatters) P94551 .AU.1 5 October 2016 -46- 2014262282 05 Oct 2016 peptides resulted in a fluorescence shift, consistent with their binding to the HLA-A2 pocket. However, EBNA Clone 315 was only able to stain T2 cells pulsed with its specific target peptide LLDFVRFMGV and not an irrelevant peptide (Figure 3B). Similar results were obtained when T2 cells were pulsed 5 with either the RMFPNAPYL or LLDFVRFMGV peptides and stained with WT1 Clone 45 scFv. While both peptides were able to stabilize the HLA-A2 molecule (Fig. 4A), WT1 Clone 45 scFv was only able to detect the T2 cells pulsed with the RMFPNAPYL peptide (Fig. 4B). This further validates their utility in detecting the native complex on the surface of cells. 10 [00166] Next, the detection sensitivity of the EBNA Clone 315 scFv using flow cytometry was evaluated in order to correlate sensitivity with antigen density using flow cytometric quantitative beads. Briefly, TAP-deficient T2 cells were pulsed with (solid, unfilled lines) or without (dashed, unfilled lines) LLDFVRFMGV (Figure 3A, top, left panel) or KLQCVDLHV peptides (Figure 15 3A top, right panel) at 20 μΜ in serum-free IMDM media at 37°C for 5 hours. The cells were then stained with a mouse-anti-human HLA-A2-FITC conjugated antibody (unfilled lines) or a control mouse IgG-i-FITC conjugated antibody (filled lines) and analyzed on the FACS machine. Peptide-pulsed T2 cells from A were stained with EBNA Clone 315 scFv (unfilled lines) or a 20 control scFv (filled lines) (Figure 3B). Only T2 cells which had been pulsed with the LLDFVRFMGV peptide (left panel), but not ones which had been pulsed with KLQCVDLHV (right panel), could be stained by the EBNA Clone 315 scFv (Figure 3B). T2 cells were incubated with decreasing concentrations of the LLDFVRFMGV peptide and subsequently stained with 25 EBNA Clone 315 scFv as above (Figure 3C). Based on geometric mean fluorescence (control scFv background subtracted), the lower limit of detection corresponds with 78 nM of peptide used for pulsing.

[00167] By titrating down the amount of peptide used for incubation with the 30 T2 cells, it was determined that concentrations as low as 78 nM were still able to produce a fluorescence signal above background when stained with EBNA Clone 315 scFv (Figure 3C). With decreasing concentrations of peptide used 8270468J (GHMatters) P94551.AU.1 50ctober2016 -47- 2014262282 05 Oct 2016 for loading, there was a corresponding reduction in overall HLA-A2 intensity (data not shown) as one would expect.

[00168] Similarly, Figure 4 shows that WT1 Clone 45 can recognize HLA-A2-5 RMFPNAPYL on peptide-pulsed T2 cells. TAP-deficient T2 cells were pulsed

with (solid, unfilled lines) or without (dashed, unfilled lines) RMFPNAPYL (left panel) or LLDFVRFMGV peptides (right panel) at 40 μΜ in serum-free IMDM media at 37°C for 5 hours. The cells were then stained with a mouse-anti-human HLA-A2-FITC conjugated antibody (unfilled lines) or a control mouse 10 lgG1-FITC conjugated antibody (filled lines) and analyzed on the FACS machine (Figure 4A). Peptide-pulsed T2 cells from A were stained with WT1 Clone 45 (unfilled lines) or a control scFv (filled lines). Only T2 cells which had been pulsed with the RMFPNAPYL peptide (left panel), but not ones which had been pulsed with LLDFVRFMGV (right panel), could be stained by 15 the WT1 Clone 45 (Figure 4B).

Example 20: Demonstrating HLA restriction of the LLDFVRFMGV peptide and EBNA Clone 315 using peptide-pulsed BLCLs [00169] The expression of these peptides on BLCLs, especially since BLCLs are used routinely as APCs (82), was examined. Two BLCL lines were used, 20 one HLA-A2+ (DIMT) and one HLA-A2' (6268A) (Fig. 5A). The BLCLs were incubated in serum-free IMDM media for 5 hours at 37°C with either the specific LLDFVRFMGV or irrelevant KLQCVDLHV peptides. When incubated with the specific peptide, only the HLA-A2+ DIMT BLCL could be stained by EBNA Clone 315 (Fig. 5B). Similarly to results seen with T2 cells, DIMT cells 25 loaded with the irrelevant peptide, or 6268A loaded with the specific/irrelevant peptide, could not be stained with EBNA Clone 315. It is interesting to note that without peptide pulsing we were unsuccessful at staining DIMT. While our staining approach has been optimized to detect low levels of antigen through signal amplification involving secondary and tertiary reagents to 30 detect the scFv, the amount of peptide that the cell naturally presents seems to be below our level of detection. 8270468_1 (GHMatters) P94551 .AU.1 5Odober2016 -48- 2014262282 05 Oct 2016 [00170] Subsequently, in an attempt to study the duration of peptide presentation on HLA-A2, a pulse-chase experiment was set up to monitor the levels of the HLA-A2-EBNA3C complex on DIMT cells over time. Initially, DIMT cells were incubated in serum-free IMDM media for 5 hours at 37°C with 5 the LLDFVRFMGV peptide. Afterwards, the cells were washed twice with RPMI + 10% FBS and further cultured in this media for an additional 5 hours and 24 hours. At each of these three time points, cells were harvested and stained with either the purified EBNA Clone 315 scFv or an irrelevant scFv. The results show that after pulsing the HLA-A2-EBNA3C complex could easily 10 be detected on the cell surface (Fig. 5C). Interestingly, even after the cells were transferred to fresh media and cultured for an additional 5 and 24 hours, the MFIC-peptide complex could still be detected, signifying that peptide-pulsed BLCLs are able to hold onto and present antigen for at least a day after the peptide had been removed from the media. This data further 15 supports the use of autologous BLCLs in the generation of antigen specific T cells and the utility of TCE-specific antibodies like EBNA Clone 315 in precise visualization of TCE expression on APCs or target cells.

Example 21: Construction of EBNA Clone 315 and WT1 Clone 45 scFv-Fc fusion proteins 20 [00171] Initially, the scFv sequences were made compatible for cloning into a scFv-Fc expression vector by using PCR to add the desired restriction enzyme sites (Nhel and Apal) to either side of the EBNA Clone 315 and WT1 Clone 45 scFv sequences. The PCR reaction was done on the Tomlinson library vector which contained the WT1 Clone 45 and EBNA Clone 315 scFv 25 sequences (Fig. 7A). After subsequent digestion using Nhel and Apal, the digested PCR products were removed from a 1% agarose gel and purified (Fig. 6B).

[00172] With regards to cloning and expression of the scFv-Fc fusion proteins, a proprietary vector obtained from Eureka therapeutics(lgG Vector) 30 was used. The first constant heavy chain (CHi) was removed from this vector, something which is typically done when generating Fc fusion proteins 8270468_1 (GHMatters) P94551 AU.1 5 October 2016 -49- 2014262282 05 Oct 2016 (83). Once generated, the vector was digested with Nhel and Apal and then ligated to the predigested PCR products from Figure 7B. The ligated products yielded a vector which expressed the EBNA Clone 315 or WT1 Clone 45 scFv genes in tandem to the CH2, 3 domains of a human lgG1 under a single CMV 5 promoter (scFv-Fc Vector; Figure 7B). After further validation using DNA sequencing, the two fusion constructs were linearized using Hindlll and ran on a 1% agarose gel. Digestion with Hindlll also allowed us to block the expression of the light chain that is still present in the vector, which for all intensive purposes was undesired. As expected, both digested plasmids ran 10 at the anticipated size (~11,000 bp) based on their location relative to the lambda Hindlll marker. Each linearized plasmid was subsequently introduced into DG44 cells and cultured in OptiCFlO media as described in Example 10 above.

Example 22. Binding kinetics and sensitivity of EBNA Clone 315 scFv-Fc on 15 recombinant HLA-A2-peptide complex and peptide-pulsed T2 cells [00173] To further understand the affinity of the interaction between EBNA Clone 315 and the HLA-A2-EBNA3C complex, surface plasmon resonance was used to determine the binding kinetics between these two proteins. First, the EBNA Clone 315 scFv-Fc was purified and its binding ability was tested 20 using ELISA (Fig. 8A) along with flow cytometry via peptide-pulsed T2 cells at varying concentrations (Fig. 8B and C). These initial studies demonstrate that the antibody maintains its binding characteristics when expressed as a fusion protein In addition, it is important to note that the flow cytometric sensitivity of the scFv and scFv-Fc were very comparable (200nM - 20nM), further 25 highlighting the utility of the scFv as a monomeric binding fragment.

[00174] Next, using the Biacore T100 (GE Healthcare), a CM5 chip (flow cells 1 and 2) was initially activated for amine coupling based on manufacturer recommendation. The purified EBNA Clone 315 scFv-Fc was subsequently immobilized onto the second flow cell and the purified HLA-A2-EBNA3C 30 complex passed over both flow cells as part of the soluble phase. After background subtraction (signal from flow cell 2 minus that of flow cell 1), the 8270468_1 (GHMatters) P94551 .AU.1 5Oclober2016 -50- 2014262282 05 Oct 2016 association rate (kon) and dissociation rate (k0ff) were determined (2.361 x 105 M'Vand 6.891 x 10'2s'1, respectively), resulting in an overall Kd (k0ff/k0n) of 291 nM using a 1:1 binding model (Fig. 9); these kinetic rates were very similar to previously isolated Fabs against different MHC-peptide complexes 5 (22, 31). Relative to published TCR:MHC Class l-peptide Kd measurements, which typically range in the neighborhood of 2-50 μΜ (84), our scFv:MHC Class l-peptide interaction seems to be at best 150-fold stronger, with the most significant improvement attributed to a slower k0ff. Previous studies which support an affinity-based T cell activation model argue that a greater 10 overall affinity or slower dissociation rate leads to higher interferon-gamma release and target cell lysis (85, 86).

[00175] Lastly, in an attempt to quantify the amount of HLA-A2-LLDFVRFMGV complex on the surface of peptide-pulsed T2 cells, we decided to use flow cytometric quantitation beads. This data will also be useful in 15 determining the detection threshold of EBNA Clone 315 scFv and scFv-Fc. First, purified EBNA Clone 315 scFv-Fc was conjugated to a fluorescent label, Alexa Fluor 647 using a commercially available kit. Subsequently, T2 cells were pulsed with 20, 10, 5, or 0 μΜ of LLDFVRFMGV peptide at 37°C for 5 hours. After pulsing, the peptide-pulsed T2 cells, along with beads containing 20 known quantities of anti-human IgGi antibodies, were incubated with the fluorescently-labeled EBNA Clone 315 scFv-Fc. Once the cells and beads were analyzed on the FACS machine, the fluorescence intensities were correlated to each other, resulting in an estimation of the number of complexes on the surface of the T2 cells relative to the quantity of peptide 25 used for pulsing. These four values (337,091 sites with 20 μΜ, 149,688 sites with 10 μΜ, 76,040 sites with 5 μΜ, and no sites with 0 μΜ) were plotted on a graph and a trendline was used to create a standard curve (R2 = 0.9948) (Fig. 10A). Furthermore, when looking at the lower end of the spectrum, we have determined that an amount less than 40 nM of peptide will correspond to less 30 than 100 complexes on the surface of the cell (Fig. 10B), placing the detection level of the EBNA Clone 315 scFv-Fc fusion within that range. 8270468_1 (GHMatters) P94551 AU.1 5 October 2016 -51- 2014262282 05 Oct 2016

Example 23: Binding and specificity studies of WT1 Clone 45 scFv-Fc on recombinant FILA-A2-peptide complex and peptide-pulsed cells [00176] In order to do further studies regarding the presentation of the RMFPNAPYL peptide on the surface of APCs, the WT1 Clone 45 scFv-Fc 5 fusion protein was first purified and validated for binding to its targeted recombinant FILA-A2-peptide complex (Fig. 11 A). As was shown with the scFv, the fusion protein maintained its binding ability on the ELISA plate.

Next, we decided to check and see if the scFv-Fc maintained its binding ability and specificity on peptide-pulsed T2 cells. T2 cells were pulsed with the 10 RMFPNAPYL peptide or an irrelevant peptide in serum-free media with β2Μ at 37°C for 5 hours. Using flow cytometry, the scFv-Fc was able to detect T2 cells which had been pulsed with the RMFPNAPYL peptide, but failed to recognize the cells pulsed with an irrelevant peptide (Fig. 11B). These two assays further validated that the fusion protein acts in the same way as the 15 original scFv.

[00177] Subsequently, we decided to test whether the binding of the RMFPNAPYL peptide and scFv-Fc fusion protein were restricted to HLA-A2. HLA-A2+ and HLA-A2' BLCLs (DIMT and 6268A, respectively) were pulsed with the RMFPNAPYL peptide in serum-free media at 37°C for 5 hours. 20 Similarly to EBNA Clone 315, the WT1 Clone 45 scFv-Fc fusion protein was only able to recognize the peptide pulsed DIMT and not the HLA-A2' 6268A BLCL (Fig. 12). These results demonstrate that the RMFPNAPYL peptide is restricted to HLA-A2 and WT1 Clone 45 is only able to recognize it in the context of this complex. 25 Example 24: Antibody-dependent cellular cytotoxicity (ADCC) of EBNA Clone 315 scFv-Fc on peptide-pulsed cells [00178] In addition to using the scFv-Fc for antigen presentation studies, we tested whether the truncated human IgGi Fc region is capable of inducing antibody-dependent cellular cytotoxicity (ADCC). In order to avoid variability 30 amongst human donor lymphocytes, and in an effort to increase the chances 8270468_1 (GHMatters) P94551 .AU.1 5October2016 -52- 2014262282 05 Oct 2016 of observing cytotoxicity, Hong-fen Guo in our laboratory generated a CD16(V)-transduced NK92MI cell line. This NK92 cell variant is transduced with both IL-2 and the human CD16 activating Fc receptor (FcyRIIIA) containing a high affinity polymorphism (valine instead of phenylalanine at 5 position 158 on CD16) responsible for an enhancement in ADCC and clinical response to antibody-based immunotherapy (87, 88).

[00179] We used this cell line in combination with the EBNA Clone 315 scFv-Fc or an irrelevant, isotype-matched scFv-Fc to test whether the fusion protein can induce NK92MI-mediated ADCC against LLDFVRFMGV-pulsed LUY 10 (HLA-A2+) BLCL. At an E:T ratio of 42:1, EBNA Clone 315 scFv-Fc led to greater killing over background (with or without an irrelevant scFv-Fc) at the two highest concentrations tested (27-32% versus 13-15%) (Fig. 13). A similar magnitude of killing (over background) was also observed with other peptide-pulsed, HLA-A2+ target BLCLs (DIMT and JG19). These results show 15 that these truncated scFv-Fc fusion proteins maintain their Fc-mediated effector functions, despite being about 33% smaller than a full immunoglobulin.

Example 25: Construction and retroviral transduction of an HLA-A2-RMFPNAPYL-specific chimeric antigen receptor into NK92MI cells

20 [00180] In order to generate a CAR specific for the HLA-A2-RMFPNAPYL complex, the WT1 Clone 45 scFv would typically be fused to intracellular signaling domains of immune-modulatory proteins found in immune effector cells. A CAR expression vector (St. Jude CAR) in which a CD19-specific scFv is fused to the CD8a hinge/transmembrane region, 4-1BB and ΰϋ3ζ chain 25 was obtained and modified so that the anti-CD19 scFv was replaced with a WT1 Clone 45 scFv. However, due to restriction enzyme incompatibility issues between the St. Jude CAR vector and the Tomlinson library vector used for PCR, the entire CAR gene, containing the WT1 Clone 45 scFv, was commercially synthesized by Genescript. The resulting WT1 pUC57 vector 30 contained the desired WT1 Clone 45 CAR sequence flanked by EcoRI and Xhol. 8270468_1 (GHMatters) P94551 .AU.1 5 October 2016 -53- 2014262282 05 Oct 2016 [00181] An additional feature to the St. Jude CAR vector is an IRES-GFP sequence downstream of the CAR sequence. This allows for direct correlation of CAR expression with GFP without having to fuse both proteins together. In order to take advantage of this feature, we digested the WT1 5 pUC57 vector and St. Jude CAR vector using EcoRI and Xhol. Afterwards, the digested and undigested plasmids were run on a 1% agarose gel along with the lambda Hindlll and 100 bp markers. The highlighted bands corresponded to the anticipated sizes of the St. Jude plasmid lacking the CAR sequence (~6500 bp) and the WT1 Clone 45 CAR sequence lacking the 10 pUC57 plasmid (~1500 bp). These bands were excised from the gel, and after DNA purification, the two were ligated together. After the ligation products were transformed into E.coli, 8 colonies were selected at random and their plasmids isolated. The isolated plasmids were then validated by sequencing to determine whether they contain the WT1 Clone 45 scFv 15 sequence in the context of the CAR. In addition, the plasmids were also digested with EcoRI and Xhol and run on a 1 % agarose gel along with lambda Hindlll and 100 bp markers to validate their sizes. After demonstrating that both bands from each plasmid yielded the expected sizes, it was determined that the cloning was successful. 20 [00182] Once the WT1 Clone 45 CAR was generated (Fig. 23), the DNA was packaged into retrovirus using the 293T-based GP2 cell line. Once the retrovirus was generated in the culture media, it was recovered and concentrated. The concentrated virus was then used to infect 500,000 to 1,000,000 NK92MI cells in NK92MI growth media. After 3-4 days of culture, 25 the NK92MI cells infected with the retrovirus were compared to mock-infected cells (infected with empty retrovirus) with regards to GFP expression using flow cytometry. While the infection efficiency was approximately 27.5%, flow assisted cytometric cell sorting (FACS) allowed us to enrich the GFP-positive population to more than 98% positive (Fig. 23). 30 Example 26: Construction and retroviral transduction of an HLA-A2-LLDFVRFMGV-specific chimeric antigen receptor into NK92MI cells 8270468_1 (GHMatters) P94551 .AU.1 5October2016 -54- 2014262282 05 Oct 2016 [00183] Since the WT1 Clone 45 CAR required us to purchase the WT1 pUC57 vector, additional restriction sites were added to this construct for greater ease when swapping different scFvs. As a result, the EBNA Clone 315 scFv sequence could directly be cloned out of the Tomlinson vector from 5 which it was derived.

[00184] The first cloning step involved the removal of the WT1 Clone 45 scFv from the WT1 pUC57 vector using Sfil and Notl. The same digestion was done to the Tomlinson vector containing the EBNA Clone 315 scFv sequence. Once digested, both plasmids were run on a 1% agarose gel along with 10 lambda Hindi 11 and 100 bp markers. The highlighted bands corresponded to the WT1 pUC57 vector without a scFv, and the EBNA Clone 315 scFv excised from the Tomlinson vector. These bands were excised from the gel, the DNA was purified, and ligated together to yield the EBNA pUC57 vector. The ligation products were subsequently transformed into E.coli, 10 colonies were 15 selected at random, their plasmids were purified, and each DNA digested with EcoRI alone or EcoRI and Xhol. As anticipated, due to an inherent Xhol site within every scFv sequence derived from the Tomlison vector (with the exception of WT1 Clone 45 scFv in the context of pUC57 since the site was removed when purchased as a CAR from Genescript), the double digestion 20 yielded three separate bands.

[00185] For the second cloning step, in which the EBNA Clone 315 CAR sequence was excised from the pUC57 vector and added to the St. Jude CAR vector, a partial digestion of the EBNA pUC57 vector using Xhol was necessary. The EBNA pUC57 plasmids isolated from the 10 colonies above 25 were combined and digested with Xhol at room temperature for 1 minute. The reaction was quickly stopped by adding it to 4 separate wells of a 1 % agarose gel and running the DNA along with uncut plasmid, lambda Hindlll and 100 bp markers. The highlighted bands were determined to be the expected size of the linearized EBNA pUC57 plasmid (~4300 bp); this linearized plasmid is a 30 result of a random cut at either of the two Xhol sites. Subsequently, to obtain the complete CAR sequence (~1500 bp), the linearized plasmid was isolated from the gel and digested completely with EcoRI. The resulting double and 82704$8_1 (GHMatters) P94551 AU.1 5 October 201$ -55- 2014262282 05 Oct 2016 single digests were run on a 1% agarose gel along with the lambda Hindlll and 100 bp markers. The highlighted band corresponded to the anticipated size of the EBNA Clone 315 CAR gene, and as a result was excised from the gel, DNA purified, and ligated to the predigested (EcoRI and Xhol) St. Jude 5 CAR vector. After the ligation products were transformed into E.coli, 10 colonies were selected at random and their plasmids isolated. The isolated plasmids were then validated by sequencing to determine whether they contain the EBNA Clone 315 scFv sequence in the context of the CAR. In addition, the plasmids were also digested with EcoRI and run on a 1% 10 agarose gel along with the lambda Hindlll marker to validate their sizes. After demonstrating that the bands yielded the expected sizes, it was determined that the cloning was successful.

[00186] After the EBNA Clone 315 CAR was generated (Fig. 14), the DNA was packaged into retrovirus and used to infect NK92MI cells in the same way 15 as with the WT1 Clone 45 CAR. After 3-4 days of culture, the GFP expression level of the infected NK92MI cells were compared to mock-infected cells. The initial infection efficiency was approximately 24%, and after flow assisted cytometric cell sorting (FACS), the GFP-positive population was enriched to more than 90% positive (Fig. 14). 20 Example 27\ EBNA Clone 315 CAR-equipped NK92MI cells can detect cells bearing the specific HLA-A2-LLDFVRFMGV complex via CD107a expression [00187] Once the NK92MI cells were enriched for EBNA Clone 315 CAR expression, they were tested for their ability to recognize the targeted HLA-A2-EBNA3C complex. As an initial readout of NK92MI activation by target

25 cells, we assayed for cell surface expression of CD107a, a marker of NK cell and T cell degranulation (90, 91). T2 cells were loaded with 20 μΜ of the targeted peptide (LLDFVRFMGV), an irrelevant peptide (YMFPNAPYL) or no peptide. Using a 1:1 E:T ratio, the T2 cells were cocultured with EBNA Clone 315 CAR-expressing NK92MI cells in the presence of an anti-CD107a-PE 30 conjugated antibody at 37°C for 5 hours. As shown in Fig. 16A, NK92MI cells equipped with the EBNA Clone 315 CAR did not react to unpulsed T2 cells or 8270468_1 (GHMatters) P94551 .AU.1 5Odober2016 -56- 2014262282 05 Oct 2016 T2 cells pulsed with the irrelevant peptide, showing CD107a levels comparable to those of NK92MI cells cultured in the absence of targets. On the other hand, when the CAR-equipped NK92MI cells were cocultured with T2 cells that had been pulsed with the LLDFVRFMGV peptide, 27% of GFP+ 5 cells expressed CD107a above background levels. These results show that after scFv engineering, the CAR is able to maintain its specificity towards the targeted HLA-A2-peptide complex.

[00188] Next, in order to get a quantitative measurement of how sensitive this CAR is at activating NK92MI cells, we titrated down the LLDFVRFMGV 10 peptide concentration used to pulse T2 cells and measured their ability to activate the CAR-equipped NK92MI cells. As can be seen in Fig. 16B, the lower limit of response by the CAR-equipped NK92MI was at a peptide concentration of 10 nM, with a clear dose response curve beginning at the 600 nM concentration. Based on our earlier quantitation studies, this peptide 15 concentration corresponds to approximately 25 complexes on the cell surface. Compared to the levels necessary for epitope detection using the EBNA Clone 315 scFv or scFv-Fc (200-20 nM), the CAR seems to be a more sensitive approach at detecting low levels of MHC-peptide complex on the surface of APCs using flow cytometric analysis. 20 [00189] While T2 cells can present any peptide of interest, BLCLs naturally present their own peptides on their MHC Class I molecules. Similarly to T2, these endogenous peptides can be replaced by simple incubation with a substitute peptide of high enough affinity. Using a 1:1 E:T ratio, HLA-A2+ (DIMT) and HLA-A2' (6268A) BLCLs were pulsed with serum-free IMDM 25 medium or medium containing the LLDFVRFMGV, cocultured with EBNA Clone 315 CAR-expressing NK92MI cells as discussed above, and assayed for CD107a expression using flow cytometry. Peptide pulsed DIMT (HLA-A2+) induced 25% of GFP+ NK92MI cells to express CD107a (Fig. 17), in contrast to 0.54% for peptide pulsed 6268A (HLA-A2') and 1.09% for unpulsed DIMT. 30 This data further demonstrates both peptide specificity and HLA-A2 exclusivity of the EBNA Clone 315 CAR. 8270468J (GHMatters) P94551.AU.1 50ctober2016 -57- 2014262282 05 Oct 2016

Example 28: EBNA Clone 315 CAR-equipped NK92MI cells can destroy cells bearing the specific HLA-A2-LLDFVRFMGV complex via 51 Cr release [00190] While CD107a expression on NK cells and T cells reflect their activation, target cell lysis can also be measured using a conventional51 Cr 5 cytotoxicity assay. First, to get an idea of how sensitive the 51 Cr cytotoxicity assay is with regards to killing HLA-A2-EBNA3C expressing targets, T2 cells were pulsed with decreasing concentrations of the LLDFVRFMGV peptide at 37°C for 3 hours and subsequently labeled with 51 Cr as described in the Materials and Methods. The labeled T2 cells were then cocultured with EBNA 10 Clone 315 CAR-expressing NK92MI cells at 37°C for 2 hours at a 3:1 E:T ratio. Similar to the results seen in the CD107a assay (Fig. 16B), EBNA Clone 315 CAR expressing NK92MI cells were able kill T2 cells in a peptide-dependent manner, with 13.2% of 2 nM peptide-pulsed T2 cells being killed compared to 10.1 % with unpulsed T2 cells (Fig. 18). Relative to that which 15 can be detected using flow cytometric antibody staining, the level of sensitivity is in the order of 10-100 fold greater in favor of the CAR using two separate assays (CD107a and 51 Cr).

[00191] Next, DIMT and 6268A BLCLs pulsed with the LLDFVRFMGV peptide (20 μΜ) in serum-free IMDM (Fig 19A and B) were used as targets in 20 the 51Cr release assay. Similar to the results from the CD107a assay (Fig. 17), only the HLA-A2+ DIMT BLCL were lysed by the CAR-equipped NK92MI cells (Fig. 19A), which could be blocked using the purified EBNA Clone 315 scFv-Fc (Fig. 19B). In addition, the ability to block cytotoxicity was not restricted to the scFv-Fc protein since a commercial anti-HLA-A2 (BB7.2) 25 antibody also possessed blocking ability (data not shown). This blocking data recapitulates results seen with other MHC-restricted, peptide-specific antibodies on antigen-specific cytolytic T cells.

[00192] Although the lytic potential of the CAR-equipped NK92MI cells was clearly evident when targets were artificially pulsed with the relevant peptide, 30 cytotoxicity against naturally processed HLA-A2-peptide complexes is of clinical relevance. Here, CAR-equipped NK92MI cells were tested against a 8270468_1 (GHMatters) P94551 .AU.1 5 October 201$ -58- 2014262282 05 Oct 2016 panel of HLA-A2+ (DIMT and JG19) and HLA-A2' (6268A and GKO) unpulsed BLCLs. While the level of cytoxicity was low, 23.0% for DIMT and 8.9% for JG19 (30:1 E:T ratio), this was highly significant when compared to 3.6% for GKO and 1.8% for 6268A (Fig. 20A). In addition, when the cytotoxicity assay 5 was performed in the presence of EBNA Clone 315 scFv-Fc, the killing capacity could be reduced by approximately 46% when compared to that with an irrelevant scFv-Fc or in the absence of antibody (Fig 20B). These findings demonstrate the utility and specificity of TCR-like CARs in reprogramming effector immune cells to engage antigen whose expression is below the 10 detection limit using conventional flow cytometry.

[00193] Lastly, since both the EBNA Clone 315 CAR and scFv-Fc fusion protein have the same variable sequences used for detecting the HLA-A2-LLDFVRFMGV complex, we decided to directly compare CAR-mediated cytotoxicity with ADCC since both approaches are currently being used 15 independently for the treatment of cancer patients. First, the DIMT BLCL was pulsed with the LLDFVRFMGV peptide at 20 μΜ in serum-free IMDM media at 37°C for 2 hours. The pulsed BLCL was then labeled with 51 Cr and cocultured with either EBNA Clone 315 CAR or CD16(V)-expressing NK92MI cells along with EBNA Clone 315 scFv-Fc or an irrelevant scFv-Fc at an E:T ratio of 15:1 20 for 3 hours at 37°C. At a EBNA Clone 315 scFv-Fc concentration of 0.5 pg/ml, CD16(V) NK92MI cells were able to kill about 30-35% of cells, compared to 10-15% with an irrelevant scFv-Fc or no antibody at all (Fig. 21). When the ADCC experiment was carried out using higher scFv-Fc concentrations, the cytotoxicity percentage did not change (data not shown). 25 On the other hand, at the same E:T ratio, EBNA Clone 315 CAR-equipped NK92MI cells were able to kill 80-90% of the same peptide-pulsed target cells; and the EBNA Clone 315 scFv-Fc was included as a blocking control (Fig. 22). These results demonstrate that the CAR-mediated killing involving NK92MI cells is a far more potent means of target cell lysis compared to 30 ADCC in our setting.

Example 29: WT1 Clone 45 CAR-equipped NK92MI cells can destroy cells bearing the specific HLA-A2-RMFPNAPYL complex via 51 Cr release 82704$8_1 (GHMatters) P94551 .AU.1 5 October 2016 -59- 2014262282 05 Oct 2016 [00194] Along with the EBNA Clone 315 CAR, we decided to test the cytolytic ability of the WT1 Clone 45 CAR in the context of NK92MI cells. First, DIMT and 6268A BLCLs were pulsed with the RMFPNAPYL peptide (40 pg/ml) in serum-free IMDM at 37°C for 3-5 hours. Subsequently, the target cells were 5 labeled with 51 Cr and cocultured with the CAR-equipped NK92MI cells at 37°C for 4 hours. Of the two peptide-pulsed BLCLs, only the HLA-A2+ DIMT could be lysed (~70% versus ~5% with 6268A) at a 10:1 E:T ratio (Fig. 24). In addition, CAR-mediated cytotoxicity could be blocked using a commercial anti-HLA-A2 antibody by approximately 45% (Fig. 25), further demonstrating 10 specificity.

[00195] Next, we decided to tested the cytolytic capacity of WT1 Clone 45 CAR-equipped NK92MI cells against cell lines which might natively express the HLA-A2-RMFPNAPYL complex. Due to previously published data (92), and conversations with Dr. Richard O’Reilly’s laboratory here at MSKCC, 15 researchers have demonstrated that WT1 can be constitutively activated in all BLCLs derived from EBV immortalization. More specifically, O’Reilly’s group was able to show WT1 transcript in the DIMT BLCL (data not shown). As a result, we first decided to test WT1 Clone 45 CAR-mediated killing against unpulsed HLA-A2+ DIMT and HLA-A2' 6268A BLCL. Similarly to what was 20 seen with the EBNA Clone 315 CAR, WT1 Clone 45 CAR-equipped NK92MI cells were able to kill unpulsed DIMT at a lower capacity than peptide-pulsed DIMT. While the level of cytotoxicity was lower, ~35% for DIMT at a 20:1 E:T ratio, it was far greater when compared to 6268A (~5%) (Fig. 26A). In addition, when the cytotoxicity assay was performed in the presence of the 25 WT1 Clone 45 scFv-Fc, the killing capacity could be reduced by approximately 43% relative to an irrelevant scFv-Fc or in the absence of antibody (Fig 26B). These findings correspond well with what was seen using the EBNA Clone 315 CAR and further demonstrate the utility and specificity of TCR-like CARs in reprogramming effector immune cells to engage antigen. 30 [00196] Lastly, CAR-mediated cytotoxicity against two cell lines which are HLA-A2-positive and previously shown to express WT1 was tested. OVCAR-3 is a cell line established from malignant ascites of a patient with progressive 8270468_1 (GHMatters) P94551 .AU.1 5 October 2016 -60- 2014262282 05 Oct 2016 adenocarcinoma of the ovary (93) and later shown to contain WT1 mRNA (94). In addition, 697 is a human pre-B cell leukemia established from bone marrow cells obtained from a child with relapsed acute lymphocytic leukemia (ALL) (95). Since then, several groups have shown that this cell line also 5 expresses high levels of both WT1 transcript and protein (96, 97). WT1 Clone 45 CAR-expressing NK92MI cells were cocultured with 51Cr labeled OVCAR-3 and 697 cells at 37°C for 4 hours. CAR-equipped NK92MI cells were able to lyse approximately 20-30% of 697 and OVCAR-3 cells at a 20:1 E:T ratio, which decreased with the number of effector cells used in the assay. This 10 data demonstrates that these two cell types are sensitive to WT1 Clone 45 CAR-equipped NK92MI cells and provides further evidence for their utility in the treatment of HLA-A2+/WT1+ malignancies.

EQUIVALENTS

[00197] The foregoing written specification is considered to be sufficient to 15 enable one skilled in the art to practice the invention. The foregoing description and Examples detail certain embodiments of the invention and describes the best mode contemplated by the inventors. It will be appreciated, however, that no matter how detailed the foregoing may appear, the invention may be practiced in many ways and the invention should be 20 construed in accordance with the appended claims and any equivalents thereof.

[00198] In the claims which follow and in the preceding description of the invention, except where the context requires otherwise due to express language or necessary implication, the word “comprise” or variations such as 25 “comprises” or “comprising” is used in an inclusive sense, i.e. to specify the presence of the stated features but not to preclude the presence or addition of further features in various embodiments of the invention. 8270468_1 (GHMatters) P94551 .AU.1 5October2016 -61- 2014262282 05 Oct 2016

References 1. Kohler, G., and C. Milstein. 1975. Continuous cultures of fused cells secreting antibody of predefined specificity. Nature 256:495-497. 2. Oldham, R. K., and R. O. Dillman. 2008. Monoclonal antibodies in cancer 5 therapy: 25 years of progress. J Clin Oncol 26:1774-1777. 3. Tassev, D. V., and N. K. Cheung. 2009. Monoclonal antibody therapies for solid tumors. Expert opinion on biological therapy 9:341-353. 4. Fridman, W. H. 1991. Fc receptors and immunoglobulin binding factors.

Faseb J 5:2684-2690. 10 5. Kratz, F., I. A. Muller, C. Ryppa, and A. Warnecke. 2008. Prodrug strategies in anticancer chemotherapy. ChemMedChem 3:20-53. 6. Magdelaine-Beuzelin, C., Q. Kaas, V. Wehbi, M. Ohresser, R. Jefferis, Μ. P. Lefranc, and H. Watier. 2007. Structure-function relationships of the variable domains of monoclonal antibodies approved for cancer treatment. Critical 15 reviews in oncology/hematology 64:210-225. 7. Schmidt, Μ. M., and K. D. Wittrup. 2009. A modeling analysis of the effects of molecular size and binding affinity on tumor targeting. Molecular cancer therapeutics 8:2861-2871. 8. Kloetzel, P. M., and F. Ossendorp. 2004. Proteasome and peptidase function 20 in MHC-class-l-mediated antigen presentation. Current opinion in immunology 16:76-81. 9. Donaldson, J. G., and D. B. Williams. 2009. Intracellular assembly and trafficking of MHC class I molecules. Traffic (Copenhagen, Denmark) 10:1745-1752. 25 10. Hunt, D. F., R. A. Henderson, J. Shabanowitz, K. Sakaguchi, H. Michel, N.

Sevilir, A. L. Cox, E. Appella, and V. H. Engelhard. 1992. Characterization of peptides bound to the class I MHC molecule HLA-A2.1 by mass spectrometry. Science (New York, N.Y255:1261 -1263. 8270468.1 (GHMatters) P94551 AU.1 5 October 201$ -62- 2014262282 05 Oct 2016 11. Hofmann, S., M. Gluckmann, S. Kausche, A. Schmidt, C. Corvey, R. Lichtenfels, C. Huber, C. Albrecht, M. Karas, and W. Herr. 2005. Rapid and sensitive identification of major histocompatibility complex class l-associated tumor peptides by Nano-LC MALDI MS/MS. Mol Cell Proteomics 4:1888- 5 1897. 12. Smith, G. P. 1985. Filamentous fusion phage: novel expression vectors that display cloned antigens on the virion surface. Science (New York, N. Y 228:1315-1317. 13. Carmen, S., and L. Jermutus. 2002. Concepts in antibody phage display. 10 Briefings in functional genomics & proteomics 1:189-203. 14. Baeuerle, P. A., and C. Reinhardt. 2009. Bispecific T-cell engaging antibodies for cancer therapy. Cancer research 69:4941-4944. 15. Bratkovic, T. 2010. Progress in phage display: evolution of the technique and its application. Cell Mol Life Sci 67:749-767. 15 16. Kourilov, V., and M. Steinitz. 2002. Magnetic-bead enzyme-linked immunosorbent assay verifies adsorption of ligand and epitope accessibility.

Analytical biochemistry 311:166-170. 17. Andersen, P. S., A. Stryhn, B. E. Hansen, L. Fugger, J. Engberg, and S.

Buus. 1996. A recombinant antibody with the antigen-specific, major 20 histocompatibility complex-restricted specificity of T cells. Proceedings of the

National Academy of Sciences of the United States of America 93:1820-1824. 18. Porgador, A., J. W. Yewdell, Y. Deng, J. R. Bennink, and R. N. Germain. 1997. Localization, quantitation, and in situ detection of specific peptide-MHC class I complexes using a monoclonal antibody. Immunity 6:715-726. 25 19. Zhong, G., C. Reis e Sousa, and R. N. Germain. 1997. Production, specificity, and functionality of monoclonal antibodies to specific peptide-major histocompatibility complex class II complexes formed by processing of exogenous protein. Proceedings of the National Academy of Sciences of the United States of America 94:13856-13861. 8270468_1 (GHMatters) P94551 .AU.1 5October2016 -63- 2014262282 05 Oct 2016 20. Dadaglio, G., C. A. Nelson, Μ. B. Deck, S. J. Petzold, and E. R. Unanue. 1997. Characterization and quantitation of peptide-MHC complexes produced from hen egg lysozyme using a monoclonal antibody. Immunity 6:727-738. 21. Lev, A., G. Denkberg, C. J. Cohen, M. Tzukerman, K. L. Skorecki, P. 5 Chames, H. R. Hoogenboom, and Y. Reiter. 2002. Isolation and characterization of human recombinant antibodies endowed with the antigen-specific, major histocompatibility complex-restricted specificity of T cells directed toward the widely expressed tumor T-cell epitopes of the telomerase catalytic subunit. Cancer research 62:3184-3194. 10 22. Chames, P., S. E. Hufton, P. G. Coulie, B. Uchanska-Ziegler, and H. R.

Hoogenboom. 2000. Direct selection of a human antibody fragment directed against the tumor T-cell epitope HLA-A1-MAGE-A1 from a nonimmunized phage-Fab library. Proceedings of the National Academy of Sciences of the United States of America 97:7969-7974. 15 23. Denkberg, G., C. J. Cohen, A. Lev, P. Chames, H. R. Hoogenboom, and Y.

Reiter. 2002. Direct visualization of distinct T cell epitopes derived from a melanoma tumor-associated antigen by using human recombinant antibodies with MHC- restricted T cell receptor-like specificity. Proceedings of the National Academy of Sciences of the United States of America 99:9421-9426. 20 24. Held, G., M. Matsuo, M. Epel, S. Gnjatic, G. Ritter, S. Y. Lee, T. Y. Tai, C. J.

Cohen, L. J. Old, M. Pfreundschuh, Y. Reiter, H. R. Hoogenboom, and C. Renner. 2004. Dissecting cytotoxic T cell responses towards the NY-ESO-1 protein by peptide/MHC-specific antibody fragments. European journal of immunology 34:2919-2929. 25 25. Held, G., A. Wadle, N. Dauth, G. Stewart-Jones, C. Sturm, M. Thiel, C. Zwick, D. Dieckmann, G. Schuler, H. R. Hoogenboom, F. Levy, V. Cerundolo, M. Pfreundschuh, and C. Renner. 2007. MHC-peptide-specific antibodies reveal inefficient presentation of an HLA-A*0201-restricted, Melan-A-derived peptide after active intracellular processing. European journal of immunology 30 37:2008-2017. 26. Klechevsky, E., M. Gallegos, G. Denkberg, K. Palucka, J. Banchereau, C. Cohen, and Y. Reiter. 2008. Antitumor activity of immunotoxins with T-cell 82704$8_1 (GHMatters) P94551 .AU.1 5 October 201$ -64- 2014262282 05 Oct 2016 receptor-like specificity against human melanoma xenografts. Cancer research 68:6360-6367. 27. Cohen, C. J., O. Sarig, Y. Yamano, U. Tomaru, S. Jacobson, and Y. Reiter. 2003. Direct phenotypic analysis of human MHC class I antigen presentation: 5 visualization, quantitation, and in situ detection of human viral epitopes using peptide-specific, MHC-restricted human recombinant antibodies. J Immunol 170:4349-4361. 28. Cohen, C. J., N. Hoffmann, M. Farago, H. R. Hoogenboom, L. Eisenbach, and Y. Reiter. 2002. Direct detection and quantitation of a distinct T-cell epitope 10 derived from tumor-specific epithelial cell-associated mucin using human recombinant antibodies endowed with the antigen-specific, major histocompatibility complex-restricted specificity of T cells. Cancer research 62:5835-5844. 29. Epel, Μ., I. Carmi, S. Soueid-Baumgarten, S. K. Oh, T. Bera, I. Pastan, J. 15 Berzofsky, and Y. Reiter. 2008. Targeting TARP, a novel breast and prostate tumor-associated antigen, with T cell receptor-like human recombinant antibodies. European journal of immunology 38:1706-1720. 30. Neumann, F., C. Sturm, M. Hulsmeyer, N. Dauth, P. Guillaume, I. F.

Luescher, M. Pfreundschuh, and G. Held. 2009. Fab antibodies capable of 20 blocking T cells by competitive binding have the identical specificity but a higher affinity to the MHC-peptide-complex than the T cell receptor. Immunology letters 125:86-92. 31. Makler, O., K. Oved, N. Netzer, D. Wolf, and Y. Reiter. 2010. Direct visualization of the dynamics of antigen presentation in human cells infected 25 with cytomegalovirus revealed by antibodies mimicking TCR specificity.

European journal of immunology 40:1552-1565. 32. Michaeli, Y., G. Denkberg, K. Sinik, L. Lantzy, C. Chih-Sheng, C. Beauverd, T. Ziv, P. Romero, and Y. Reiter. 2009. Expression hierarchy of T cell epitopes from melanoma differentiation antigens: unexpected high level 30 presentation of tyrosinase-HI_A-A2 Complexes revealed by peptide-specific, MHC-restricted, TCR-like antibodies. J Immunol 182:6328-6341. 8270468_1 (GHMatters) P94551 .AU.1 5October2016 -65- 2014262282 05 Oct 2016 33. Weidanz, J. A., P. Piazza, H. Hickman-Miller, D. Woodburn, T. Nguyen, A. Wahl, F. Neethling, M. Chiriva-lnternati, C. R. Rinaldo, and W. H. Hildebrand. 2007. Development and implementation of a direct detection, quantitation and validation system for class I MHC self-peptide epitopes. Journal of 5 immunological methods 318:47-58. 34. Verma, B., O. E. Hawkins, F. A. Neethling, S. L. Caseltine, S. R. Largo, W. H. Hildebrand, and J. A. Weidanz. 2009. Direct discovery and validation of a peptide/MHC epitope expressed in primary human breast cancer cells using a TCRm monoclonal antibody with profound antitumor properties. Cancer 10 Immunol Immunother 59:563-573. 35. Verma, B., F. A. Neethling, S. Caseltine, G. Fabrizio, S. Largo, J. A. Duty, P. Tabaczewski, and J. A. Weidanz. 2010. TCR mimic monoclonal antibody targets a specific peptide/HLA class I complex and significantly impedes tumor growth in vivo using breast cancer models. J Immunol 184:2156-2165. 15 36. Wittman, V. P., D. Woodburn, T. Nguyen, F. A. Neethling, S. Wright, and J. A.

Weidanz. 2006. Antibody targeting to a class I MHC-peptide epitope promotes tumor cell death. J Immunol 177:4187-4195. 37. Reiter, Y., A. Di Carlo, L. Fugger, J. Engberg, and I. Pastan. 1997. Peptide-specific killing of antigen-presenting cells by a recombinant antibody-toxin

20 fusion protein targeted to major histocompatibility complex/peptide class I complexes with T cell receptor-like specificity. Proceedings of the National Academy of Sciences of the United States of America 94:4631-4636. 38. Stewart-Jones, G., A. Wadle, A. Hombach, E. Shenderov, G. Held, E.

Fischer, S. Kleber, N. Nuber, F. Stenner-Liewen, S. Bauer, A. McMichael, A. 25 Knuth, H. Abken, A. A. Hombach, V. Cerundolo, E. Y. Jones, and C. Renner. 2009. Rational development of high-affinity T-cell receptor-like antibodies. Proceedings of the National Academy of Sciences of the United States of America 106:5784-5788. 39. Willemsen, R. A., R. Debets, E. Hart, H. R. Hoogenboom, R. L. Bolhuis, and 30 P. Chames. 2001. A phage display selected fab fragment with MHC class I- restricted specificity for MAGE-A1 allows for retargeting of primary human T lymphocytes. Gene therapy 8:1601-1608. 8270468_1 (GHMatters) P94551 .AU.1 5Oclober2016 -66- 2014262282 05 Oct 2016 40. Chames, P., R. A. Willemsen, G. Rojas, D. Dieckmann, L. Rem, G. Schuler, R. L. Bolhuis, and H. R. Hoogenboom. 2002. TCR-like human antibodies expressed on human CTLs mediate antibody affinity-dependent cytolytic activity. J Immunol 169:1110-1118. 5 41. Willemsen, R. A., C. Ronteltap, P. Chames, R. Debets, and R. L. Bolhuis. 2005. T cell retargeting with MHC class l-restricted antibodies: the CD28 costimulatory domain enhances antigen-specific cytotoxicity and cytokine production. J Immunol 174:7853-7858. 42. Doubrovina, E. S., Μ. M. Doubrovin, S. Lee, J. H. Shieh, G. Heller, E. Pamer, 10 and R. J. O'Reilly. 2004. In vitro stimulation with WT1 peptide-loaded Epstein-

Barr virus-positive B cells elicits high frequencies of WT1 peptide-specific T cells with in vitro and in vivo tumoricidal activity. Clin Cancer Res 10:7207-7219. 43. Hiasa, A., H. Nishikawa, M. Hirayama, S. Kitano, S. Okamoto, H. Chono, S. 15 S. Yu, J. Mineno, Y. Tanaka, N. Minato, I. Kato, and H. Shiku. 2009. Rapid alphabeta TCR-mediated responses in gammadelta T cells transduced with cancer-specific TCR genes. Gene therapy 16:620-628. 44. Ochi, T., H. Fujiwara, and M. Yasukawa. 2010. Application of adoptive T-cell therapy using tumor antigen-specific T-cell receptor gene transfer for the 20 treatment of human leukemia. Journal of biomedicine & biotechnology 2010:521248. 45. Davies, D. M., and J. Maher. 2010. Adoptive T-cell immunotherapy of cancer using chimeric antigen receptor-grafted T cells. Archivum immunologiae et therapiae experimentalis 58:165-178. 25 46. Yee, C. 2010. Adoptive therapy using antigen-specific T-cell clones. Cancer journal (Sudbury, Mass 16:367-373. 47. Bucks, C. M., J. A. Norton, A. C. Boesteanu, Y. M. Mueller, and P. D. Katsikis. 2009. Chronic antigen stimulation alone is sufficient to drive CD8+ T cell exhaustion. J Immunol 182:6697-6708. 30 48. Okamoto, S., J. Mineno, H. Ikeda, H. Fujiwara, M. Yasukawa, H. Shiku, and I.

Kato. 2009. Improved expression and reactivity of transduced tumor-specific 8270468_1 (GHMatters) P94551 .AU.1 5 October 2016 -67- 2014262282 05 Oct 2016 TCRs in human lymphocytes by specific silencing of endogenous TCR. Cancer research 69:9003-9011. 49. Cartellieri, Μ., M. Bachmann, A. Feldmann, C. Bippes, S. Stamova, R. Wehner, A. Temme, and M. Schmitz. 2010. Chimeric antigen receptor- 5 engineered T cells for immunotherapy of cancer. Journal of biomedicine & biotechnology 2010:956304. 50. Eshhar, Z., T. Waks, G. Gross, and D. G. Schindler. 1993. Specific activation and targeting of cytotoxic lymphocytes through chimeric single chains consisting of antibody-binding domains and the gamma or zeta subunits of 10 the immunoglobulin and T-cell receptors. Proceedings of the National

Academy of Sciences of the United States of America 90:720-724. 51. Weijtens, Μ. E., R. A. Willemsen, D. Valerio, K. Stam, and R. L. Bolhuis. 1996. Single chain lg/gamma gene-redirected human T lymphocytes produce cytokines, specifically lyse tumor cells, and recycle lytic capacity. J Immunol 15 157:836-843. 52. Lamers, C. H., S. C. Langeveld, C. M. Groot-van Ruijven, R. Debets, S. Sleijfer, and J. W. Gratama. 2007. Gene-modified T cells for adoptive immunotherapy of renal cell cancer maintain transgene-specific immune functions in vivo. Cancer Immunol Immunother 56^875-^1883. 20 53. Lamers, C. H„ S. Sleijfer, A. G. Vulto, W. H. Kruit, M. Kliffen, R. Debets, J. W.

Gratama, G. Stoter, and E. Oosterwijk. 2006. Treatment of metastatic renal cell carcinoma with autologous T-lymphocytes genetically retargeted against carbonic anhydrase IX: first clinical experience. J Clin Oncol 24:e20-22. 54. Kershaw, Μ. H., J. A. Westwood, L. L. Parker, G. Wang, Z. Eshhar, S. A. 25 Mavroukakis, D. E. White, J. R. Wunderlich, S. Canevari, L. Rogers-Freezer, C. C. Chen, J. C. Yang, S. A. Rosenberg, and P. Hwu. 2006. A phase I study on adoptive immunotherapy using gene-modified T cells for ovarian cancer. Clin Cancer Res 12:6106-6115. 55. Till, B. G., M. C. Jensen, J. Wang, E. Y. Chen, B. L. Wood, H. A. Greisman, 30 X. Qian, S. E. James, A. Raubitschek, S. J. Forman, A. K. Gopal, J. M. Pagel, C. G. Lindgren, P. D. Greenberg, S. R. Riddell, and O. W. Press. 2008. 8270468_1 (GHMatters) P94551 .AU.1 5 October 201$ -68- 2014262282 05 Oct 2016

Adoptive immunotherapy for indolent non-Hodgkin lymphoma and mantle cell lymphoma using genetically modified autologous CD20-specific T cells. Blood 112:2261-2271. 56. Pule, M. A., B. Savoldo, G. D. Myers, C. Rossig, Η. V. Russell, G. Dotti, Μ. H. 5 Huls, E. Liu, A. P. Gee, Z. Mei, E. Yvon, H. L. Weiss, H. Liu, C. M. Rooney, H. E. Heslop, and Μ. K. Brenner. 2008. Virus-specific T cells engineered to coexpress tumor-specific receptors: persistence and antitumor activity in individuals with neuroblastoma. Nature medicine 14:1264-1270. 57. Subramanian, C., S. Hasan, M. Rowe, M. Hottiger, R. Orre, and E. S. 10 Robertson. 2002. Epstein-Barr virus nuclear antigen 3C and prothymosin alpha interact with the p300 transcriptional coactivator at the CH1 and CH3/HAT domains and cooperate in regulation of transcription and histone acetylation. Journal of virology 76:4699-4708. 58. Epstein, M. A., G. Henle, B. G. Achong, and Y. M. Barr. 1965. Morphological 15 and Biological Studies on a Virus in Cultured Lymphoblasts from Burkitt's

Lymphoma. The Journal of experimental medicine 121:761-770. 59. Cohen, J. I. 2000. Epstein-Barr virus infection. The New England journal of medicine 343:481-492.

60. Garrido, J. L., S. Maruo, K. Takada, and A. Rosendorff. 2009. EBNA3C 20 interacts with Gadd34 and counteracts the unfolded protein response.

Virology journal 6:231. 61. Kerr, B. Μ., N. Kienzle, J. M. Burrows, S. Cross, S. L. Silins, M. Buck, E. M. Benson, B. Coupar, D. J. Moss, and T. B. Sculley. 1996. Identification of type B-specific and cross-reactive cytotoxic T-lymphocyte responses to Epstein- 25 Barr virus. Journal of virology 70:8858-8864. 62. Yang, L, Y. Han, F. Suarez Saiz, and M. D. Minden. 2007. A tumor suppressor and oncogene: the WT1 story. Leukemia 21:868-876. 63. Sugiyama, H. 2010. WT1 (Wilms' tumor gene 1): biology and cancer immunotherapy. Japanese journal of clinical oncology 40:377-387. 8270468_1 (GHMatters) P94551 .AU.1 5October2016 -69- 2014262282 05 Oct 2016 64. Coppes, M. J., C. E. Campbell, and B. R. Williams. 1993. The role of WT1 in Wilms tumorigenesis. Faseb J 7:886-895. 65. O'Reilly, R. J., T. Dao, G. Koehne, D. Scheinberg, and E. Doubrovina. 2010. Adoptive transfer of unselected or leukemia-reactive T-cells in the treatment 5 of relapse following allogeneic hematopoietic cell transplantation. Seminars in immunology 22:162-172. 66. Rezvani, K., J. M. Brenchley, D. A. Price, Y. Kilical, E. Gostick, A. K. Sewell, J. Li, S. Mielke, D. C. Douek, and A. J. Barrett. 2005. T-cell responses directed against multiple HLA-A*0201-restricted epitopes derived from Wilms' 10 tumor 1 protein in patients with leukemia and healthy donors: identification, quantification, and characterization. Clin Cancer Res 11:8799-8807. 67. Borbulevych, O. Y., P. Do, and B. M. Baker. 2010. Structures of native and affinity-enhanced WT1 epitopes bound to HLA-A*0201: implications for WT1-based cancer therapeutics. Molecular immunology 47:2519-2524. 15 68. Gao, L, I. Bellantuono, A. Elsasser, S. B. Marley, Μ. Y. Gordon, J. M.

Goldman, and H. J. Stauss. 2000. Selective elimination of leukemic CD34(+) progenitor cells by cytotoxic T lymphocytes specific for WT1. Blood 95:2198-2203. 69. Mailander, V., C. Scheibenbogen, E. Thiel, A. Letsch, I. W. Blau, and U. 20 Keilholz. 2004. Complete remission in a patient with recurrent acute myeloid leukemia induced by vaccination with WT 1 peptide in the absence of hematological or renal toxicity. Leukemia 18:165-166. 70. Keilholz, U., A. Letsch, A. Busse, A. M. Asemissen, S. Bauer, I. W. Blau, W. K. Hofmann, L. Uharek, E. Thiel, and C. Scheibenbogen. 2009. A clinical and 25 immunologic phase 2 trial of Wilms tumor gene product 1 (WT1) peptide vaccination in patients with AML and MDS. S/oocM 13:6541-6548. 71. Garboczi, D. N., P. Ghosh, U. Utz, Q. R. Fan, W. E. Biddison, and D. C.

Wiley. 1996. Structure of the complex between human T-cell receptor, viral peptide and HLA-A2. Nature 384:134-141. 30 72. Garboczi, D. N., U. Utz, P. Ghosh, A. Seth, J. Kim, E. A. VanTienhoven, W. E.

Biddison, and D. C. Wiley. 1996. Assembly, specific binding, and 8270468.1 (GHMatters) P94551 AU.1 5 October 2016 -70- 2014262282 05 Oct 2016 crystallization of a human TCR-alphabeta with an antigenic Tax peptide from human T lymphotropic virus type 1 and the class I MHC molecule HLA-A2. J Immunol 157:5403-5410. 73. Altman, J. D., P. A. Moss, P. J. Goulder, D. H. Barouch, M. G. McHeyzer- 5 Williams, J. I. Bell, A. J. McMichael, and Μ. M. Davis. 1996. Phenotypic analysis of antigen-specific T lymphocytes. Science (New York, N.Y274:94- 96. 74. Busch, D. Η., I. M. Pilip, S. Vijh, and E. G. Pamer. 1998. Coordinate regulation of complex T cell populations responding to bacterial infection. 10 Immunity 8:353-362. 75. Schatz, P. J. 1993. Use of peptide libraries to map the substrate specificity of a peptide-modifying enzyme: a 13 residue consensus peptide specifies biotinylation in Escherichia coli. Bio/technology (Nature Publishing Company) 11:1138-1143. 15 76. de Wildt, R. M., C. R. Mundy, B. D. Gorick, and I. M. Tomlinson. 2000.

Antibody arrays for high-throughput screening of antibody-antigen interactions. Nature biotechnology 18:989-994. 77. Imai, C., K. Mihara, M. Andreansky, I. C. Nicholson, C. H. Pui, T. L. Geiger, and D. Campana. 2004. Chimeric receptors with 4-1BB signaling capacity 20 provoke potent cytotoxicity against acute lymphoblastic leukemia. Leukemia 18:676-684. 78. Kim, Y. J., S. H. Kim, P. Mantel, and B. S. Kwon. 1998. Human 4-1 BB regulates CD28 co-stimulation to promote Th1 cell responses. European journal of immunology 28:881-890. 25 79. Hurtado, J. C., Y. J. Kim, and B. S. Kwon. 1997. Signals through 4-1 BB are costimulatory to previously activated splenic T cells and inhibit activation-induced cell death. J Immunol 158:2600-2609. 80. Jackson, R. J., Μ. T. Howell, and A. Kaminski. 1990. The novel mechanism of initiation of picornavirus RNA translation. Trends in biochemical sciences 30 15:477-483. 8270468_1 (GHMatters) P94551 .AU.1 5October2016 -71- 2014262282 05 Oct 2016 81. Baas, E. J., Η. M. van Santen, M. J. Kleijmeer, H. J. Geuze, P. J. Peters, and H. L. Ploegh. 1992. Peptide-induced stabilization and intracellular localization of empty HLA class I complexes. The Journal of experimental medicine 176:147-156. 5 82. Koehne, G., K. M. Smith, T. L. Ferguson, R. Y. Williams, G. Heller, E. G.

Pamer, B. Dupont, and R. J. O'Reilly. 2002. Quantitation, selection, and functional characterization of Epstein-Barr virus-specific and alloreactive T cells detected by intracellular interferon-gamma production and growth of cytotoxic precursors. Blood 99:1730-1740. 10 83. Gould, L. H., J. Sui, H. Foellmer, T. Oliphant, T. Wang, M. Ledizet, A.

Murakami, K. Noonan, C. Lambeth, K. Kar, J. F. Anderson, A. M. de Silva, M. S. Diamond, R. A. Koski, W. A. Marasco, and E. Fikrig. 2005. Protective and therapeutic capacity of human single-chain Fv-Fc fusion proteins against West Nile virus. Journal of virology 79:14606-14613. 15 84. van der Merwe, P. A., and S. J. Davis. 2003. Molecular interactions mediating T cell antigen recognition. Annual review of immunology 21:659-684. 85. Sykulev, Y., R. J. Cohen, and Η. N. Eisen. 1995. The law of mass action governs antigen-stimulated cytolytic activity of CD8+ cytotoxic T lymphocytes. Proceedings of the National Academy of Sciences of the United States of 20 America 92:11990-11992. 86. Schodin, B. A., T. J. Tsomides, and D. M. Kranz. 1996. Correlation between the number of T cell receptors required for T cell activation and TCR-ligand affinity. Immunity 5:137-146. 87. Cartron, G., L. Dacheux, G. Salles, P. Solal-Celigny, P. Bardos, P. Colombat, 25 and H. Watier. 2002. Therapeutic activity of humanized anti-CD20 monoclonal antibody and polymorphism in IgG Fc receptor FcgammaRllla gene. Blood 99:754-758. 88. Weng, W. K., R. S. Negrin, P. Lavori, and S. J. Horning. Immunoglobulin G Fc receptor FcgammaRllla 158 V/F polymorphism correlates with rituximab- 30 induced neutropenia after autologous transplantation in patients with non-

Hodgkin's lymphoma. J Clin Oncol 28:279-284. 8270468_1 (GHMatters) P94551 AU.1 5 October 2016 -72- 2014262282 05 Oct 2016 89. George, A. J., J. Stark, and C. Chan. 2005. Understanding specificity and sensitivity of T-cell recognition. Trends in immunology 26:653-659. 90. Betts, M. R., J. M. Brenchley, D. A. Price, S. C. De Rosa, D. C. Douek, M. Roederer, and R. A. Koup. 2003. Sensitive and viable identification of 5 antigen-specific CD8+ T cells by a flow cytometric assay for degranulation.

Journal of immunological methods 281:65-78. 91. Alter, G., J. M. Malenfant, and M. Altfeld. 2004. CD107a as a functional marker for the identification of natural killer cell activity. Journal of immunological methods 294:15-22. 10 92. Spinsanti, P., U. de Grazia, A. Faggioni, L. Frati, A. Calogero, and G. Ragona. 2000. Wilms' tumor gene expression by normal and malignant human B lymphocytes. Leukemia & lymphoma 38:611-619. 93. Hamilton, T. C., R. C. Young, W. M. McKoy, K. R. Grotzinger, J. A. Green, E. W. Chu, J. Whang-Peng, A. M. Rogan, W. R. Green, and R. F. Ozols. 1983. 15 Characterization of a human ovarian carcinoma cell line (NIH:OVCAR-3) with androgen and estrogen receptors. Cancer research 43:5379-5389. 94. Viel, A., F. Giannini, E. Capozzi, V. Canzonieri, C. Scarabelli, A. Gloghini, and M. Boiocchi. 1994. Molecular mechanisms possibly affecting WT1 function in human ovarian tumors. International journal of cancer 57:515-521. 20 95. Findley, H. W., Jr., M. D. Cooper, T. H. Kim, C. Alvarado, and A. H. Ragab. 1982. Two new acute lymphoblastic leukemia cell lines with early B-cell phenotypes. Blood 60:1305-1309. 96. Krug, L. Μ., T. Dao, A. B. Brown, P. Maslak, W. Travis, S. Bekele, T. Korontsvit, V. Zakhaleva, J. Wolchok, J. Yuan, H. Li, L. Tyson, and D. A. 25 Scheinberg. 2010. WT1 peptide vaccinations induce CD4 and CD8 T cell immune responses in patients with mesothelioma and non-small cell lung cancer. Cancer Immunol Immunother 59:1467-1479. 97. Chaise, C., S. L Buchan, J. Rice, J. Marquet, H. Rouard, M. Kuentz, G. E. Vittes, V. Molinier-Frenkel, J. P. Farcet, H. J. Stauss, Μ. H. Delfau-Larue, and 30 F. K. Stevenson. 2008. DNA vaccination induces WT1-specific T-cell responses with potential clinical relevance. Blood *\ 12:2956-2964. 8270468_1 (GHMatters) P94551 .AU.1 5October2016 -73- 2014262282 05 Oct 2016 98. Zhang, L, J. R. Conejo-Garcia, D. Katsaros, P. A. Gimotty, M. Massobrio, G. Regnani, A. Makrigiannakis, H. Gray, K. Schlienger, Μ. N. Liebman, S. C. Rubin, and G. Coukos. 2003. Intratumoral T cells, recurrence, and survival in epithelial ovarian cancer. The New England journal of medicine 348:203-213. 5 99. Odunsi, K., and L. J. Old. 2007. Tumor infiltrating lymphocytes: indicators of tumor-related immune responses. Cancer Immun 7:3. 100. Sun, J. C., J. N. Beilke, and L. L. Lanier. 2010. Immune memory redefined: characterizing the longevity of natural killer cells. Immunological reviews 236:83-94. 10 101. Tam, Y. K., B. Miyagawa, V. C. Ho, and H. G. Klingemann. 1999.

Immunotherapy of malignant melanoma in a SCID mouse model using the highly cytotoxic natural killer cell line NK-92. Journal of hematotherapy 8:281 -290. 102. Korbelik, M., and J. Sun. 2001. Cancer treatment by photodynamic therapy 15 combined with adoptive immunotherapy using genetically altered natural killer cell line. International journal of cancer 93:269-274. 103. Arai, S., R. Meagher, M. Swearingen, H. Myint, E. Rich, J. Martinson, and H. Klingemann. 2008. Infusion of the allogeneic cell line NK-92 in patients with advanced renal cell cancer or melanoma: a phase I trial. Cytotherapy 10:625-20 632. 104. Yasukawa, Μ., H. Ohminami, S. Kaneko, Y. Yakushijin, Y. Nishimura, K. Inokuchi, T. Miyakuni, S. Nakao, K. Kishi, I. Kubonishi, K. Dan, and S. Fujita. 1998. CD4(+) cytotoxic T-cell clones specific for bcr-abl b3a2 fusion peptide augment colony formation by chronic myelogenous leukemia cells in a b3a2- 25 specific and HLA-DR-restricted manner. Blood 92:3355-3361.

105. Chmielewski, M., A. Hombach, C. Heuser, G. P. Adams, and H. Abken. 2004. T cell activation by antibody-like immunoreceptors: increase in affinity of the single-chain fragment domain above threshold does not increase T cell activation against antigen-positive target cells but decreases selectivity. J 30 Immunol 173:7647-7653. 8270468_1 (GHMatters) P94551 .AU.1 5October2016 -74- 2014262282 05 Oct 2016 106. Tomkinson, B., E. Robertson, and E. Kieff. 1993. Epstein-Barr virus nuclear proteins EBNA-3A and EBNA-3C are essential for B-lymphocyte growth transformation. Journal of virology 67:2014-2025. 107. Boissel, L. B., M. Van Etten, RA. and Klingemann, HG. 2009. Transfection of 5 NK Cells with mRNA or Lentivirus Expressing Chimeric Antigen Receptors

Results in Highly Efficient Killing of Lymphoid Malignancies and Compares Favorably with Monoclonal Antibody-Directed ADCC. In 51st ASH Annual Meeting and Exposition, Ernest N. Morial Convention Center, New Orleans, LA. 10 108. Mulder, A., C. Eijsink, M. J. Kardol, Μ. E. Franke-van Dijk, S. H. van der Burg, M. Kester, Doxiadis, II, and F. H. Claas. 2003. Identification, isolation, and culture of HLA-A2-specific B lymphocytes using MHC class I tetramers. J Immunol 171:6599-6603. 109. Mulder, A., C. Eijsink, M. G. Kester, M. E. Franke, M. J. Kardol, Μ. H. 15 Heemskerk, C. van Kooten, F. A. Verreck, J. W. Drijfhout, F. Koning,

Doxiadis, II, and F. H. Claas. 2005. Impact of peptides on the recognition of HLA class I molecules by human HLA antibodies. J Immunol 175:5950-5957. 110. Bansal, H., S. Bansal, M. Rao, K. P. Foley, J. Sang, D. A. Praia, R. K. Blackman, W. Ying, J. Barsoum, M. R. Baer, K. Kelly, R. Swords, G. E. 20 Tomlinson, M. Battiwalla, F. J. Giles, K. P. Lee, and S. Padmanabhan. 2010.

Heat shock protein 90 regulates the expression of Wilms tumor 1 protein in myeloid leukemias. Blood 116:4591-4599. 111. O'Reilly, R. J. 2008. Epstein-Barr virus sustains tumor killers. Nature medicine 14:1148-1150. 25 112. Milner, E., E. Barnea, I. Beer, and A. Admon. 2006. The turnover kinetics of major histocompatibility complex peptides of human cancer cells. Mol Cell Proteomics 5:357-365. 113. Vicent, S., R. Chen, L. C. Sayles, C. Lin, R. G. Walker, A. K. Gillespie, A. Subramanian, G. Hinkle, X. Yang, S. Saif, D. E. Root, V. Huff, W. C. Hahn, 30 and E. A. Sweet-Cordero. 2010. Wilms tumor 1 (WT1) regulates KRAS-driven 8270468_1 (GHMatters) P94551 .AU.1 50ctOber2016 -75- 2014262282 05 Oct 2016 oncogenesis and senescence in mouse and human models. The Journal of clinical investigation 120:3940-3952. 114. Maruo, S., B. Zhao, E. Johannsen, E. Kieff, J. Zou, and K. Takada. 2011. Epstein-Barr virus nuclear antigens 3C and 3A maintain lymphoblastoid cell 5 growth by repressing p16INK4A and p14ARF expression. Proceedings of the

National Academy of Sciences of the United States of America 108:1919-1924. 6270468_1 (GHMatters) P94551.AU.1 5 October 2016

Claims

1. An isolated antigen-binding protein or antigen-binding fragment thereof comprising one of: (A) an antigen binding region having the amino acid sequence of one of SEQ ID NOS: 5, 8, 10, 13, 14, 17, or 20; (B) an antigen binding region comprising a VH and a VL respectively, with amino acid sequences selected from SEQ ID NOs: (i) 24 and 25; (ii) 26 and 27; (iii) 28 and 29; (iv) 30 and 31; (v) 32 and 33; (vi)34 and 35 and (vii) 36 and 37; or (C) (i) the following three light chain (LC) complementarity determining regions (CDRs): (a) a LC CDR1 comprising the amino acid sequence of SEQ ID NO: 56; and (b) a LC CDR2 and CDR3 comprising respectively, the amino acid sequence of SEQ ID NOs: 58 and 65, 59 and 66, 60 and 67, 61 and 68, 61 and 69, 62 and 70 and 63 and 71; and (C) (ii) the following three heavy chain (HC) CDRs: (a) a HC CDR1 comprising the amino acid sequence of SEQ ID NO: 38; and (b) a HC CDR2 and CDR3 comprising respectively the amino acid sequence of one of SEQ ID NOs: 41 and 49, 42 and 50, 43 and 51,44 and 52, 45 and 53, 46 and 54 and 47 and 55.

2. The isolated antigen-binding protein of claim 1, wherein the isolated antigen-binding protein is an antibody.

3. The isolated antigen-binding protein of claim 1, wherein the antigenbinding protein is a chimeric antigen receptor (CAR).

4. The isolated antigen-binding protein of claim 2, wherein the antibody is a full-length antibody, a substantially intact antibody, a Fab fragment, a F(ab')2 fragment or a single chain variable fragment (scFv).

5. The isolated antigen-binding protein of claim 1, wherein the antigenbinding protein specifically binds to an epitope on an HLA/peptide complex.

6. The isolated antigen-binding protein of claim 5, comprising: (A) an antigen binding region having the amino acid sequence of SEQ ID NO: 5; (B) an antigen binding region comprising a VH and a VL respectively, with the amino acid sequences SEQ ID NOs: 24 and 25; or (C) (i) (a) a light chain(LC) complementarity determining region 1 (CDR1) comprising the amino acid sequence of SEQ ID NO: 56; (b) a LC CDR2 comprising the amino acid sequence of SEQ ID NO: 58; and (c) a LC CDR3 comprising the amino acid sequence of SEQ ID NO: 65; and (C) (ii) (a) a heavy chain (HC) complementarity determining region 1 (CDR1) comprising the amino acid sequence of SEQ ID NO: 39; (b) a HC CDR2 comprising the amino acid sequence of SEQ ID NO: 41; and (c) a HC CDR3 comprising the amino acid sequence of SEQ ID NO: 49.

7. The isolated antigen-binding protein of claim 6, wherein the peptide of the HLA/peptide complex has the amino acid sequence LLDFVRFMGV (SEQ ID NO:4).

8. The isolated antigen-binding protein of claim 1, comprising: (A) an antigen binding region having the amino acid sequence of SEQ ID NO: 8; (B) an antigen binding region comprising a VH and a VL respectively, with the amino acid sequences SEQ ID NOs: 26 and 27; or (C) (i) (a) a light chain(LC) complementarity determining region 1 (CDR1) comprising the amino acid sequence of SEQ ID NO: 56; (b) a LC CDR2 comprising the amino acid sequence of SEQ ID NO: 59; and (c) a LC CDR3 comprising the amino acid sequence of SEQ ID NO: 66; and (C) (ii) (a) a heavy chain (HC) complementarity determining region 1 (CDR1) comprising the amino acid sequence of SEQ ID NO: 38; (b) a HC CDR2 comprising the amino acid sequence of SEQ ID NO: 42; and (c) a HC CDR3 comprising the amino acid sequence of SEQ ID NO: 50.

9. The isolated antigen-binding protein of claim 1, comprising: (A) an antigen binding region having the amino acid sequence of SEQ ID NO: 10; (B) an antigen binding region comprising a VH and a VL respectively, with the amino acid sequences SEQ ID NOs: 28 and 29; or (C) (i) (a) a light chain(LC) complementarity determining region 1 (CDR1) comprising the amino acid sequence of SEQ ID NO: 56; (b) a LC CDR2 comprising the amino acid sequence of SEQ ID NO: 60; and (c) a LC CDR3 comprising the amino acid sequence of SEQ ID NO: 67; and (C) (ii) (a) a heavy chain (HC) complementarity determining region 1 (CDR1) comprising the amino acid sequence of SEQ ID NO: 38; (b) a HC CDR2 comprising the amino acid sequence of SEQ ID NO: 43; and (c) a HC CDR3 comprising the amino acid sequence of SEQ ID NO: 51.

10. The isolated antigen-binding protein of claim 5, wherein the peptide of the HLA/peptide complex has the amino acid sequence RLTRFLSRV (SEQ ID NO: 7).

11. The isolated antigen-binding protein of claim 1, comprising: (A) an antigen binding region having the amino acid sequence of SEQ ID NO: 13; (B) an antigen binding region comprising a VH and a VL respectively, with the amino acid sequences SEQ ID NOs: 30 and 31; or (C) (i) (a) a light chain(LC) complementarity determining region 1 (CDR1) comprising the amino acid sequence of SEQ ID NO: 56; (b) a LC CDR2 comprising the amino acid sequence of SEQ ID NO: 61; and (c) a LC CDR3 comprising the amino acid sequence of SEQ ID NO: 68; and (C) (ii) (a) a heavy chain (HC) complementarity determining region 1 (CDR1) comprising the amino acid sequence of SEQ ID NO: 38; (b) a HC CDR2 comprising the amino acid sequence of SEQ ID NO: 44; and (c) a HC CDR3 comprising the amino acid sequence of SEQ ID NO: 52.

12. The isolated antigen-binding protein of claim 1, comprising: (A) an antigen binding region having the amino acid sequence of SEQ ID NO: 14; (B) an antigen binding region comprising a VH and a VL respectively, with the amino acid sequences SEQ ID NOs: 32 and 33; or (C) (i) (a) a light chain(LC) complementarity determining region 1 (CDR1) comprising the amino acid sequence of SEQ ID NO: 56; (b) a LC CDR2 comprising the amino acid sequence of SEQ ID NO: 61; and (c) a LC CDR3 comprising the amino acid sequence of SEQ ID NO: 69; and (C) (ii) (a) a heavy chain (HC) complementarity determining region 1 (CDR1) comprising the amino acid sequence of SEQ ID NO: 38; (b) a HC CDR2 comprising the amino acid sequence of SEQ ID NO: 45; and (c) a HC CDR3 comprising the amino acid sequence of SEQ ID NO: 53.

13. The isolated antigen-binding protein of claim 1, comprising: (A) an antigen binding region having the amino acid sequence of SEQ ID NO: 17; (B) an antigen binding region comprising a VH and a VL respectively, with the amino acid sequences SEQ ID NOs: 34 and 35; or (C) (i) (a) a light chain(LC) complementarity determining region 1 (CDR1) comprising the amino acid sequence of SEQ ID NO: 56; (b) a LC CDR2 comprising the amino acid sequence of SEQ ID NO: 62; and (c) a LC CDR3 comprising the amino acid sequence of SEQ ID NO: 70; and (C) (ii) (a) a heavy chain (HC) complementarity determining region 1 (CDR1) comprising the amino acid sequence of SEQ ID NO: 38; (b) a HC CDR2 comprising the amino acid sequence of SEQ ID NO: 46; and (c) a HC CDR3 comprising the amino acid sequence of SEQ ID NO: 54.

14. The isolated antigen-binding protein of claim 5, wherein the peptide of the HLA/peptide complex has the amino acid sequence RIITSTILV (SEQ ID NO: 12).

15. The isolated antigen-binding protein of claim 1, comprising one of: (A) an antigen binding region having the amino acid sequence of SEQ ID NO: 20; (B) an antigen binding region comprising a VH and a VL respectively, with the amino acid sequences SEQ ID NOs: 36 and 37; or (C) (i) (a) a light chain(LC) complementarity determining region 1 (CDR1) comprising the amino acid sequence of SEQ ID NO: 56; (b) a LC CDR2 comprising the amino acid sequence of SEQ ID NO: 63; and (c) a LC CDR3 comprising the amino acid sequence of SEQ ID NO: 71; and (C) (ii) (a) a heavy chain (HC) complementarity determining region 1 (CDR1) comprising the amino acid sequence of SEQ ID NO: 38; (b) a HC CDR2 comprising the amino acid sequence of SEQ ID NO: 47; and (c) a HC CDR3 comprising the amino acid sequence of SEQ ID NO: 55.

16. The isolated antigen-binding protein of claim 5, wherein the peptide of the HLA/peptide complex has the amino acid sequence LLEEMFLTV (SEQ ID NO: 19).

17. The isolated antigen-binding protein of claim 5, wherein the HLA of the HLA/peptide complex is HLA-A2.

18. A fusion protein comprising the antigen-binding protein of any one of claims 1 to 17.

19. An isolated single-chain variable fragment (scFv) comprising one of the amino acid sequences set forth in SEQ ID NOS: 5, 8, 10, 13, 14, 17 and 20.

20. An isolated scFv comprising a VH and a VL linked by an amino acid spacer, wherein the VH and VL respectively comprise the amino acid sequences set forth in SEQ ID NOS: (i) 24 and 25; (ii) 26 and 27; (iii) 28 and 29; (iv) 30 and 31; (v) 32 and 33; (vi) 34 and 35; and (vii) 36 and 37.

21. An immunoconjugate comprising a first component which is the antigen-binding protein or fragment thereof of any one of claims 1 to 17.

22. The immunoconjugate of claim 21, comprising a second component having a second amino acid sequence.

23. The immunoconjugate of claim 22, further comprising a cytotoxin.

24. The immunoconjugate of claim 22, wherein the second component is a binding protein or antibody having a binding specificity for a target that is different from the HLA-peptide complex.

25. A bispecific antibody comprising the antigen-binding protein of claim 1 of any one of claims 1 to 17.

26. A pharmaceutical composition comprising the antibody binding protein of any one of claims 1 to 17.