WO2017173360A2 - Combination treatments directed toward programmed death ligand-1 (pd-l1) positive cancers - Google Patents

Abstract

Provided are methods of treating cancers using a sugar analog, including methods of sensitizing cancers to EGF inhibitor-, PD-1 inhibitor-, and/or PD-L1 inhibitor-based therapies by administering a sugar analog, such as, for example, 2-Deoxy-D-glucose (2-DG), F-Fucose, and 2-F-peracetyl-Fucose.

Description

DESCRIPTION

COMBINATION TREATMENTS DIRECTED TOWARD PROGRAMMED DEATH LIGAND-1 (PD-Ll) POSITIVE CANCERS

[0001] This application claims the benefit of U.S. Provisional Application Serial No.

62/316,178 filed on March 31, 2016, the entirety of which is incorporated herein by reference.

RELATED FIELDS

[0002] The present disclosure relates generally to the fields of medicine, molecular biology and oncology. More particularly, it concerns methods for characterizing and treating cancers.

BACKGROUND

[0003] Perpetuation of T-cell activation has drastically reshaped the treatment of a broad spectrum of malignant cancer. For instance, the development of ipilimumab, the first FDA approved checkpoint blockade targeting T-cell response made treating metastatic melanoma probable (Hodi, F.S. et al., 2010, NEJM, 363:711-723; Robert, C. et al., 2013, Clin. Cancer Res., 19:2232-2239; and Robert, C. et al, 2011, NEJM, 364:2517-2526). While the anti-cytotoxic T lymphocyte antigen-4 (CTLA-4) monoclonal antibody showed promising results in treating melanoma patients, a second-generation checkpoint inhibitors targeting either PD-1 or PD-Ll have demonstrated better clinical activity and safety in phase III clinical trials (Topalian, S.L. et al., 2012, NEJM, 366:2443-54; and Brahmer, J.R et al, 2012, NEJM, 366:2455-2465). Because PD-Ll also possesses oncogenic potential that induces cancer cells progression (Topalian, S.L. et al., Id. ; Page, D.B. et al, 2014, Ann. Rev. Med. , 65: 185-202), in addition to its immunosuppression activity, targeting the PD-1/PD-L1 interaction provides dual efficacy by blocking immunosuppression via PD-1 while reducing cell progression via PD-Ll and is expected to have more sensitive outcome (Topalian, S.L. et al., Id ; Brahmer, J.R. et al, Id ; and Hamid, O., 2013, NEJM, 369: 134-144). The US FDA has approved two anti-PD-1 therapeutic antibodies for treatment of certain cancers: KEYTRUDA^® (pembrolizumab) and OPDIVO^® (nivolumab). However, the pathophysiological function and regulatory mechanism of PD-Ll remains incompletely defined.

[0004] Reawakening silenced immune response, particularly effector T-cells, has been recently added to a repertoire of treatment options after surgical removal, chemotherapy, radiotherapy, and targeted therapies. While the use of anti-CTLA-4 monoclonal antibody (Dunn et al., 2002, Nature Immunology, 3:991-998; and Leach et al, 1996, Science, 271: 1734-36) initially demonstrated success in treating metastatic melanoma, it has been shown to also induce an autoimmune response. Unlike anti-CTLA-4 antibodies, which affect only immune cells, anti-PD-Ll antibodies and anti-PD-1 antibodies act at a cellular level and at tumor sites to block the interaction between PD-1 -expressing effector T-cells and PD-Ll - expressing tumor cells. This creates a dual impact from both the tumor cell and the T-cell, thereby limiting the adverse effects and providing better therapeutic efficacy (Okazaki, T. et al., 2013, Nature Immunology, 14: 1212-1218). There remains a need for new and more effective therapeutics and methodologies that successfully target the PD-1/PD-L1 pathway and activate effector cells of the immune system to attack the tumor cells and treat cancers.

SUMMARY

[0005] The inventors have found that glycosylation of PD-Ll (also known as CD274,

PDCD1L1, or B7-H1) expressed on tumor cells potentiates or enhances binding to PD-1 on immune effector cells, such as T cells. PD-1/PD-L1 binding in vivo suppresses the immune response against the tumor cells. Disrupting the interaction between PD-Ll on the tumor cell and PD-1 on the immune effector cell prevents the immune suppression such that immune cells are active against the tumor cells, resulting in tumor cell killing. The present inventors have found that certain sugar analogs that are inhibitors of sugar transferring enzymes (also synonymously called sugar transferases or glycosyltransferases), for example, hexokinase or fucosyltransferase, prevent formation of the N-glycan structures found on the PD-Ll protein and inhibit the glycosylation of PD-Ll expressed on tumor cells. Reducing or eliminating the glycosylation of PD-Ll inhibits or reduces its interaction with PD-1 on T cells and thwarts the immunosuppressive activity caused by PD-1 -expressing T cells binding to PD-Ll on tumor cells, thus making the tumor cells susceptible to killing by the T cells which can participate in an active immune response against the tumor. Glycosylation of PD-Ll may also stabilize its expression on the surface of tumor cells. In the absence of glycosylation, the PD-Ll is more rapidly internalized and degraded in the tumor cell, resulting in less PD-Ll expressed on the tumor cell surface, and, in turn, less cell surface-expressed PD-Ll available to interact with PD-1 and mediate immunosuppression. Administration of these sugar analogs may potentiate and enhance the anti-cancer activity of anti-PD-1 and/or PD-Ll agents or other chemo therapeutic agents. [0006] Accordingly, provided herein are methods of treating a subject having a cancer, comprising administering to the subject an effective amount of a compound that is an inhibitor of a sugar transferring enzyme (also called a glycosyltransferase or sugar transferase). Subjects are preferably human but may also include non-human subjects such as companion animals and livestock. Examples of sugar transferring enzymes include, but are not limited to, hexokinase and fucosyltransferase. In one embodiment, inhibitors of sugar transferring enzymes include 2-Deoxy-D-glucose (2-DG), 2-F-Fucose and 2-F-peracetyl- Fucose. In a specific embodiment the inhibitor of the sugar transferring enzyme is not 2-DG. The cancer is preferably positive for glycosylated PD-L1. "Treating" or "treatment" includes treating, preventing, reducing the incidence of, ameliorating symptoms of, or providing a therapeutic benefit, and, in the context of cancer, includes reducing, preventing, or inhibiting tumor cell proliferation or killing of tumor or cancer cells, reducing tumor size, inhibiting or preventing metastasis and/or the invasiveness of a tumor, and preventing the spread or recurrence of a tumor or cancer. In certain aspects, the PD-L1 -positive cancer expresses PD- LI having elevated glycosylation, N-linked glycosylation, or N-glycosylation relative to a control cell. In further aspects, the PD-L1 is glycosylated or has elevated glycosylation at positions N35, N192, N200 and/or N219 of human PD-L1 protein (as set forth in SEQ ID NO: 1).

[0007] Particular embodiments relate to methods of treating cancer with combinations of an inhibitor of a sugar transferring enzyme (sugar transferase) and one or more of a PD-1 inhibitor or a PD-L1 inhibitor. Provided are methods of treating cancer in a subject, preferably a human subject, in need thereof, comprising administering (a) an effective amount of an inhibitor of a sugar transferring enzyme and (b) an effective amount of a PD-L1 inhibitor or a PD-1 inhibitor. In embodiments, the sugar transferring enzyme inhibitor is a fucosyltransferase inhibitor, such as 2-DG, 2-F-Fucose, or 2-F-peracetyl-Fucose. The PD-L1 inhibitor is preferably an antibody that binds PD-L1, preferably glycosylated PD-L1, and, particularly is an antibody, that blocks the binding of PD-L1 to PD-1. The PD-1 inhibitor is preferably an antibody that binds PD-1 and blocks the binding of PD-1 to PD-L1. In other embodiments, the sugar analog may be administered in combination with a CTLA4 inhibitor, such as ipilimumab.

[0008] In particular aspects, methods of treatment are provided which involve combinations of the sugar transferring enzyme inhibitor, particularly the hexokinase or fucosyltransferase inhibitor, and one or more targeted anti-cancer agents, including for example, tyrosine kinase inhibitors for cancers positive for a tyrosine kinase, such as EGF (or other growth factor) receptor. In some aspects, the method comprises administering to a subject in need thereof, e.g., suffering from cancer, (a) an effective amount of a sugar transferase inhibitor and (b) an effective amount of an EGF pathway inhibitor to the subject. In certain aspects, the EGF pathway inhibitor is a tyrosine kinase inhibitor. In particular aspects, the EGF pathway inhibitor is gefitinib, erlotinib, lapatinib, cetuximab, icotinib or AG 1478. In certain aspects, the sugar transferase inhibitor is a hexokinase or fucosyltransferase inhibitor, and, in particular aspects, is 2-DG, 2-F-Fucose, or 2-F-peracetyl- Fucose. In a particular embodiment, the sugar transferase inhibitor is not 2-DG. A cancer cell positive for a tyrosine kinase receptor, such as a growth factor receptor, is one that expresses tyrosine kinase receptors on the cell surface, wherein such receptors are bindable by ligands such as growth factors.

[0009] In certain aspects, the method comprises administering, in combination, at least a sugar transferring enzyme/glycosyltransferase inhibitor with a PD-1 or PD-L1 inhibitor and a targeted cancer therapeutic, such as a tyrosine kinase receptor inhibitor. In certain embodiments, the sugar transferring enzyme inhibitor is a hexokinase or fucosyltransferase inhibitor. In certain embodiments, the tyrosine kinase receptor inhibitor is preferably an inhibitor of EGFR. Thus, a method is provided for treating cancer in a subject comprising administering (a) an effective amount of a sugar transferring enzyme inhibitor; (b) an effective amount of a PD-1 inhibitor or a PD-L1 inhibitor; and (c) an effective amount of an EGF pathway inhibitor to the subject. In embodiments, a hexokinase or fucosyltransferase inhibitor is administered to the subject in step (a). In specific embodiments, the hexokinase or fucosyltransferase inhibitor is 2-DG, 2-F-Fucose, or 2-F- peracetyl-Fucose.

[0010] In certain embodiments, the PD-1 or PD-L1 inhibitor is an antibody that is directed against PD-1 and binds to PD-1, or is an antibody that is directed against PD-L1 and binds PD-L1. For example, the antibodies against PD-L1 could be atezolizumab, durvalumab, or avelumab, and the antibodies against PD-1 could be nivolumab, pembrolizumab, or pidilizumab. In certain embodiments, the PD-1 or PD-L1 inhibitor is an antibody that preferentially binds to glycosylated PD-L1 as compared to unglycosylated PD- Ll and inhibits binding of PD-1 to PD-L1. In certain embodiments, the PD-1 or PD-L1 inhibitor is an antibody that preferentially binds to glycosylated PD-1 as compared to unglycosylated PD-1 and inhibits binding of PD-1 to PD-L1. Nonlimiting examples of antibodies that specifically and preferentially bind glycosylated PD-L1 compared to non- glycosylated PD-L1 are STM004, STM115, STM073 and STM108, or humanized or chimeric forms thereof, as described in co-pending PCT Application No. PCT/US 16/24691 and Provisional Patent Application No. 62/314,652, the contents of which are hereby incorporated by reference in their entireties.

[0011] In certain aspects, the EGF pathway inhibitor is a receptor tyrosine kinase inhibitor. In particular aspects, the EGF pathway inhibitor is gefitinib, erlotinib, lapatinib, cetuximab, icotinib or AG1478.

[0012] In some aspects, the subject has, without limitation, a breast cancer, lung cancer, head & neck cancer, prostate cancer, esophageal cancer, tracheal cancer, skin cancer, brain cancer, liver cancer, bladder cancer, stomach cancer, pancreatic cancer, ovarian cancer, uterine cancer, cervical cancer, testicular cancer, colon cancer, rectal cancer, gall bladder cancer, or skin cancer. In certain aspects, the subject has a melanoma, non- small-cell lung cancer (NSCLC), or renal cell carcinoma (RCC). In particular aspects, the subject has a metastatic cancer.

[0013] In certain aspects, the agents are administered concurrently or sequentially.

[0014] A further embodiment provides a method of sensitizing a subject to PD-1 inhibitor and/or PD-L1 inhibitor therapy, comprising administering to the subject an effective amount of a sugar transferring enzyme inhibitor, e.g., a hexokinase or fucosyltransferase inhibitor, in particular, an effective amount of 2-DG, 2-F-Fucose, or 2-F-peracetyl-Fucose. A further embodiment provides a method of sensitizing a subject to targeted cancer therapy, such as a EGF pathway inhibitor or a tyrosine kinase receptor inhibitor, comprising administering to the subject an effective amount of a sugar transferring enzyme inhibitor, e.g., a hexokinase or fucosyltransferase inhibitor, and, in particular, an effective amount of 2-DG, 2-F-Fucose, or 2-F-peracetyl-Fucose. In some aspects, the subject is resistant or refractory to PD-1 and/or PD-L1 inhibitor therapy or to the targeted cancer therapy.

[0015] Further provided are compositions comprising a sugar transferring enzyme inhibitor, e.g., a hexokinase or fucosyltransferase inhibitor, and a PD-1 or PD-L2 inhibitor. Also provided are compositions comprising a sugar transferring enzyme inhibitor, e.g., a hexokinase or fucosyltransferase inhibitor, and an EGF pathway inhibitor. The invention also provides compositions comprising a sugar transferring enzyme inhibitor, e.g., a hexokinase or fucosyltransferase inhibitor, a PD- 1 or PD-L2 inhibitor, and an EGF pathway inhibitor. [0016] In a further embodiment, there is provided a method of characterizing a sample comprising measuring the level of PD-Ll glycosylation in the sample. In some aspects, the method further comprises measuring the level of PD-Ll glycosylation at positions N35, N192, N200 and/or N219. In certain aspects, the sample is a biological sample from a cancer patient, and the level of PD-Ll glycosylation is used to characterize the cancer. By way of example, cancers whose component cells express higher levels of cell surface PD-Ll glycosylation compared to controls may be amenable to treatment with the methods as described herein.

[0017] In some embodiments, cancer cells may be treated by methods and compositions of the embodiments. Cancer cells that may be treated with cell targeting constructs according to the embodiments include, but are not limited to, cells from the bladder, blood, bone, bone marrow, brain, breast, colon, esophagus, gastrointestine, gum, head, kidney, liver, lung, nasopharynx, neck, ovary, cervix, prostate, skin, stomach, testis, tongue, or uterus. In addition, the cancer may specifically be of one of the following histological types, while not being limited to these: neoplasm, malignant; carcinoma; carcinoma, undifferentiated; giant and spindle cell carcinoma; small cell carcinoma; papillary carcinoma; squamous cell carcinoma; lymphoepithelial carcinoma; basal cell carcinoma; pilomatrix carcinoma; transitional cell carcinoma; papillary transitional cell carcinoma; adenocarcinoma; gastrinoma, malignant; cholangiocarcinoma; hepatocellular carcinoma; combined hepatocellular carcinoma and cholangiocarcinoma; trabecular adenocarcinoma; adenoid cystic carcinoma; adenocarcinoma in adenomatous polyp; adenocarcinoma, familial polyposis coli; solid carcinoma; carcinoid tumor, malignant; branchiolo- alveolar adenocarcinoma; papillary adenocarcinoma; chromophobe carcinoma; acidophil carcinoma; oxyphilic adenocarcinoma; basophil carcinoma; clear cell adenocarcinoma; granular cell carcinoma; follicular adenocarcinoma; papillary and follicular adenocarcinoma; nonencapsulating sclerosing carcinoma; adrenal cortical carcinoma; endometroid carcinoma; skin appendage carcinoma; apocrine adenocarcinoma; sebaceous adenocarcinoma; ceruminous adenocarcinoma; mucoepidermoid carcinoma; cystadenocarcinoma; papillary cystadenocarcinoma; papillary serous cystadenocarcinoma; mucinous cystadenocarcinoma; mucinous adenocarcinoma; signet ring cell carcinoma; infiltrating duct carcinoma; medullary carcinoma; lobular carcinoma; inflammatory carcinoma; Paget's disease, mammary; acinar cell carcinoma; adenosquamous carcinoma; adenocarcinoma w/squamous metaplasia; thymoma, malignant; ovarian stromal tumor, malignant; thecoma, malignant; granulosa cell tumor, malignant; androblastoma, malignant; Sertoli cell carcinoma; leydig cell tumor, malignant; lipid cell tumor, malignant; paraganglioma, malignant; extra-mammary paraganglioma, malignant; pheochromocytoma; glomangiosarcoma; malignant melanoma; amelanotic melanoma; superficial spreading melanoma; malignant melanoma in giant pigmented nevus; epithelioid cell melanoma; blue nevus, malignant; sarcoma; fibrosarcoma; fibrous histiocytoma, malignant; myxosarcoma; liposarcoma; leiomyosarcoma; rhabdomyosarcoma; embryonal rhabdomyosarcoma; alveolar rhabdomyosarcoma; stromal sarcoma; mixed tumor, malignant; mullerian mixed tumor; nephroblastoma; hepatoblastoma; carcinosarcoma; mesenchymoma, malignant; brenner tumor, malignant; phyllodes tumor, malignant; synovial sarcoma; mesothelioma, malignant; dysgerminoma; embryonal carcinoma; teratoma, malignant; struma ovarii, malignant; choriocarcinoma; mesonephroma, malignant; hemangiosarcoma; hemangioendothelioma, malignant; kaposi's sarcoma; hemangiopericytoma, malignant; lymphangiosarcoma; osteosarcoma; juxtacortical osteosarcoma; chondrosarcoma; chondroblastoma, malignant; mesenchymal chondrosarcoma; giant cell tumor of bone; Ewing's sarcoma; odontogenic tumor, malignant; ameloblastic odontosarcoma; ameloblastoma, malignant; ameloblastic fibrosarcoma; pinealoma, malignant; chordoma; glioma, malignant; ependymoma; astrocytoma; protoplasmic astrocytoma; fibrillary astrocytoma; astroblastoma; glioblastoma; oligodendroglioma; oligodendroblastoma; primitive neuroectodermal; cerebellar sarcoma; ganglioneuroblastoma; neuroblastoma; retinoblastoma; olfactory neurogenic tumor; meningioma, malignant; neurofibrosarcoma; neurilemmoma, malignant; granular cell tumor, malignant; malignant lymphoma; Hodgkin's disease; paragranuloma; malignant lymphoma, small lymphocytic; malignant lymphoma, large cell, diffuse; malignant lymphoma, follicular; mycosis fungoides; other specified non-Hodgkin's lymphomas; malignant histiocytosis; multiple myeloma; mast cell sarcoma; immunoproliferative small intestinal disease; leukemia; lymphoid leukemia; plasma cell leukemia; erythroleukemia; lymphosarcoma cell leukemia; myeloid leukemia; basophilic leukemia; eosinophilic leukemia; monocytic leukemia; mast cell leukemia; megakaryoblastic leukemia; myeloid sarcoma; and hairy cell leukemia. In specific embodiments, the cancer is a BLBC.

[0018] As used herein in the specification, "a" or "an" may mean one or more. As used herein in the claim(s), when used in conjunction with the word "comprising", the words "a" or "an" may mean one or more than one. [0019] The use of the term "or" in the claims is used to mean "and/or" unless explicitly indicated to refer to alternatives only or the alternatives are mutually exclusive, although the disclosure supports a definition that refers to only alternatives and "and/or." As used herein "another" may mean at least a second or more.

[0020] Throughout this application, the term "about" is used to indicate that a value includes the inherent variation of error for the device, the method being employed to determine the value, or the variation that exists among the study subjects.

[0021] As used herein, "effective amount" is an amount of a compound or composition that, when administered to a patient with cancer, will have the intended therapeutic effect, e.g., alleviation, amelioration, palliation or elimination of one or more manifestations of cancer in the patient. A therapeutic effect does not necessarily occur by administration of one dose, and may occur only after administration of a series of doses.

Thus, a therapeutically effective amount may be administered in one or more administrations.

The effective amount may also be therapeutically effective in the context of a combination therapy, even if the amount of the agent may not be therapeutically effective when administered alone.

[0022] As used herein, the term "programmed death ligand- 1" or "PD-L1 " refers to a polypeptide (the terms "polypeptide" and "protein" are used interchangeably herein) or any native PD-L1 from any vertebrate source, including mammals such as primates (e.g. , humans, cynomolgus monkey (cyno)), dogs, and rodents (e.g. , mice and rats), unless otherwise indicated, and, in certain embodiments, included various PD-L1 isoforms, related PD-L1 polypeptides, including SNP variants thereof.

[0023] An exemplary amino acid sequence of human PD-L1 (UniProtKB/Swiss-Prot:

Q9NZQ7.1 ; GI: 83287884), is provided below:

MRIFAVFIFM TYWHLLNAFT VTVPKDLYVV EYGSNMTIEC KFPVEKQLDL AALIVYWEME DKNIIQFVHG EEDLKVQHSS YRQRARLLKD QLSLGNAALQ ITDVKLQDAG VYRCMISYGG ADYKRITVKV NAPYNKINQR ILVVDPVTSE HELTCQAEGY PKAEVIWTSS DHQVLSGKTT TTNSKREEKL FNVTSTLRIN TTTNEIFYCT FRRLDPEENH TAELVIPELP LAHPPNERTH LVILGAILLC LGVALTFIFR LRKGRMMDVK KCGIQDTNSK KQSDTHLEET (SEQ ID NO: 1). In SEQ ID NO: 1, the amino terminal amino acids 1- 18 constitute the signal sequence of the human PD-Ll protein. Accordingly, the mature human PD-Ll protein consists of amino acids 19-290 of SEQ ID NO: 1.

[0024] Other aspects, features and advantages of the described embodiments will become apparent from the following detailed description and illustrative examples.

BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The following drawings form part of the present specification and are included to further demonstrate certain aspects and embodiments as described herein. The aspects and embodiments may be better understood by reference to one or more of these drawings in combination with the detailed description of specific embodiments presented herein.

[0026] FIGS. 1A-1H - PD-Ll is glycosylated in cancer cells. A. Expression of PD-

Ll protein in primary breast cancer patient samples. Box plots showing the expression level of PD-Ll protein in primary breast cancer patient samples (left). * indicates statistically significant by the Mann- Whitney test. Western blot analysis of PD-Ll in representative breast cancer patient samples (right). B. Western blot analysis of PD-Ll in four breast cancer, four melanoma and three lung and three colon cancer cells. C. Western blot analysis of PD-Ll expression in shCTRL and two independent shPD-Ll stable clones of A431 cells. PD-Ll was transient transfected into shPD-Ll#5 clone. D. Glycoprotein staining of purified PD-Ll protein with or without PNGase F treatment. Coomassie blue staining panel represents total amount of PD-Ll protein. The upper bands appear in lane 4 and 5 are from the loading of PNGase F. (-) Ctrl, a control for non-glycoprotein; (+) Ctrl, a control for glycoprotein. E. Glycosylation pattern of PD-L1-GFP, HA-PD-L1, and PD-Ll-Flag proteins. Cell lysates were treated with PNGase F and Endo H and analyzed by Western blot. F. GFP- tagged PD-Ll full length (WT), extracellular domain (ECD), or intracellular domain (ICD) was transiently expressed in 293T cells. Cells were then treated with or without 5 g/ml tunicamycin (TM) overnight. Protein expression of PD-Ll was examined using Western blot. G. Schematic diagram of various PD-Ll protein expression constructs used in this study. Full-length PD-Ll was separated into extracellular domain (ECD) and intracellular domain (ICD). SP, signal peptide, and TM, transmembrane domain. Four putative NXT motifs in the ECD domain are labeled in red. The numbers indicate amino acid positions. H. Western blot analysis of the protein expression pattern of PD-Ll WT and its NQ mutants. Non-glycosylated form in lane 14 indicates PD-Ll WT with overnight treatment with tunicamycin (TM). Black circle, glycosylated PD-Ll; arrowhead, non-glycosylated PD-Ll. [0027] FIGS. 2A-2G - Glycosylation stabilizes PD-Ll expression and is required for cancer cells-associated immunosuppression. A-B. Western blot analysis of PD-Ll protein in PD-Ll -Flag expressing 293T cells. Cells were treated with 20 mM cycloheximide (CHX) (A) and 5 μΜ MG132 (B), as indicated intervals and analyzed by Western blot. The intensity of PD-Ll protein was quantified by a densitometer. C. Protein stability of PD-Ll WT, N35Q, N192Q, N200Q, N219Q, and 4NQ (determined as described in A). Quantification of protein half-life of PD-Ll WT and four NQ mutants (bottom) by a densitometer. D. Interaction of PD-1 and PD-Ll proteins with or without TM or anti-PD-Ll antibody treatment. Confocal image shows bound PD-l/Fc fusion protein on membrane of PD-Ll WT expressing 293T cells (left). Quantification of bound PD-1 protein in PD-L1/PD- 1 interaction assay (right). E. Binding affinity of PD-1 with PD-Ll WT or 4NQ proteins. The lysate of PD-Ll WT or 4NQ expressing 293T cells were incubated with or without PD-l/Fc fusion protein and then PD-Ll proteins were immunoprecipitated with anti-Flag antibody analyzed by Western blot. F. Co-IP measuring the interaction of PD-1 and PD-Ll in PD-Ll WT or 4NQ expressing 293T cells. G. Levels of soluble IL-2 in co-culture of Jurkat T-cell and PD-Ll WT or 4NQ expressing cells. Cells were pretreated with MG132 prior to D, E, and G experiment. Black circle, glycosylated PD-Ll; arrowhead, non-glycosylated PD-Ll; TM, tunicamycin; * indicates statistically significant by Student's t test. All error bars are expressed as mean ±SD of 3 independent experiments. [0028] FIGS. 3A-3I - EGF signaling induces PD-Ll glycosylation. A. Western blot analysis of glycosylation of PD-Ll protein in BT549, A431, BT549-shCTRL, and BT549-shEGFR cells treated with 25 ng/ml EGF, 25 ng/ml IGF-1, 10 ng/ml HGF, 25 ng/ml FGF, and 100 nM TGF for 10 hr. B. Interaction of PD-1 and PD-Ll in EGF and/or TKI treated PD-Ll WT- and 4NQ-expressing cells. C. Levels of soluble IL-2 in co-culture of Jurkat T-cells and PD-Ll WT- or 4NQ-expressing cells with EGF and/or TKI treatment. D. Heatmap analysis of N-glycosyltransferase genes expression in BLBC using TCGA dataset.

E. qRT-PCR of N-glycosyltransferases mRNA expressions on EGF and/or TKI treated cells.

F. Venn diagram of N-glycosyltransferase expression in BLBC (D), upregulated N- glycosyl transferase genes by EGF (E), and positively correlated N-glycosyltransferase genes to EGFR in breast cancer TCGA dataset. G. Western blot analysis of PD-Ll protein in EGF and/or TKI treated BT549 and MDA-MB 468 shCTRL and shB3GNT3 cells. H. Interaction of PD-1 and PD-Ll in EGF and/or TKI treated BT549 shCTRL and shB3GNT3 cells. I. Representative images from IHC staining of EGFR, B3GNT3, PD-Ll, and Granzyme B in primary breast cancer patients. Black circle, glycosylated PD-Ll; arrowhead, non- glycosylated PD-Ll; TM, tunicamycin; TKI, gefitinib; * indicates statistically significant by Student's t test. All error bars are expressed as mean ±SD of 3 independent experiments.

[0029] FIGS. 4A-4I - Blocking PD-Ll glycosylation with 2-DG and TKI in BLBC cells. A. Western blot analysis of PD-Ll protein in the cells treated with several indicated inhibitors. 293T cells were transfected with PD-Ll and treated with 1 g/ml tunicamycin (TM), 5 g/ml swainsonine (SW), 50 g/ml castanospermine (CST), 10 mM 2- DG, 1 μΜ gefitinib, 1 μΜ lapatinib, 1 μΜ erlotinib, 1 μΜ cetuximab, and 1 μΜ AG1478. B. Western blot analysis of PD-Ll protein in 2-DG- and 3-BP-treated cells. C. Western blot analysis of PD-Ll protein in glucose-, pyruvate-, or glutamine-depleted condition. PD-Ll WT expressing cells were either cultured with or without glucose (25 mM, 5 mM, 1 mM, 0 mM), without pyruvate, or without glutamine. D. Flag-PD-Ll WT or 4NQ expressing cells were incubated with 20 μΜ 2-DG IRDye® 800 (green color) overnight. PD-Ll proteins were then immunoprecipitated with anti-Flag antibody and analyzed by Western blot (red color). Green color represents PD-Ll with 2-DG IRDye® 800 incorporated (middle). E-F. Western blot analysis of glycosylated PD-Ll protein (E) and levels of soluble IL-2 (F) in TKI and/or 2-DG treated BLBC or non-BLBC cells. G. Binding of soluble PD-1 to PD-Ll -expressing BT549 cells treated with TKI/2-DG and/or anti-PD-Ll antibody. H. Soluble IL-2 levels in PD-Ll expressing BT549 cells treated with TKI/2-DG and/or anti-PD-Ll antibody. I. A proposed model showing upregulation of B3GNT3 N-linked glycotransferase by EGFR induces PD-Ll glycosylation to facilitate PD-1 interaction, therefore promoting immunosuppression in T-cells. Immunosuppression can be inhibited by 2-DG/TKI and anti- PD-Ll antibody treatment. Black circle, glycosylated PD-Ll; arrowhead, non-glycosylated PD-Ll; TKI, gefitinib; * indicates statistically significant by Student's t test. All error bars are expressed as mean ± SD of 3 independent experiments.

[0030] FIGS. 5A-5D - Expression of PD-Ll protein in cancer cells. A. Western blot analysis of PD-Ll in lung cancer cells. B. Western blot analysis of PD-Ll in colon cancer cells. C. Western blot analysis of PD-Ll in breast cancer cells. D. Western blot analysis of PD-Ll in ovarian cancer cells. Black circle, glycosylated PD-Ll; arrow head, non-glycosylated PD-Ll.

[0031] FIGS. 6A-6D - PD-Ll is glycosylated in cancer cells. A. Western blot analysis of PD-Ll in cancer cells using different anti-PD-Ll antibodies. Four BLBC cells, HCC1937, SUM149, MB-231, and BT20, and two non-BLBC cells, MB-483 and MB-474, were selected to test the expression of PD-L1 using different antibodies. B. Dual-expression construct for Flag-PD-Ll and shRNA of PD-L1. C. Western blot analysis of PD-L1 in shCTRL and two independent shPD-Ll stable clones of MDA-MB-231 and A431 cells. D. Glycosylation pattern of PD-L1 protein in MDA-MB-231 and A431 cells. Cell lysates were treated with PNGase F and analyzed by Western blot. Black circle, glycosylated PD-L1; arrow head, non-glycosylated PD-L1.

[0032] FIGS. 7A-7E - Expression of glycosylated and non-glycosylated PD-L1 protein. A. Western blot analysis of PD-Ll-Myc, PD-Ll-Flag, and HA-PD-Ll proteins in tunicamycin (TM) treated cells. B. Western blot analysis of PD-Ll-GFP WT, ECD, and ICD proteins in tunicamycin (TM) treated cells. C. Western blot analysis of PD-Ll-Myc, PD-Ll- Flag, HA-PD-Ll, PD-Ll-GFP WT, ECD, and ICD proteins in tunicamycin (TM) treated cells. The intensity of glycosylated (black bar) or non-glycosylated PD-L1 (red bar) protein was determined by a densitometry quantification (bottom). D. The mean of the intensity of glycosylated (black bar) or non-glycosylated PD-L1 (red bar) protein obtained from the bottom of C. Error bars represent SD. E. Glycosylation pattern of PD-L1 protein in PD-L1 expressing HEK 293T cells. Cell lysates were treated with PNGase F or O-glycosidase and analyzed by Western blot. Black circle, glycosylated PD-L1; arrow head, non-glycosylated PD-L1.

[0033] FIGS. 8A and 8B - N-glycosylation sites of PD-L1 protein. A sequence alignment of the PD-L1 amino acid sequences from different species is shown. Four NXT motifs, N35, N192, N200, and N219 are highlighted in red, and two non-NXT motifs, N63 and N204, are highlighted in green. Red box, conserved NXT motif. FIG. 8A: Consensus sequence = SEQ ID NO: 42; Q9NZQ7 HUMAN = SEQ ID NO: 43; Q9EP73 MOUSE = SEQ ID NO: 44; D4AE25 RAT = SEQ ID NO: 44; C5NU1 l BOVIN = SEQ ID NO: 46; Q4QTK1 PIG = SEQ ID NO: 47; and F7DZ76 HORSE = SEQ ID NO: 48. FIG. 8B: Consensus sequence = SEQ ID NO: 49; Q9NZQ7 HUMAN = SEQ ID NO: 50; Q9EP73 MOUSE = SEQ ID NO: 51; D4AE25 RAT = SEQ ID NO: 52; C5NU11 BOVIN = SEQ ID NO: 53; Q4QTK1 PIG = SEQ ID NO: 54; and F7DZ76 HORSE = SEQ ID NO: 55.

[0034] FIGS. 9A-9H - LC-MS/MS-based identification of N-glycopeptides. LC-

MS/MS -based identification of N-glycopeptides corresponding to each of the four N- glycosylation sites, Ν35 (A and E), Ν192 (B and F), Ν200 (C and G), and Ν219 (D and H) of PD-L1 (from HEK 293 cells). The LC-MS profiles (A-D) were shown as spectra averaged over a period of elution time (as labeled in figures) when a representative subset of glycoforms were detected. For each N-glycosylation site, one representive HCD MS² spectrum (E-H) is shown to exemplify its identification based on detection of yl ion (tryptic peptide backbone carrying the GlcNAc attached to the N- glycosylated Asn), along with the b and y ions defining its peptide sequence. The cartoon symbols used for the glycans (see inset) conform to the standard representation recommended by the Consortium for Functional Glycomics: Additional Hex and HexNAc were tentatively assigned as either lacNAc (Gal- GlcNAc) or lacdiNAc (GalNAc-GlcNAc) extension from the trimannosyl core (Man3- GlcNAc2), which can either be core fucosylated or not. The sequences showing N-linked glycosylation sites in FIGS. 9E-9H are set forth in SEQ ID NOs.: 56-59, respectively.

[0035] FIGS. 10A-10H - PD-Ll glycosylation changes protein stability and immunosuppression function of PD-Ll. A-B. Western blot analysis of PD-Ll protein in PD-Ll-Flag expressing HEK 293T cells. Cells were treated with 20 mM cycloheximide (CHX) (A) and 5 μΜ MG132 (B), as indicated intervals and analyzed by Western blot. C. Western blot analysis of PD-Ll in tunicamycin (TM) treated A431 cells. Bottom panel shows a densitometry quantification of PD-Ll protein. D. Schematic diagrams of PD-Ll/PD-1 interaction assay. E. Confocal image shows membrane localized PD-Ll WT or 4NQ proteins. F. Membrane localization of PD-Ll WT or 4NQ proteins. After biotinylation of membrane localization of PD-Ll WT or 4NQ proteins, the biotinylated proteins were pull- downed by streptavidin agarose. Membrane localized PD-Ll WT or 4NQ proteins were examined by Western blot. The ratio of membrane bound PD-Ll WT or 4NQ protein, which were obtained from the densitometry quantification, is showed in bottom. G. The interaction of PD-1 and PD-Ll protein in PD-Ll WT, N35Q, N192Q, N200Q, N219Q, or 4NQ cells. All error bars are expressed as mean ±SD of 3 independent experiments. H. Schematic diagram of IL-2 ELISA assay. PD-Ll stable clones were co-cultured with PD-1 overexpressed Jurkat T-cell. IL-2 expression from Jurkat cells were measured using ELISA. Detailed methodology is described in Materials and Methods section. Black circle, glycosylated PD-Ll; arrow head, non-glycosylated PD-Ll

[0036] FIGS. 11A-11D - EGF signaling induces PD-Ll glycosylation. A. Western blot analysis of PD-Ll protein in EGF, TKI, and/or tunicamycin (TM) treated cells. B. Western blot analysis of PD-Ll protein in EGF treated vector control (pBABE puro), EGFR WT, or EGFR K721A (no kinase activity) expressing CHO cells. C. The glycosylation of PD-Ll protein on different concentration of EGF or tunicamycin treatment. A431 and HeLa cells were serum- starved with or without 5 g/ml tunicamycin overnight and then treated with the indicated concentration of EGF for 10 nr. D. The mRNA expression of PD-Ll in EGF and/or TKI treated BT549 cells using qRT-PCR. All error bars are expressed as mean ±SD of 3 independent experiments. Black circle, glycosylated PD-Ll; arrow head, non- glycosylated PD-Ll.

[0037] FIGS. 12A-12E - A subset of N-glycosyltransferase is correlated with increased EGFR expression in BLBC. A. EGFR mRNA expression is higher in basal subtype of breast cancer (BLBC) in Agilent microarray (left) and RNAseq_V2 (right) datasets. B. EGFR protein expression is higher in basal subtype of breast cancer (BLBC) in RPPA (reverse phase protein array; TCPA (The Cancer Proteome Atlas)) dataset. C. The correlation between N-glycosyltransferase genes and basal subtype of breast cancer. D. EGFR mRNA expression was highly correlated with B3GNT3 mRNA expression in breast cancer TCGA (The Cancer Genome Atlas; NIH) dataset. E. Kaplan-Meier overall survival curves of EGFR and B3GNT3 in NKI295 cohort.

[0038] FIG. 13 - EGF signaling induces PD-Ll glycosylation BLBC cells.

Western blot analysis of glycosylated PD-Ll and B3GNT3 proteins in basal like breast cancer (BLBC) and non-BLBC cells. Graphs depicting the quantification of protein expression are shown below the blot.

[0039] FIGS. 14A-14E - 2-DG and TKI block PD-Ll glycosylation in BLBC cells. A. Western blot analysis of PD-Ll protein in cells treated with several indicated EGFR inhibitors. MDA-MB436, Hela, and PD-Ll WT-expressing Hela cells were treated with 1 μg/ml tunicamycin (TM), 5 g/ml swainsonine (SW), 50 g/ml castanospermine (CST), 10 mM 2-DG, 1 μΜ gefitinib, 1 μΜ lapatinib, 1 μΜ erlotinib, 1 μΜ cetuximab, or 1 μΜ AG1478. B. A longer exposed image of FIG. 4B, showing Western blot analysis of PD-Ll protein in 2-DG and 3-BP treated cells. C. A longer exposed image of FIG. 4C, showing Western blot analysis of PD-Ll protein in glucose, pyruvate, or glutamine depleted conditions. PD-Ll WT expressing cells were either cultured with or without glucose (25 mM, 5 mM, 1 mM, 0 mM), without pyruvate, or without glutamine. D-E. Western blot analysis of PD-Ll protein in TKI- and/or 2-DG-treated BLBC (D) or other cancer (E) cells. Black circle, glycosylated PD-Ll; arrow head, non-glycosylated PD-Ll. [0040] FIGS. 15A-15G - 2-DG/gefitinib sensitizes anti-PD-1 antibody immunotherapy in BLBC cells in 4T1 syngeneic mouse model. A. Tumor growth of 4T1- luc cells in B ALB/c mice following treatment with 2-DG/gefitinib and/or anti-PD- 1 antibody. The treatment protocol is depicted above the images in (A). In vivo tumor growth of 4Tl-luc cells on days 3 and 15 was shown by bioluminescence imaging using IVIS100 in representative mice treated as indicated. B. Images and box plots showing the tumor size/volume in mice treated with 2-DG/gefitinib and/or anti-PD-1 antibody. Tumors were measured and dissected at the endpoint. n = 9 mice per group. C. Survival plots of mice bearing syngeneic 4Tl-luc derived tumors following treatment with 2-DG/gefitinib and/or anti-PD-1 antibody. Significance was determined by log -rank test. *p < 0.05 and **p < 0.001 ; n = 10 mice per group. D. Flow cytometry of CD8 marker on CD3+ T cells isolated from tumors of 2-DG/gefitinib and/or anti-PD-1 antibody treated mice. E. Intracellular cytokine staining of IFNy in CD8+ CD3+ T cell populations. Significance was determined by two-way ANOVA, with *p < 0.05 and **p < 0.001 ; n = 7 mice per group. F. 4T1 tumor mass sections were triple-stained with antibodies against PD-L1, CD8, and granzyme B (GB). Hoechst dye was used for nuclear staining. Magnified images and white allows show that GB is secreted from activated CD8+ cytotoxic T cells. Scale bar, 100 μιη or 50 μιη (magnified images). G. Western blot analysis of the protein expression pattern of PD-L1, granzyme B, and polio virus receptor (PVR) proteins in 4T1 tumor mass from animals treated with 2-DG/gefitinib and/or anti-PD-1 antibody. * indicates statistically significant by Student's t test. All error bars are expressed as mean +SD of 3 independent experiments.

[0041] FIGS. 16A-16D - Nontoxicity of combination treatment of 2-DG/gefitinib and anti-PDl antibody immunotherapy in an EMT6 tumored syngeneic mouse model.

A. A graph showing the effect of the treatments described in FIGS. 15A-15G on the body weight of mice. B. Results of liver and kidney function measurements from treated mice at the end of the experiments. C. The growth of EMT6 tumor cells in BALB/c mice following treatment with 2-DG/Gefitinib and/or anti-PD-1 antibody. The treatment protocol is as depicted in FIG. 15A above. Images and box plots showing the tumor size/volume in 2- DG/Gefitinib and/or anti-PD-1 antibody treated mice. Tumors were measured and dissected at the endpoint. (n = 8 mice per group). D. Intracellular cytokine staining of IFNy in CD8+ CD3+ T cell populations in the EMT6 syngeneic mouse model. Significance was determined by two-way ANOVA, with *p < 0.05 and **p < 0.001 ; n = 8 mice per group. Error bars represent mean +SD of 3 independent experiments. * indicates statistically significant by Student's t test.

[0042] FIG. 17 illustrates the chemical structures of Fucose, 2-F-Fucose, and 2-F- peracetyl-Fucose.

[0043] FIG. 18 - 2-DG, 2-F-Fucose, and 2-F-peracetyl-Fucose reduce glycosylation of PD-Ll. Shown in FIG. 18 are Western blot analyses of PD-Ll in BT-549 PD-Ll stable clones expressing either WT PD-Ll or a PD-Ll glycosylation variant (N35Q, N35/3NQ, N192/3NQ, N200/3NQ, or N219/3NQ) as described herein. Cells were either untreated (Mock, lane 1) or treated with the sugar transferring enzyme inhibitors 2-DG (10 mM, lane 2), 2-F-Fucose ("2-F-Fuc", 250 μΜ, lane 3), or 2-F-peracetyl-Fucose ("2-F-Pa-Fu", 500 μΜ, lane 4).

[0044] FIG. 19 - 2-DG, 2-F-Fucose, and 2-F-peracetyl-Fucose reduce PD-Ll interaction with PD-1. FIG. 19 shows a graph of PD-1 protein bound to PD-Ll over time (hr) following treatment of PD-Ll -expressing cells with 2-DG, 2-F-Fucose ("2-F-Fuc"), or 2- F-peracetyl-Fucose (2-F-Ac3-Fuc") versus untreated cells (control, "CTRL"). A reduction in PD-1 binding to PD-Ll is seen versus control in the cells treated with a sugar transferring enzyme inhibitor.

DESCRIPTION OF THE EMBODIMENTS

[0045] The inventors have determined and demonstrated important modifications of the PD-Ll protein in cancer cells that promote survival signaling and aid in cancer cell escape from T-cell immune surveillance. In particular, it was found that glycosylation of PD- Ll is crucial to the PD-Ll and PD-1 interaction and that limiting, altering, or removing glycosylation of the PD-Ll protein could enhance and promote killing by T effector cells of the immune system. Furthermore, the substitution of sugars (e.g., glucose or fructose) by sugar analogs such as 2-DG (2-deoxy-D-glucose), 2-F-Fucose, and/or 2-F-peracetyl-Fucose, which are inhibitors of sugar transferring or glycosyltransferase enzymes, such as hexokinase or fucosyltransferase, blocked PD-Ll glycosylation and reduced PD-Ll binding to PD-1. The reduction in PD-Ll binding to PD-1 reduced immune suppression and promoted immune function against tumor cells. The sugar analogs can potentiate the action of anti-PD-Ll and anti-PD-1 therapies against cancers, particularly cancers whose cells are positive for PD-Ll protein expression. The combination of anti-PD-Ll antibody treatment and the blockage of PD-Ll glycosylation by sugar analogs, e.g., 2-DG, 2-F-Fucose, and/or 2-F-peracetyl-Fucose, produced a much stronger T cell immune response against cancer/tumor cells. Thus, combined treatment of cancer with an inhibitor of PD-L1 glycosylation (e.g., a sugar analog, for example, without limitation, 2-DG, 2-F-Fucose, and/or 2-F-peracetyl-Fucose) and an inhibitor of the PD-l/PD-Ll pathway could significantly enhance the efficacy and benefits of such treatment.

[0046] In addition, sugar analogs, e.g., glucose and/or fructose analogs, combined with inhibitors of tyrosine kinase receptors (TKRs), e.g., EGFR inhibitors, potentiated the activity of the TKR inhibitors, e.g., EGFR inhibitors, in blocking PD-L1 glycosylation and PD-L1 binding to PD-1. Thus, the treatment of cancer with a combination of an inhibitor of PD-L1 glycosylation (e.g., a sugar analog) and an inhibitor of the EGF pathway, e.g., an EGFR inhibitor, provides significant enhancement of the efficacy of the cancer treatment. In an embodiment, the sugar analog is 2-DG, 2-F-Fucose, and/or 2-F-peracetyl-Fucose.

[0047] In other embodiments, PD-L1 glycosylation can be used as a biomarker to predict the ability of cancer cells to escape immune surveillance by cells of the immune system, such as effector T cells that have cytotoxic function.

EXAMPLES

[0048] The following examples are intended to demonstrate embodiments and preferred embodiments as described in the disclosure. Those of skill in the art will appreciate that the examples describe techniques that were discovered by the inventors to function well in the described embodiments, and thus can be considered to constitute preferred modes for their practice. However, it will also be appreciated that many changes may be made in the specific embodiments as disclosed and still obtain a like and/or similar result without departing from the spirit and scope of that which is described herein.

Example 1 - Materials and Methods

[0049] Cell culture, stable transfectants, and transfection. All cells were obtained from American Type Culture Collection (ATCC). These cells were grown in in DMEM/F12 or RPMI 1640 medium supplemented with 10% fetal bovine serum (FBS). PD-L1 stable transfectants in MDA-MB-468, BT549, and 293T cells were selected using puromycin (InvivoGen, San Diego, CA, USA). For transient transfection, cells were transiently transfected with DNA using SN liposomes²⁰ and lipofectamine™ 2000 (Life Technologies, Carlsbad, CA, USA). [0050] Generation of stable cells using lentiviral infection. The lentiviral-based shRNA (pGIPZ plasmids) used to knockdown expression of PD-Ll (Shen, J. et al., 2013, Nature, 497:383-387) was purchased from the shRNA/ORF Core Facility (UT MD Anderson Cancer Center). Based on knock-down efficiency of PD-Ll protein expression in MDA-MB- 231 or A431 cells, two shPD-Ll clones were selected for this study. The mature antisense sequences are as follows: TCAATTGTCATATTGCTAC (shPD-Ll #1, SEQ ID NO: 2), TTGACTCCATCTTTCTTCA (shPD-Ll #5, SEQ ID NO: 3). Using a pGIPZ-shPD-Ll/Flag- PD-L1 dual expression construct to knock down endogenous PD-Ll and reconstitute Flag- PD-Ll simultaneously, endogenous PD-Ll knock-down and Flag-PD-Ll WT- or 4NQ mutant-expressing cell lines were established. To generate lentivirus-expressing shRNA for PD-Ll and Flag-PD-Ll, 293T cells were transfected with pGIPZ-non-silence (for vector control virus), pGIPZ-shPD-Ll, or pGIPZ-shPD-Ll/ PD-Ll WT, or pGIPZ-shPD-Ll/ PD-Ll 4NQ mutant with FuGENE 6 transfection reagent. Twenty-four hours after transfection, the medium was changed, then the medium was collected at 24-hour intervals. The collected medium containing lentivirus were centrifuged to eliminate cell debris, and filtered through 0.45-μιη filters. Cells were seeded at 50% confluence 12 hours before infection, and the medium was replaced with medium containing lentivirus. After infection for 24 hours, the medium was replaced with fresh medium and the infected cells were selected with 1 μg/ml puromycin (InvivoGen).

[0051] Plasmids. Human PD-Ll clone was obtained from the shRNA/ORF Core

Facility (UT MD Anderson Cancer Center, Houston, TX, USA) and cloned into pCDH lentiviral expression vectors to establish PD-Ll-Flag or PD-Ll-Myc expression cell lines. In addition, it also cloned into pEGFP-Nl and pCMV-HA mammalian cell expression vectors for transient transfection. Using pCDH / PD-Ll-Flag expression vector as a template, PD- Ll-Flag NQ mutants (N35Q, N192Q, N200Q, N219Q, and 4NQ (N35Q/N192Q/N200Q/N219Q) were produced by performing a site direct mutagenesis (see Table 1 below). To create a pGIPZ-shPD-Ll/Flag-PD-Ll dual expression construct to knock down endogenous PD-Ll and reconstitute Flag-PD-Ll simultaneously, a shPD-Ll construct (shPD-Ll #5) which targets 3'-UTR region of PD-Ll mRNA was first selected. Next, the Flag-PD-Ll wild type (WT) or 4NQ mutant were cloned into pGIPZ-shPD-Ll (Thermo Scientific, Pittsburgh, PA, USA) which expressed shRNA for endogenous PD-Ll. All constructs were confirmed using enzyme digestion and DNA sequencing. Table 1. Primers for site direct mutagenesis.

Primers Sequences (5'to 3')

Forward gtggtagagtatggtagccaaatgacaattgaatgcaaa (SEQ ID NO: 4)

N35Q

Reverse tttgcattcaattgtcatttggctaccatactctaccac

(SEQ ID NO: 5)

Forward gagaggagaagcttttccaggtgaccagcacactgag (SEQ ID NO: 6)

N192Q

Reverse ctcagtgtgctggtcacctggaaaagcttctcctctc

(SEQ ID NO: 7)

Forward gaccagcacactgagaatccagacaacaactaatgagat (SEQ ID NO: 8)

N200Q

Reverse atctcattagttgttgtctggattctcagtgtgctggtc

(SEQ ID NO: 9)

Forward gagagaggagaagcttttccaagtgaccagcacactgaga

(SEQ ID NO: 10)

N219Q

Reverse tctcagtgtgctggtcacttggaaaagcttctcctctctc (SEQ ID NO: 11)

[0052] qRT-PCR assays were performed to measure the expression of mRNA (Shen,

J. et al., 2013, Nature, 497:383-387; and Chang, C.J. et al, 2011, Nature Cell Biology, 13:317-323), (see Table 2 below). Cells were washed twice with PBS and immediately lysed in QIAzol. The lysed sample was subjected to total RNA extraction using RNeasy Mini Kit (Qiagen, Hilden, Germany). To measure the expression of mRNA, cDNA was synthesized from 1 μg purified total RNA by Superscript III First-Strand cDNA synthesis system using random hexamers (Life Technologies) according to the manufacturer's instructions. qPCR was performed using real-time PCR machine (iQ5, BioRad, Hercules, CA, USA). All data analysis was performed using the comparative Ct method. Results were first normalized to internal control β-actin mRNA.

Table 2. Primers for qRT-PCR.

Gene Sequences (5' to 3')

Forward gcataacgaacctaaccctcag

(SEQ ID NO: 12)

B4GALT2

Reverse gcccaatgtccactgtgata

(SEQ ID NO: 13)

Forward gtaacctcagtcacctgcc

(SEQ ID NO: 14)

B4GALT3

Reverse attccgctccacaatctctg

(SEQ ID NO: 15)

Forward tcttcaacctcacgctcaag

(SEQ ID NO: 16)

B3GNT3

Reverse gtgtgcaaagacgtcatcatc

(SEQ ID NO: 17)

B3GAT1 Forward caccatcaccctcctttctattc Gene Sequences (5' to 3')

(SEQ ID NO: 18)

Reverse gaacaacaggtctgggatttct

(SEQ ID NO: 19)

Forward gccttttgccatcgacatg

(SEQ ID NO: 20)

B3GAT2

Reverse agtcagattcttgcatccctg

(SEQ ID NO: 21)

Forward caaggagagcattaggaccaag

(SEQ ID NO: 22)

ST6GAL1

Reverse ccccattaaacctcaggactg

(SEQ ID NO: 23)

Forward tcgtcatggtgtggtattcc

(SEQ ID NO: 24)

ST3GAL4

Reverse caggaagatgggctgatcc

(SEQ ID NO: 25)

Forward gaccgcactcatcttacacc

(SEQ ID NO: 26)

MAN2A2

Reverse ggaggttggctgaaggaatac

(SEQ ID NO: 27)

Forward tcccctgctttaaccatcg

(SEQ ID NO: 28)

MAN2B1

Reverse ttgtcacctatactggcgttg

(SEQ ID NO: 29)

Forward ctgagtgatggaacgagtgag

(SEQ ID NO: 30)

UGGT1

Reverse tagagatgaccagatgcaacg

(SEQ ID NO: 31)

Forward gagtccaacttcacggcttat

(SEQ ID NO: 32)

MGAT3

Reverse agtggtccaggaagacataga

(SEQ ID NO: 33)

Forward tgtgagggaaagatcaagtgg

(SEQ ID NO: 34)

MGAT5

Reverse gctctccaaggtaaatgaggac

(SEQ ID NO: 35)

Forward ccactgagttcgtcaagagg

(SEQ ID NO: 36)

MOGS

Reverse acttccttgccatctgtcac

(SEQ ID NO: 37)

Forward tggctcgctgataagttctg

(SEQ ID NO: 38)

GNPTAB

Reverse gtgagtctggtttgggagaag

(SEQ ID NO: 39)

Forward gcaaagacctgtacgccaaca

(SEQ ID NO: 40)

ACTB

Reverse tgcatcctgtcggcaatg

(SEQ ID NO: 41) [0053] Antibodies and chemicals. The following antibodies were used: Flag (F3165;

Sigma- Aldrich, St. Louis, MO, USA), Myc (11667203001 ; Roche Diagnostics, Indianapolis, IN, USA), HA (11666606001 ; Roche Diagnostics), PD-L1 (13684; Cell Signaling Technology, Danvers, MA, USA), PD-L1 (329702; BioLegend, San Diego, CA, USA,), PD- LI (GTX117446; GeneTex, Irvine, CA, USA), PD-L1 (AF156; R&D Systems, Minneapolis, MN, USA), PD-1 (ab52587; Abeam, Cambridge, MA, USA), B3GNT3 (abl90458; Abeam), B3GNT3 (18098-1-AP; Proteintech, Chicago, IL, USA,) Granzyme B (ab4059; Abeam), EGFR (4267; Cell Signaling Technology), cc-Tubulin (B-5-1-2; Sigma- Aldrich), β-Actin (A2228; Sigma-Aldrich). Epidermal growth factor (EGF), cycloheximide, tunicamycin, swainsonine, castanospermine, and 2-deoxy-glucose (2-DG) were purchased from Sigma- Aldrich. Gefitinib, erlotinib, lapatinib, cetuximab, and AG1478 were obtained from Calbiochem Corp (Billerica, MA, USA).

[0054] Immunoblot analysis, immunocytochemistry and immunoprecipitation.

Immunoblot analysis was performed as described previously (Lim, S.O. et al., 2008, Gastroenterology, 135:2128-2140; and Lee, D.F et al., 2007, Cell, 130:440-455). Image acquisition and quantification of band intensity were performed using Odyssey® infrared imaging system (LI-COR Biosciences, Lincoln, NE, USA). For immunoprecipitation (pulldown), the cells were lysed in buffer (50 mM Tris HCl, pH 8.0, 150 mJVI NaCl, 5 mM ethylenediaminetetraacetic acid (EDTA) and 0.5% Nonidet P-40 (NP-40)) and centrifuged at 16, 000 x g for 30 minutes to remove debris. Cleared lysates were subjected to immunoprecipitation with antibodies. To measure 2-DG incorporation in PD-L1 protein, cells were incubated with IRDye® 800CW 2-DG Optical probe (LI-COR Biosciences) for overnight, and then immunoprecipitation was performed. For immunocytochemistry, cells were fixed in 4% paraformaldehyde at room temperature for 15 minutes, permeabilized in 5% Triton X-100 for 5 minutes, and then stained using primary antibodies. The secondary antibodies used were anti-mouse Alexa Fluor 488 or 594 dye conjugate and/or anti-rabbit Alexa Fluor 488 or 594 dye conjugate (Life Technologies). Nuclei were stained with 4', 6- diamidino-2-phenylindole (DAPI blue) (Life Technologies). After mounting, the cells were visualized using a multiphoton confocal laser-scanning microscope (Carl Zeiss, Thornwood, NY, USA).

[0055] PD-L1 and PD-1 (PD-Ll/PD-1 ) interaction assay. To measure PD-1 and PD-

Ll proteins interaction, cells were fixed in 4% paraformaldehyde at room temperature for 15 minutes and then incubated with recombinant human PD-1 Fc protein (R&D Systems) for 1 hour. The secondary antibodies used were anti-human Alexa Fluor 488 dye conjugate (Life Technologies). Nuclei were stained with 4', 6-diamidino-2-phenylindole (DAPI blue) (Life Technologies). The fluorescence intensity of Alexa Fluor 488 dye was then measured using a microplate reader Synergy Neo (BioTeK, Winooski, VT, USA) and normalized to the intensity by total protein quantity. To take an image, after mounting, the cells were visualized using a confocal laser-scanning microscope (Carl Zeiss).

[0056] Co-culture experiments and IL-2 expression measurement. Co-culture of

Jurkat T cells and tumor cells and IL-2 expression measurement was performed as described previously (Sheppard, K.A. et al., 2004, FEBS Letters, 574:37-41). To analyze the effect of tumor cells on T cell inactivation, tumor cells were co-cultured with activated Jurkat T cells expressing human PD-1, which were activated with Dynabeads® Human T- Activator CD3/CD28 (Life Technologies). Co-cultures at 5: 1 (Jurkat : tumor cell) ratio were incubated for 12 or 24 hours. Secreted IL-2 level in medium were measured as described by the manufacturer (Human IL-2 ELISA Kits, Thermo Scientific).

[0057] Glycosylation analysis ofPD-Ll. To confirm glycosylation of PD-L1 protein, cell lysates were treated with PNGase F, Endo H, O-glycosidase (New England BioLabs, Ipswich, MA, USA) as described by the manufacturer. To stain glycosylated PD-L1 protein, purified PD-L1 protein was stained using the Glycoprotein Staining Kit (Peirce/Thermo Scientific) as described by the manufacturer.

[0058] Immunohistochemical staining of human breast tumor tissue samples.

Immunohistochemical (IHC) staining was performed as described previously (Lee, D.F et al., 2007, Cell, 130:440-455; Lo, H.W. et al., 2007, Cancer Res. , 67:9066-9076; and Chang, C.J. et al., 2011, Cancer Cell, 19:86-100). Briefly, tissue specimens were incubated with antibodies against PD-L1, EGFR, B3GNT3, or Granzyme B, and with a biotin-conjugated secondary antibody, and then incubated with an avidin-biotin-peroxidase complex. Visualization was performed using amino-ethylcarbazole chromogen. For statistical analysis, Fisher's exact test and Spearman rank correlation coefficient were used and a p-value less than 0.05 was considered statistically significant. According to histological scoring, the intensity of staining was ranked into four groups: high (score 3), medium (score 2), low (score 1), and negative (score 0).

[0059] Identification of N-glycopeptide. Purified His tagged PD-L1 protein was reduced with 10 mM dithiothreitol at 37°C for 1 hr, alkylated with 50 mM iodoacetamide in 25 niM ammonium bicarbonate buffer for 1 hr in the dark at room temperature, and then treated overnight with sequencing grade trypsin at an enzyme-to-substrate ratio of 1:50 at 37°C. The digested products were then diluted with formic acid to final concentration with 0.1% and further cleaned up by ZipTip C18 (Millipore) before LC-MS/MS analysis. LC- MS/MS data were acquired at the Academia Sinica Mass Spectrometry Facility at IBC. The peptide mixture was analyzed by nanospray LC-MS/MS on an Orbitrap Fusion Tribrid (Thermo Scientific) coupled to an UltiMate 3000 RSLCnano System (Dionex) with trap column Acclaim PepMap 100 (2 cm x 100 μιη i.d) (Dionex). Peptide mixtures were loaded onto a Acclaim PepMap RSLC 25 cm x 75 μιη i.d. column (Dionex) and separated at a flow rate of 500 nL/min using a gradient of 5% to 35% solvent B (100% acetonitrile with 0.1% formic acid) in 60 min. Solvent A was 0.1% formic acid in water. The parameters used for MS and MS/MS data acquisition under the HCD parallel with CID mode were: top speed mode with 3 s cycle time; FTMS: scan range (m/z) = 400-2000; resolution = 120 K; AGC target = 2 x 10⁵; maximum injection time (ms) = 50; FTMSn (HCD): isolation mode = quadrupole; isolation window = 1.6; collision energy (%) = 30 with stepped collision energy 5%; resolution = 30 K; AGC target = 5 ^χ 10⁴; maximum injection time (ms) =60; ITMSn (CID): isolation mode = quadrupole; isolation window = 1.6; collision energy (%) = 30; AGC target = 1 x 10⁴. Raw data was converted to Mascot generic format (MGF) by Proteome Discoverer 1.4. For glycopeptide identification, the HCD MS² data were searched using Byonic (version 2.0-25) with the following search parameters: peptide tolerance = 2 ppm; fragment tolerance = 6 ppm; missed cleavages = 1; modifications: carbamidomethyl cysteine (fixed), methionine oxidation (common 2), deamidation at N (rare 1). The glycopeptide hits suggested by Byonic were further checked manually by combining HCD and CID MS² results.

[0060] Statistical analysis. Data in bar graphs represents mean fold change relative to untreated or control groups with standard deviation of three independent experiments. Statistical analyses were performed using SPSS (Ver. 20, SPSS, Chicago, IL). The correlation between protein expression and BLBC subset was analyzed using Spearman's correlation and Mann- Whitney test. Student's t test was performed for experimental data. A P value < 0.05 was considered statistically significant.

Example 2 - Protein Expression Analysis of PD-L1

[0061] To unravel the underlying mechanism of PD-L1, the protein expression of PD-

Ll in human tumor tissues and cancer cell lines was examined. FIGS. 1A and IB and 5A- 5B illustrate protein expression in lung, breast, colon and ovarian cancer cell lines, and FIG. 6A shows different PD-Ll antibodies. It was observed that a majority of PD-Ll was detected at -45 kDa (black circle) but a smaller fraction at 33 kDa (arrowhead) also appeared. Knocking down PD-Ll by lentiviral short-hairpin RNA (shRNA) targeting either the coding sequence (shPD-Ll#l) or the 3'UTR (shPD-Ll#5) downregulated expression of both the 33- and 45-kDa form of PD-Ll (see FIG. 6B). Reconstitution of PD-Ll restored expression of both forms in the shPD-Ll#5 clone (see FIG. 1C; FIG. 6C shows the vector design). These results suggest that both bands are PD-Ll and that the higher molecular weight of PD-Ll may be a result of posttranslational modifications.

[0062] Glycosylated proteins frequently produce a heterogeneous pattern in the

Western blot, which resembles the higher molecular weight (-45 kDa) of PD-Ll. To test whether the pattern observed for PD-Ll corresponded to the glycosylated form, MDA-MB 231 and HeLa cells were treated with recombinant glycosidase (Peptide-N-Glycosidase F; PNGase F) to remove N-glycan structure and then subjected to Western blot analysis. As shown in FIG. 6D, a significant portion of the 45-kDa PD-Ll was reduced to 33-kDa upon PNGase F treatment. Consistently, positive staining of the glycan structure was observed in purified His-tagged PD-Ll, but not in the presence of PNGase F (FIG. ID). These results demonstrate that the higher molecular weight of PD-Ll is indeed the glycosylated form.

[0063] To recapitulate the expression of PD-Ll in cells, various overexpression constructs were produced to mimic endogenous expression. To avoid possible cleavage at the N-terminus signaling peptide, different tag sequences were fused at either the N- or C- terminus (FIGS. 7A-7B, top). Similar to the results from endogenous PD-Ll expression, transient transfection of all GFP-, HA-, Flag-, or Myc- tagged PD-Ll had a -15 kDa molecular-weight shift from its actual size on the Western blot (FIGS. IE and 7A-7B). In contrast to PNGase F treatment, which removes the all N-glycan structure on PD-Ll, addition of recombinant glycosidase, i.e., endoglycosidase H (Endo H), only partially reduced PD-Ll glycosylation, suggesting that complex type of N-linked glycan structures (containing both high mannose and hybrid types) exist predominantly on PD-Ll (Stanley, P. 2011, Cold Spring Harbor Perspectives in Biology, 3). Furthermore, glycosylation of PD-Ll was completely inhibited when cells were treated with N-linked glycosylation inhibitor, tunicamycin (TM) (FIGS. IF and 7A-7D), but not with O-glycosidase (FIG. 7E). Together, these results indicate that PD-Ll is extensively N-linked glycosylated in the cells tested (Heifetz, A. et al., 1979, Biochemistry, 18:2186-2192). Example 3 - Glycosylation Analysis

[0064] Western blot analysis using two PD-Ll -specific antibodies (anti-PD-Ll antibody and anti-hB7-Hl antibody) indicated that PD-Ll glycosylation occurred on the extracellular domain (ECD) of the PD-Ll protein, recognized by the anti-hB7-Hl antibody, but not the intracellular domain (ICD) of the protein, recognized by anti-PD-Ll antibody (FIGS. IF and 7C). To pinpoint the glycosylation sites, a sequence alignment of the PD-Ll amino acid sequences from different species was performed to search for evolutionarily conserved NXT motifs, a consensus N-glycosylation recognition sequence (Schwarz, F. and Aebi, M., 2011, Curr. Opin. Struc. Biol , 21:576-582). Consistent with the earlier prediction (Cheng, X. et al., 2013, /. Biol. Chem. , 288: 11771-11785; Vigdorovich, V. et al., 2013, Structure, 21 :707-717), four NXT motifs were identified (FIGS. 1G and 8). To confirm if these sequences were indeed glycosylated, the tryptic peptides of a purified human PD-Ll were analyzed by nano LC-MS/MS. Glycopeptides carrying complex type N-glycans were identified for each of the 4 N-glycosylation sites (FIGS. 9A-9H), consistent with the apparent resistance to Endo H treatment (FIG. IE). A series of asparagine (N) to glutamine (Q) substitutions was generated to determine the specific glycosylation site(s) on the PD-Ll protein. All four mutants, N35Q, N192Q, N200Q, and N219Q, exhibited a certain degree of reduction in glycosylation compared with the WT PD-Ll (FIG. 1H, lanes 2, 3, 4, and 5). No detectable differences in glycosylation were observed for the three non-NXT NQ PD-Ll mutants (FIG. 1H, lanes 11, 12, and 13). In addition, PD-Ll glycosylation was completely ablated in the PD-Ll 4NQ protein variant in which all four asparagines were mutated to glutamine as indicated by the absence of signals corresponding to the glycosylated form at 45 kDa (FIG. 1H, lane 10 and lane 14). Based on the crystal structure of PD-1/PD-L1 complex (Lin, D. Y. et al., 2008, PNAS USA, 105:3011-3016), these four glycosylation sites of PD-Ll (N35, N192, N200 and N219) are exposed on the surface of the protein. Mutation of PD-Ll glycosylation sites (PD-Ll 4NQ) did not affect the overall structure based on the prediction. These results suggest that PD-Ll exists exclusively as N-glycosylated glycoprotein in the cells and that all four NXT motifs are glycosylated.

[0065] The levels of glycosylated PD-Ll observed were significantly higher than those of the non-glycosylated form of PD-Ll (FIGS. 1A-1B and 7C-7D). Therefore, the inventors investigated whether glycosylation stabilizes the PD-Ll protein. To this end, the protein turnover rate was measured in the presence of the protein synthesis inhibitor cycloheximide (CHX) to determine the effect of glycosylation on PD-Ll protein stability. The turnover rate of Flag-tagged PD-Ll was much faster in TM- than in DMSO-treated HEK 293T cells (FIGS. 2A and 10A-10B). Similar results were observed for the endogenous PD- Ll in A431 cells (FIG. IOC). Under proteasome inhibition by MG132, the levels of non- glycosylated PD-Ll steadily increased over a 12-hour period (FIG. 2B). Consistently, non- glycosylated PD-Ll 4NQ was rapidly degraded with a half-life as short as 4 hours (FIG. 2C), similar to that of the non-glycosylated PD-Ll WT (FIG. 2A), suggesting that impaired glycosylation leads to degradation of PD-Ll via the 26S proteasome.

Example 4 - Binding Affinity of PD-Ll

[0066] PD-Ll is a key immune suppressor through its binding with PD-1 during cancer progression. Thus, the binding affinity of WT PD-Ll and glycosylation-deficient mutant PD-Ll was compared to PD-1. To this end, PD-Ll WT and 4NQ mutant PD-Ll were stably expressed in MDA-MB-468 and HEK-293T cells, and stable clones with similar amounts of PD-Ll WT and 4NQ expression were then incubated with recombinant PD-l/Fc fusion protein, followed by the addition of anti-human IgG (Fc specific) fluorescence conjugate for signal amplification (Cheng, X. et al., 2013, /. Biol. Chem., 288: 11771-11785), (FIG. 10D). While there were no significant changes in membrane localization between glycosylated and non-glycosylated PD-Ll (FIG. 10E (confocal image) and FIG. 10F (biotinylation pull-down)), a marked difference was observed in PD-1 binding on the cell membrane between stable transfectants treated with or without TM (FIG. 2D, quantitation shown on the right). Furthermore, ablation of PD-Ll glycosylation by TM treatment or expression of the 4NQ mutant reduced its association with PD-1 (FIGS. 2E and 10G). In vitro binding experiments also demonstrated the loss of the PD-Ll 4NQ/PD-1/Fc interaction, suggesting that glycosylation of PD-Ll enhances its association with PD-1 (FIG. 2F). Because immunosuppression is associated with cell-cell interaction involving cytotoxic T- cells expressing PD-1 and PD-Ll -expressing cells, a co-culture experiment was performed to measure the IL-2 expression levels, an indicator of T-cell immune response that is suppressed by the PD-Ll and PD-1 interaction, in Jurkat T-cells after exposure to PD-Ll stable clones (Yang, W. et al., 2008, Invest. Ophthalmol. & Visual Set , 49:2518-2525), (FIG. 10H). Consistently, the loss of PD-Ll glycosylation impaired its interaction with PD-1 and enhanced the release of soluble IL-2 (FIG. 2G). These results reveal that the integrity of PD- Ll glycophenotype is required for its immunosuppressive function. Example 5 - Signaling and Regulation of PD-Ll

[0067] To delineate which upstream signaling governs PD-Ll glycosylation, several cancer cell lines were treated with common oncogenic growth factors that often are linked to tumor malignancy. Among those examined, EGF, but not IGF-1, HGF, FGF, or TGF-β, strongly induced PD-Ll glycosylation in BT549 and A431 cells (FIG. 3A, top). The increase in EGF-induced PD-Ll glycosylation was suppressed by EGFR tyrosine kinase inhibitor (TKI) gefitinib or by knocking down EGFR in BT549 cells (FIGS. 3A and 11A). Consistently, increasing the expression of EGFR, but not kinase dead mutant K721A, in an EGFR-low background (CHO cells) induced PD-Ll glycosylation upon EGF treatment (FIG. 11B). In addition, in the presence of TM, EGF failed to induce either the non-glycosylation form of PD-Ll or PD-Ll mRNA expression, thus excluding that transcription and stability of non-glycosylation form of PD-Ll could contribute to the increase of glycosylated PD-Ll (FIGS. 11 A, C, and D). Together, the results suggest that EGF- mediated EGFR kinase activity stimulates PD-Ll glycosylation. Consistently, EGF-mediated EGFR kinase was required for the EGF-induced PD-Ll and PD-1 interaction (FIG. 3B) and suppression of the T-cell immune response (FIG. 3C).

[0068] To understand how EGF regulates PD-Ll glycosylation, the gene expression profile was analyzed using the TCGA data sets. EGFR was found to be highly expressed in basal like breast cancer (BLBC) using Agilent, RNAseq_V2 and RPPA data sets (FIGS. 12A- 12B). Upon examination of the gene expression profile of EGFR and 50 N- glycosyltransferases using the public TCGA and RNAseq_V2 data sets (Barretina, J. et al., 2012, Nature, 483:603-607), it was found that five N-linked glycosylation-related enzymes were most highly correlated with EGFR in BLBC as compared with other breast cancer types, such as luminal A and B (FIGS. 3D and 12C). To identify which N-linked glycosyltransferase is specifically upregulated by EGF, the inventors analyzed the mRNA expression levels of several glycosyltransferases in MDA-MB 468 and BT-549 BLBC cells under a 24-hour EGF treatment with or without gefitinib (FIG. 3E). Among three tested criteria (genes upregulated by EGF in BLBC cells; genes upregulated in BLBC patients; and genes positively correlated with EGFR), only B3GNT3 was highly expressed in the BLBC, which correlated strongly with EGFR level and responded to EGF kinase activity (FIGS. 3F and 12D). In addition, high EGFR and B3GNT3 expression in the breast cancer patient dataset showed poor survival in compared to those with low or no expression (FIG. 12E). Consistently, downregulation of B3GNT3 by shRNA in BT549 and MDA-MB-468 cells reduced both EGF- mediated PD-Ll glycosylation (FIG. 3G) and PD-1 interaction with PD- Ll (FIG. 3H). To further validate the pathological relevance of the identified mechanism, the expression of EGFR, B3GNT3, PD-Ll, and the cytotoxic T-cell activation indicator, granzyme B, was evaluated in human breast tumor specimens by immunohistochemical (IHC) staining. PD-Ll correlated positively with EGFR (P = 0.007) and B3GNT3 (P = 0.0001), but negatively with granzyme B (P = 0.043) (FIG. 31 and Table 3 below). B3GNT3 and PD-Ll expression were also higher in BLBC than in non-BLBC cell lines (FIG. 13). Together, the data suggest that EGFR signaling, which is frequently activated in BLBC, upregulates B3GNT3 gene expression to catalyze PD-Ll glycosylation, stabilize PD-Ll, and enhance its interaction with PD-1, so as to escape T-cell immune surveillance (FIG. 41).

Table 3. Relationships between expression of PD-Ll, EGFR, Granzyme B, and

B3GNT3 in surgical specimens of breast cancer

Expression of PD-Ll

- / + ++ /+++ Total P value

26 28 54

- / +

(45.6%) (25.2%) (32.1%)

31 83 114

EGFR ++ /+++ P = 0.007*

(54.4%) (74.8%) (67.9%)

57 111 168

Total

(100%) (100%) (100%)

50 92 142

- / +

(68.5%) (81.4%) (76.3%)

Granzyme 23 21 44

++ /+++ P = 0.043** B (31.5%) (18.6%) (23.7%)

73 113 186

Total

(100%) (100%) (100%)

58 29 87

- / +

(84.1%) (24.8%) (46.8%)

11 88 99

B3GNT3 ++ /+++ P = 0.0001*

(15.9%) (75.2%) (53.2%)

69 117 186

Total

(100%) (100%) (100%)

*Positive correlation between PD-Ll and EGFR was analyzed using the PASS Pearson Chi- Square test. A P value of less than 0.05 was set as the criterion for statistical significance. **Inverse correlation between PD-Ll and granzyme B was analyzed using the PASS Pearson Chi-Square test. A P value < 0.05 was set as the criterion for statistical significance.

[0069] In accordance with the studies described herein, an assessment of which compounds might affect PD-Ll glycosylation was carried out. To this end, HEK293-PD-L1 cells were treated with EGFR inhibitors (gefitinib, erlotinib, lapatinib, cetuximab, and AG1478), glycosylation inhibitors, such as swainsonine (SW) or castanospermine (CST) and the glycolysis inhibitor 2-deoxy-glucose (2-DG). While the addition of EGFR inhibitors and glycosylation inhibitors had little or no effect on pre-existing PD-Ll glycosylation (FIG. 4A, lanes 3 and 4 and 6-10), both glycosylated and non-glycosylated PD-Ll were inhibited by 2- DG (FIG. 4A, lane 5, and FIG. 14A). Interestingly, another glycolysis inhibitor, 3- bromopyruvate (3-BP) did not inhibit PD-Ll glycosylation (FIG. 4B), suggesting that inhibition of glycolysis per se is not the cause of blocking PD-Ll glycosylation. In addition, since the GlcNAc (N-acetylglucosamine), monosaccharide derivative of glucose exists in the N-glycan structure of PD-Ll protein (FIGS. 9A-H), it was tested whether a depletion of glucose inhibits glycosylation of PD-Ll protein. Similar to the effects of 2-DG, depletion of glucose, but not pyruvate or glutamine, inhibited PD-Ll glycosylation (FIG. 4C and FIG. 14C, with longer exposure). These data provide support for 2-DG as a glucose analog interfering with the glycan structure of PD-Ll, resulting in degradation of PD-Ll. To confirm this, fluorescent-labeled 2-DG (IRdye® 800 2-DG) was used to measure the conjugation of 2-DG on PD-Ll (FIG. 4D) which indeed showed that 2-DG was incorporated into WT PD-Ll, but not into 4NQ mutant PD-Ll. This result further suggests a novel cellular regulatory function of 2-DG in inhibiting the protein glycan structure formation.

[0070] Clinical trial results have revealed that patients with metastatic melanoma who receive immunotherapy and do not respond at the early stage show no benefit to this immune intervention. While it has been observed that most patients respond to targeted therapy at early treatment, the risk of drug resistance limits the therapeutic benefit. In view of this situation, a combination of targeted and immune therapy could provide both acute and durable therapeutic benefit. Since low concentrations of 2-DG reduced PD-Ll glycosylation, it was also investigated as to whether destabilization of PD-Ll by 2-DG reduced EGFR- induced T-cell immunosuppression by examining PD-Ll expression in BLBC, lung cancer, and colon cancer cells treated with gefitinib, 2-DG alone, or gefitinib plus 2-DG. The combination of the two drugs was found to downregulated PD-Ll glycosylation much more effectively (FIG. 4E, quantification of Western blot from FIGS. 14D-14E). This combination also enhanced IL-2 release (FIG. 4F) from Jurkat T-cells co-cultured with BLBC, but not luminal breast cancer cells, as luminal breast cancer cells express very low levels of EGFR. These observations suggest that destabilization of PD-Ll by metabolic restriction and TKI may enhance the therapeutic efficacy of anti-PD-Ll immune therapy. To test this hypothesis, cells were treated with 2-DG/gefitinib and with or without human anti- PD-L1 antibody. 2-DG/TKI significantly increased the immune response potential of anti- PD-L1 antibody by reducing the interaction of PD-L1 with PD-1 (FIG. 4G) and enhancing IL-2 expression in T-cells (FIG. 4H).

[0071] In examples, the inventors have revealed an important biological event of PD-

Ll in the response of cancer cells to survival signaling to escape T-cell immune surveillance. They found that glycosylation of PD-L1 is a key signature for PD-L1 and PD-1 interaction. Combined treatment of glucose substitution by 2-DG together with EGFR TKI effectively blocks PD-L1 glycosylation as demonstrated by the experiments described herein. In some instances, the effect may be synergistic. Furthermore, a combination of anti-PD-Ll antibody treatment and blockage of PD-L1 glycosylation demonstrated much stronger T cell immune response (FIG. 4H). These results and combination treatments offer advantages toward opening a new avenue for effective anti-cancer strategies and cancer treatments. In addition, these findings provide the basis to further explore glycosylation of PD-L1 as a novel biomarker to predict the ability of cancer cells to escape immune surveillance and offer a new direction to understand the effect of glycosylation on the immune checkpoint molecule blockade. In addition, an essential node of PD-L1 glycosylation in the EGF/EGFR/SJGNrJ PD-Ll/PD-l/IL-2 signaling axis in crosstalk between cancer cells and T-cells has been identified (FIG. 41).

Example 6 Effect of 2-DG, gefitinib and anti-PD-1 antibody treatment in a tumored mouse model

[0072] In vivo experiments using a 4T1 or EMT6 syngeneic tumor mouse model were performed to further validate the combination treatment of 2-DG, gefitinib and PD-Ll/PD-1 blockade immune therapy in animals. Results from these experiments suggest that treatment with 2-DG and with the tyrosine kinase receptor (EGFR) inhibitor gefitinib enhanced the efficacy of anti-PD-1 antibody treatment in the 4Tl-lucif erase (4T1-Luc) syngeneic BALB/c model. Tumor size was reduced and mouse survival improved in mice treated with 2-DG and gefitinib and in mice treated with anti-PD-1 antibody (FIGS. 15A-15D), with no changes in body weight (FIG. 16A) and with minimal cytotoxicity (FIG. 16B). A tumor-infiltrated, activated CD8+ T cell population also significantly increased in mice treated with 2-DG and gefitinib and in mice treated with anti-PD-1 antibody (FIGS. 15E, 15F) through blocking PD-L1 glycosylation. The effect of the drug treatment combination was also observed in an EMT6 syngeneic mouse model (FIG. 16C). Together, these data corroborated the finding that inhibition of PD-Ll glycosylation by 2-DG treatment and gefitinib treatment could sensitize cancer cells (such as BLBC) to PD-1 blockade treatment, and that such a treatment combination, or similar types of treatment combinations, could significantly increase the immune response potential of an anti-PD- 1 antibody.

Example 7 - Inhibition of PD-Ll Glycosylation by Sugar Analogs

[0073] The sugar analogs 2-F-Fucose and 2-F-peracetyl-Fucose were tested for their ability to inhibit glycosylation of the PD-Ll protein and for their effect on the PD-1/PD-L1 interaction. To assess the impact of the inhibitors on PD-Ll glycosylation, BT-549 PD-Ll stable clones expressing either wild-type PD-Ll or expressing glycosylation mutants of PD- LI, namely, a N35Q PD-Ll mutant, and mutants in which three of the four glycosylation sites had a glutamine (Q) for asparagine (N) substitution, namely N35/3NQ, N192/3NQ, N200/3NQ, and N219/3NQ forms of PD-Ll, were treated with 2-DG (10 mM), 2-F-Fucose (250 μΜ), or 2-F-peracetyl-Fucose (500 μΜ). The results are shown in FIG. 18, revealing that the glycosylation inhibitors reduced PD-Ll glycosylation.

[0074] To assess the impact of the sugar analog inhibitors on PD-Ll/PD-1 binding, cells expressing wild type PD-Ll were treated with 2-DG (10 mM), 2-F-Fucose (250 μΜ), or 2-F-peracetyl-Fucose (500 μΜ) and then the cells were assayed for binding to PD-1. FIG. 19 demonstrates that all three sugar analogs inhibited binding of PD-Ll to PD-1 compared to control cells that had not been treated with the sugar analogs.

[0075] The contents of all patents, published applications and publications as referenced herein are hereby incorporated by reference in their entireties.

Claims

What Is Claimed Is:

I. A method of treating a cancer in a subject in need thereof, said method comprising administering to the subject an effective amount of a sugar transferase inhibitor.

2. The method according to claim 1, further comprising administering to the subject a PD-1 or PD-L1 inhibitor.

3. The method according to any of claims 1 to 3, wherein the sugar transferase inhibitor is a sugar analog.

4. The method according to any of claims 1 to 3 wherein the sugar transferase inhibitor is not 2-deoxy glucose (2-DG).

5. The method according to claim 3, wherein the sugar transferase inhibitor is a hexokinase inhibitor or a fucosyltransferase inhibitor.

6. The method according to claim 5, wherein the hexokinase inhibitor or fucosyltransferase inhibitor is 2-deoxy glucose (2-DG), 2-F-Fucose, or 2-F-peracetyl-Fucose.

7. The method according to any one of claims 1 to 6, wherein the PD-1 or PD-L1 inhibitor is an anti-PD-1 antibody or an anti-PD-Ll antibody.

8. The method according to claim 7, wherein the subject has a PD-L1 positive cancer.

9. The method according to claim 8, wherein cells of the PD-L1 positive cancer expresses PD-L1 having elevated glycosylation relative to a control cell.

10. The method according to claim 9, wherein the PD-L1 has elevated glycosylation at positions N35, N192, N200 and/or N219.

II. The method according to claim 10, wherein the sugar transferase inhibitor prevents or blocks glycosylation of PD-L1 expressed by subject's PD-L1 positive cancer.

12. The method according to any of claims 1 to 11, wherein the cancer is an EGFR positive cancer.

13. The method according to claim 12, which further comprises administering to the subject an EGF pathway inhibitor.

14. The method according to claim 13, wherein the EGF pathway inhibitor is a tyrosine kinase receptor inhibitor.

15. The method according to claim 13, wherein the EGF pathway inhibitor is an antibody.

16. The method according to claim 13, wherein the EGF pathway inhibitor is gefitinib, erlotinib, lapatinib, cetuximab, icotinib, or AG1478.

17. A method of treating a cancer in a subject in need thereof, said method comprising administering to the subject an effective amount of (a) a sugar transferase inhibitor, and (b) an EGF pathway inhibitor.

18. The method according to claim 17, wherein the sugar transferase inhibitor is a sugar analog.

19. The method according to claim 17 or claim 18, wherein the sugar transferase inhibitor is a hexokinase inhibitor or fucosyltransferase inhibitor.

20. The method according to claim 19, wherein the hexokinase inhibitor or fucosyltransferase inhibitor is 2-DG, 2-F-Fucose, or 2-F-peracetyl-Fucose.

21. The method according to any one of claims 17 to 20, wherein the subject has a PD-L1 positive cancer.

22. The method according to claim 21, wherein cells of the PD-L1 positive cancer expresses PD-L1 having elevated glycosylation relative to a control cell.

23. The method according to claim 22, wherein the PD-L1 has elevated glycosylation at positions N35, N192, N200 and/or N219.

24. The method according to any one of claims 17 to 23, wherein the cancer is an EGFR positive cancer.

25. The method according to any one of claims 17 to 24, wherein the EGF pathway inhibitor is a tyrosine kinase receptor inhibitor.

26. The method according to any one of claims 17 to 24, wherein the EGF pathway inhibitor is an antibody.

27. The method according to any one of claims 17 to 24, wherein the EGF pathway inhibitor is gefitinib, erlotinib, lapatinib, cetuximab, icotinib or AG1478.

28. The method according to any one of claims 1 to 27, wherein the subject has a breast cancer, lung cancer, head & neck cancer, prostate cancer, esophageal cancer, tracheal cancer, skin cancer brain cancer, liver cancer, bladder cancer, stomach cancer, pancreatic cancer, ovarian cancer, uterine cancer, cervical cancer, testicular cancer, colon cancer, rectal cancer, gall bladder cancer, or skin cancer.

29. The method according to any one of claims 1 to 27, wherein the subject has a melanoma, non-small-cell lung cancer (NSCLC), renal cell carcinoma (RCC) or bladder cancer.

30. The method according to any one of claims 1 to 29, wherein the subject has a metastatic cancer.

31. The method according to any one of claims 1 to 30, wherein (a) is administered before or after the (b).

32. The method according to any one of claims 1 to 30, wherein (a) and (b) are administered concurrently.

33. The method according to any one of claims 1 to 30, comprising administering a composition comprising an effective amount of (a) and an effective amount of (b).

34. The method according to any one of claims 6 to 16 or 20 to 33, wherein one or more of 2-DG, 2-F-Fucose, and 2-F-peracetyl-Fucose is radiolabeled.

35. A method of sensitizing a PD-Ll positive cancer in a subject to a PD-1 inhibitor and/or to PD-Ll inhibitor therapy, said method comprising administering to the subject an effective amount of a sugar transferase inhibitor.

36. The method according to claim 34, wherein the sugar transferase inhibitor is a sugar analog.

37. The method according to claim 35 or claim 36, wherein the sugar analog is not 2-DG.

38. The method according to claim 35 or claim 36, wherein the sugar transferase inhibitor is a hexokinase inhibitor or fucosyltransferase inhibitor.

39. The method according to claim 38, wherein the hexokinase inhibitor or fucosyltransferase inhibitor is 2-DG, 2-F-Fucose, or 2-F-peracetyl-Fucose.

40. The method according to any one of claims 35 to 39, wherein the subject is resistant to a PD-1 inhibitor or to a PD-Ll inhibitor therapy.

41. A method of sensitizing a PD-Ll and EGFR positive cancer in a subject to EGF pathway inhibitor therapy and/or to PD-1 inhibitor therapy and/or to PD-Ll inhibitor therapy, said method comprising administering to the subject an effective amount of a sugar transferase inhibitor.

42. The method according to claim 39, wherein the sugar transferase inhibitor is a sugar analog.

43. The method of claim 41 or 42, wherein the sugar analog is not 2-DG.

44. The method according to claim 39 or claim 40, wherein the sugar transferase inhibitor is a hexokinase inhibitor or fucosyltransferase inhibitor.

45. The method according to claim 44, wherein the hexokinase inhibitor or fucosyltransferase inhibitor is 2-DG, 2-F-Fucose, or 2-F-peracetyl-Fucose.

46. The method according to any of claims 41 to 45, wherein the subject is resistant to EGF pathway inhibitor therapy, PD-1 inhibitor therapy or a PD-L1 inhibitor therapy.

47. A composition comprising (a) an effective amount of a sugar transferase inhibitor and (b) an effective amount of a PD-1 inhibitor or a PD-L1 inhibitor.

48. The composition according to claim 47, wherein the sugar transferase inhibitor is a sugar analog.

49. The composition according to claim 47 or claim 48, wherein the sugar transferase inhibitor is a hexokinase inhibitor or fucosyltransferase inhibitor.

50. The composition according to claim 49, wherein the hexokinase inhibitor or fucosyltransferase inhibitor is 2-DG, 2-F-Fucose, or 2-F-peracetyl-Fucose.

51. The composition according to any one of claims 47 to 50, further comprising an effective amount of EGF pathway inhibitor.

52. The composition according to claim 51, wherein the EGF pathway inhibitor is an EGFR inhibitor.

53. A method of characterizing a sample comprising measuring the level of PD-L1 glycosylation in the sample.

54. The method according to claim 53, further comprising measuring the level of PD-L1 glycosylation at positions N35, N192, N200 and/or N219 on the PD-L1 protein.

55. The method according to claim 53 or claim 54, wherein the sample is from a cancer patient and the level of PD-L1 glycosylation is used to characterize the cancer.