Using methylcholanthrene-induced fibrosarcomas to study tumor immunology.

Mouse models of cancer are essential in furthering our understanding both of the mechanisms that drive tumor development and the immune response that develops in parallel, and also in providing a platform for testing novel anti-cancer therapies. The majority of solid tumor models available rely on the injection of existing cancer cell lines into naïve hosts which, while providing quick and reproducible model systems, typically lack the development of a tumor microenvironment that recapitulates those seen in human cancers. Administration of the carcinogen 3-methylcholanthrene (MCA), allows tumors to develop in situ, forming a tumor microenvironment with an established stroma and vasculature. This article provides a detailed set of protocols for the administration of MCA into mice and the subsequent monitoring of tumors. Protocols are also provided for some of the routinely used downstream applications that can be used for MCA tumors.

Comparison of peptide-major histocompatibility complex tetramers and dextramers for the identification of antigen-specific T cells.

Fluorochrome-conjugated peptide-major histocompatibility complex (pMHC) multimers are widely used for flow cytometric visualization of antigen-specific T cells. The most common multimers, streptavidin-biotin-based 'tetramers', can be manufactured readily in the laboratory. Unfortunately, there are large differences between the threshold of T cell receptor (TCR) affinity required to capture pMHC tetramers from solution and that which is required for T cell activation. This disparity means that tetramers sometimes fail to stain antigen-specific T cells within a sample, an issue that is particularly problematic when staining tumour-specific, autoimmune or MHC class II-restricted T cells, which often display TCRs of low affinity for pMHC. Here, we compared optimized staining with tetramers and dextramers (dextran-based multimers), with the latter carrying greater numbers of both pMHC and fluorochrome per molecule. Most notably, we find that: (i) dextramers stain more brightly than tetramers; (ii) dextramers outperform tetramers when TCR-pMHC affinity is low; (iii) dextramers outperform tetramers with pMHC class II reagents where there is an absence of co-receptor stabilization; and (iv) dextramer sensitivity is enhanced further by specific protein kinase inhibition. Dextramers are compatible with current state-of-the-art flow cytometry platforms and will probably find particular utility in the fields of autoimmunity and cancer immunology.

Highly prevalent colorectal cancer-infiltrating LAP⁺ Foxp3⁻ T cells exhibit more potent immunosuppressive activity than Foxp3⁺ regulatory T cells.

Although elevated CD4⁺Foxp3⁺ regulatory T cell (Treg) frequencies within tumors are well documented, the functional and phenotypic characteristics of CD4⁺Foxp3⁺ and CD4⁺Foxp3⁻ T cell subsets from matched blood, healthy colon, and colorectal cancer require in-depth investigation. Flow cytometry revealed that the majority of intratumoral CD4⁺Foxp3⁺ T cells (Tregs) were Helios⁺ and expressed higher levels of cytotoxic T-lymphocyte antigen 4 (CTLA-4) and CD39 than Tregs from colon and blood. Moreover, ∼30% of intratumoral CD4⁺Foxp3⁻ T cells expressed markers associated with regulatory functions, including latency-associated peptide (LAP), lymphocyte activation gene-3 (LAG-3), and CD25. This unique population of cells produced interleukin-10 (IL-10) and transforming growth factor-ß (TGF-ß), and was ∼50-fold more suppressive than Foxp3⁺ Tregs. Thus, intratumoral Tregs are diverse, posing multiple obstacles to immunotherapeutic intervention in colorectal malignancies.

The impact of regulatory T cells on carcinogen-induced sarcogenesis.

Burnet proposed in the 1950's that the immune system is engaged in identifying and destroying abnormal cancerous cells. This process, termed immune surveillance, has been at the centre of intense debate for decades. Results using immunodeficient mice lend support to the immune surveillance hypothesis. We surmised that immune surveillance would be hampered by the inhibitory effect of naturally occurring FoxP3(+) regulatory T cells, a population of T cells shown to be present at an increased frequency in a variety of human tumours. The carcinogen, methylcholanthrene was injected subcutaneously into mice and the steady development of fibrosarcomas was observed over approximately 200 days. These fibrosarcomas were strikingly infiltrated with FoxP3(+) regulatory T cells implying that these cells impinge upon immune-mediated rejection of the tumour. This was confirmed by partial ablation of FoxP3(+) regulatory T-cell activity, which resulted in a marked reduction in tumour incidence. The reduction of tumour incidence was ablated in mice that lacked interferon gamma. These data offer strong support for the concept of immune surveillance and indicate that this process is limited by the inhibitory effect of FoxP3(+) regulatory T cells.

Immune tolerance to hepatitis C virus acquired during engraftment of bone marrow transplant.

The CD4+ T-cell response appears to be important for clearance of hepatitis C virus (HCV) in the majority of individuals. We have recently described a series of human leucocyte antigen (HLA)-DR11-restricted T-cell epitopes derived from HCV proteins which enables distinct populations of memory CD4+ T cells to be detected and counted in all nonviraemic HCV subjects. We examined the case of an HLA-DR11+ recipient of a haematopoietic stem-cell transplant who was concurrently infected with HCV from an HLA-DR11+ donor sibling. An acute HCV hepatitis developed and was treated with type I interferon. After successful viral clearance, the recipient demonstrated a selective lack of HCV epitope-specific CD4+ T cells and absence of serological responses compared with the treated donor. The recipient had no evidence of any nonspecific immunosuppression. The subsequent effects of concurrent infection during immune reconstitution are not known in adult humans, but data from murine models suggest this can lead to a skewing of the T-cell repertoire because of thymic selection. From the above observations, it is plausible that the introduction of foreign viral antigen into the thymus may lead to subsequent acquired central tolerance.

Characterization of novel HLA-DR11-restricted HCV epitopes reveals both qualitative and quantitative differences in HCV-specific CD4+ T cell responses in chronically infected and non-viremic patients.

The CD4+ T cell response is of critical importance in determining the fate of many viral infections. Clearance of HCV has a strong association with the MHC class II antigen HLA-DR11 suggesting a key role for CD4+ T cells. We used an epitope-prediction program to identify multiple novel HLA-DR11-restricted epitopes derived from several HCV proteins. These epitopes then allowed us to explore the qualitative and quantitative aspects of specific CD4+ T cell responses in HLA-DR11+ patients. Irrespective of the time since viral clearance, all the non-viremic patients recognized four epitopes with a high frequency of IFN-gamma-producing memory CD4+ T cells. There appeared to be two subpopulations of memory cells, immediate "effector" memory cells (Th1 phenotype) and resting "central" memory cells (Th1/0). Chronically infected patients revealed an almost complete absence of HCV epitope-specific IFN-gamma-producing T cells. However, three of these epitopes induced IL-10 production (down-regulatory) raising the question as to whether these cells play a role in viral persistence. The frequency and phenotype of memory cells is likely to reflect the magnitude of the initial immune response, and suggests that a high frequency of IFN-gamma-secreting CD4+ T cells to multiple epitopes are important in clearance of HCV.

In vivo antigen challenge in celiac disease identifies a single transglutaminase-modified peptide as the dominant A-gliadin T-cell epitope.

Celiac disease (CD) is an increasingly diagnosed enteropathy (prevalence, 1:200-1:300) that is induced by dietary exposure to wheat gliadins (as well as related proteins in rye and barley) and is strongly associated with HLA-DQ2 (alpha1*0501, beta1*0201), which is present in over 90% of CD patients. Because a variety of gliadin peptides have been identified as epitopes for gliadin-specific T-cell clones and as bioactive sequences in feeding studies and in ex vivo CD intestinal biopsy challenge, it has been unclear whether a 'dominant' T-cell epitope is associated with CD. Here, we used fresh peripheral blood lymphocytes from individual subjects undergoing short-term antigen challenge and tissue transglutaminase-treated, overlapping synthetic peptides spanning A-gliadin to demonstrate a transient, disease-specific, DQ2-restricted, CD4 T-cell response to a single dominant epitope. Optimal gamma interferon release in an ELISPOT assay was elicited by a 17-amino-acid peptide corresponding to the partially deamidated peptide of A-gliadin amino acids 57-73 (Q65E). Consistent with earlier reports indicating that host tissue transglutaminase modification of gliadin enhances gliadin-specific CD T-cell responses, tissue transglutaminase specifically deamidated Q65 in the peptide of A-gliadin amino acids 56-75. Discovery of this dominant epitope may allow development of antigen-specific immunotherapy for CD.

Induction and exhaustion of lymphocytic choriomeningitis virus-specific cytotoxic T lymphocytes visualized using soluble tetrameric major histocompatibility complex class I-peptide complexes.

This study describes the construction of soluble major histocompatibility complexes consisting of the mouse class I molecule, H-2Db, chemically biotinylated beta2 microglobulin and a peptide epitope derived from the glycoprotein (GP; amino acids 33-41) of lymphocytic choriomeningitis virus (LCMV). Tetrameric class I complexes, which were produced by mixing the class I complexes with phycoerythrin-labeled neutravidin, permitted direct analysis of virus-specific cytotoxic T lymphocytes (CTLs) by flow cytometry. This technique was validated by (a) staining CD8+ cells in the spleens of transgenic mice that express a T cell receptor (TCR) specific for H-2Db in association with peptide GP33-41, and (b) by staining virus-specific CTLs in the cerebrospinal fluid of C57BL/6 (B6) mice that had been infected intracranially with LCMV-DOCILE. Staining of spleen cells isolated from B6 mice revealed that up to 40% of CD8(+) T cells were GP33 tetramer+ during the initial phase of LCMV infection. In contrast, GP33 tetramers did not stain CD8+ T cells isolated from the spleens of B6 mice that had been infected 2 mo previously with LCMV above the background levels found in naive mice. The fate of virus-specific CTLs was analyzed during the acute phase of infection in mice challenged both intracranially and intravenously with a high or low dose of LCMV-DOCILE. The results of the study show that the outcome of infection by LCMV is determined by antigen load alone. Furthermore, the data indicate that deletion of virus-specific CTLs in the presence of excessive antigen is preceded by TCR downregulation and is dependent upon perforin.

Use of eluted peptide sequence data to identify the binding characteristics of peptides to the insulin-dependent diabetes susceptibility allele HLA-DQ8 (DQ 3.2).

HLA-DQ8 (A1*0301, B1*0302) and -DQ2 (A1*0501, B1*0201) are both associated with diseases such as insulin-dependent diabetes mellitus and coeliac disease. We used the technique of pool sequencing to look at the requirements of peptides binding to HLA-DQ8, and combined these data with naturally sequenced ligands and in vitro binding assays to describe a novel motif for HLA-DQ8. The motif, which has the same basic format as many HLA-DR molecules, consists of four or five anchor regions, in the positions from the N-terminus of the binding core of n, n + 3, n + 5/6 and n + 8, i.e. P1, P4, P6/7 and P9. P1 and P9 require negative or polar residues, with mainly aliphatic residues at P4 and P6/7. The features of the HLA-DQ8 motif were then compared to a pool sequence of peptides eluted from HLA-DQ2. A consensus motif for the binding of a common peptide which may be involved in disease pathogenesis is described. Neither of the disease-associated alleles HLA-DQ2 and -DQ8 have Asp at position 57 of the beta-chain. This Asp, if present, may form a salt bridge with an Arg at position 79 of the alpha-chain and so alter the binding specificity of P9. HLA-DQ2 and -DQ8 both appear to prefer negatively charged amino acids at P9. In contrast, HLA-DQ7 (A1*0301, B1*0301), which is not associated with diabetes, has Asp at beta 57, allowing positively charged amino acids at P9. This analysis of the sequence features of DQ-binding peptides suggests molecular characteristics which may be useful to predict epitopes involved in disease pathogenesis.