Results gathered in this study suggest that a status of “immunoprivileged self” in tumors barricades specific Teff cells. This suggestion portends that it might be very difficult, if possible, to circumvent autoimmunity toxicity in a systemic immunotherapy against cancer, unless a substantial antigenic difference is identified between the tumor target and healthy tissue. Therefore, targeting immunoregulatory elements at the tumor site would be desirable. Indeed, local delivery of engineered dendritic cells secreting anti-CTLA4 antibodies promoted immunity against melanoma in

mice without eliciting autoimmunity [44]. A nexus of immunosuppressive elements evolved at the tumor site likely suppress self-antigen-specific T lymphocytes as well as bona fide tumor-specific

T cells. A subtle reduction of CTLA4 in Teff cells by RNAi silencing could substantially overcome the tumor barrier, suggesting see more a practical approach to enhance the efficacies of antigen-specific T cells for cancer therapies. Transgenic and knockout mouse models constructed for auto-immunity studies Inhibitor Library were transitioned to study autoimmune mechanisms in antitumor immunity. A detailed description of the use of these models in the current study is provided in a supplementary table (Supporting information Table 1). BDC2.5/NOD, Foxp3-deficient C57BL/6 (B6) and NOD, NOD.Foxp3DTR, Rag-deficient-BDC2.5/NOD, and CTLA4 shRNA (CTLA4KD7) and PL4 transgenic mice were described previously [24, 29, 34, 35, 45, 46]. CTLA4KD7 and PL4 mice were backcrossed onto B6 background for more than ten generations, and then crossed with BALB-neuT [36], FIR (Foxp3-IRIS-RFP “knockin”) mice [47], or OT1 transgenic line [33]. All animals were maintained in a specific pathogen-free barrier facility and the studies are approved by the Institutional Animal Care and Use Committee at the University of Miami. The NIT-1 insulinoma, EL4 lymphoma, and E.G7-OVA lymphoma cell Alanine-glyoxylate transaminase lines were obtained from ATCC (Manassas, VA, USA) and implanted subcutaneously

at 5 × 106/mouse for insulinoma and 5 × 105 for lymphoma. For the NIT-1 model, tumor burden was quantified by measuring blood glucose levels and tumor mass. The tumor and pancreas samples were fixed in formalin solution. Paraffin-embedded sections were stained with hematoxylin and eosin (H-E) and examined by microscopy. Scoring for pancreas pathology was determined as follows: 0, intact islet with no lymphocytes in the islet area; 1, lymphocytes within the vicinity of the islet, but no infiltration; 2, peripheral insulitic lesion; 3, near or complete destruction of the islet. Flow cytometry analyses were conducted with a standard procedure [29]. The cells were stained with fluorescent-antibody conjugates to determine cells phenotype.