Felts, Sara J.

Abstract Description This study dissects mechanisms of spontaneous tumor rejection leading to better understanding of successful immune responses to cancer. Cancers persist despite expression of tumor-specific neoantigens promoting ongoing anti-tumor immune responses. While some occult tumor cells likely are cleared by effective anti-tumor immune responses, the targeted antigens are not easily identifiable as those cells spontaneously disappear. Congenic mouse breast carcinoma models expressing a defined protein mimicking neoantigens were used to dissect the protective spontaneous immune response. BALB/c (MHC H-2d) mice challenged with BALB/c breast tumors expressing the rat-erbB2 oncoprotein succumb to their tumors despite demonstrated tumor immunogenicity, mimicking known patient responses to cancer. Meanwhile, H-2b positive BALB.B and H-2d/H-2b F1 hybrid mice spontaneously eliminate genetically matched tumors in an MHC-II dependent manner. CD4+ T-cells and B-cells or the rat-erbB2-specific antibodies they produce are crucial mediators of this protective response. Differences in the tumor microenvironments in rat-erbB2+ tumor-bearing BALB.B and BALB/c mice at the time of fate decision in the models reflect the differences between effective and ineffective anti-tumor immunity. Identifying the nature of the optimal MHC-II antigens, how they activate protective CD4+ T-cells, and the interactions that promote protective antibodies will better inform future cancer therapies. Funding Sources Bernie Fineman Research Fund Mayo Clinic Center for Biomedical Discovery Mayo Clinic Department of Immunology Mayo Clinic Graduate School of Biomedical Sciences NIH Traning Grant 5 R25 GM 55252 Topic Categories Tumor Immunology: Cellular Responses and Tumor Microevironment (TIME)

Abstract Description Antigen receptor (CAR) T cells can be generated in vivo within the natural lymph node (LN) environment as the immune response to a virus infection unfolds. This in situ approach entails priming LNs with intradermal inoculations of replication defective adenoviruses encoding MHC class I and class II allo/xeno-antigens. Universal transplant antigens were devised for LN priming by introducing tryptophan into amino acid position 162 of the MHC HLA-A2:01 class I heavy chain and by using the mouse MHC class II beta chain to form chimeric class II antigens with endogenous class II alleles. The virus encoded mismatched MHC antigen presenting molecules are expressed functionally in the LN following intradermal inoculation provoking polyclonal T cells responses. Three days later CAR T cells are generated by directly injecting retroviruses encoding CARs of interest into the responding LN. Mouse CD19-specifc CAR T cells created in this manner in B6-DQ8 and DQ8/A2 mice effectively target endogenous B cells and are therapeutic in male and female Emu-myc lymphomas demonstrating the potential for the generating functional CAR T cells directly in vivo. While in situ CAR T cells are readily produced in B6, B6-IA0/DQ8, B6-IA0/DQ8/A2, and B6-IA0/DQ6 LNs, functional differences among the strains emerged in the periphery suggesting that properties of the autoimmune-associated HLA DQ8 class II allele is particularly reactive during the production of CAR T cells in vivo. Funding Sources MustangBio, Inc Mayo Clinic Center of Biomedical Research Mayo Clinic Department of Immunology Mayo Clinic Graduate School of Biomedical Science Fraternal Order of Eagles Mayo Clinic Ventures NIH R21 AI 097337 NIH R01 CA 178804 NIH T32 AI 07425 Topic Categories Tumor Immunology: Checkpoints, Prevention, and Treatment (TIPT)

Background Cancers persist despite expression of immunogenic neoantigens and ongoing antitumor immune responses. While some occult tumors likely are cleared by effective antitumor immune responses, the targeted antigens are not easily identifiable as those tumors spontaneously disappear. Methods We used mouse models with a defined antigenic protein mimicking tumor-specific neoantigens to address the nature of these spontaneous anti-tumor immune responses. Results BALB/c (H-2d ) mice challenged with BALB/c breast tumors expressing the rat-erbB2 oncoprotein succumb to their tumors despite ongoing immune responses targeting tumor-specific model antigens. Meanwhile, congenic BALB.B (H-2b ) and H-2d/H-2b F1 hybrid mice spontaneously eliminate genetically matched tumors in a major histocompatibility complex (MHC)-II dependent manner. Adoptive transfer and immune cell depletion strategies revealed CD4+ T cells and CD20+ B cells are crucial mediators of the protective response in H-2b mice. Furthermore, passive transfer of immune serum from mice rejecting their tumors confers resistance in tumor antigen-tolerant animals with an inversely proportional relationship between tumor outgrowth and the amount of rat-erbB2 specific antibody present in tumor-bearing mice. Introduction of the rat-erb2 ectodomain into other H-2b tumor models also promotes their spontaneous tumor rejection. Notably, the tumor microenvironments differ in rat-erbB2+ tumor-bearing BALB.B and BALB/c mice at the time of fate decision in the models reflecting the differences between effective and ineffective tumor immune responses. Conclusions We find that the effective antitumor immunity targeting neoantigens in these breast cancer models is determined by MHC-II-restricted presentation of optimal cancer-associated antigens. These responses are dependent on CD4+ T cells, B cells, and antigen-specific antibodies.

Abstract T cells expressing chimeric antigen receptors (CAR T cells) are manufactured outside the body using non-physiological activation and in vitro culture protocols. The procedure takes 3 weeks and costs approximately $450,000 to administer. We have devised a scheme using mouse models for manufacturing CAR T cells in just 3 days employing a natural, polyclonal anti-viral immune response inside a lymph node (LN) using off the shelf reagents that can be applied widely in populations. CAR T cells emerge from the LN within a week, invade solid tumor, and eliminate the targeted antigen-positive tumor cells. Polyclonal activation of T cells is achieved by modifying the antigen presenting properties of lymph node resident professional antigen presenting cells (APC) through viral expression of allogeneic major histocompatibility antigens. The ensuing polyclonal activation occurs in the context of virus activated APC within the normal aarchitecture of the LN and confers robust T cell activation signals directing the development of cytolytic T cells. During the course of T cell activation, the T cells are transduced in situ with retroviruses encoding selected CARs. This approach could be applied to any number of different CARs to target a variety of tumors. T cells activated in the lymph node are functionally distinct from those activated using standard CAR T manufacturing protocols. We anticipate that CAR T cells generated in situ to natural stimuli will exhibit more normal immune regulatory properties that will impact their therapeutic utility. Because our approach uses off the shelf reagents and is administered using standard clinical procedures, cost to patients should be lowered and access to therapy increased.

Abstract To test the hypothesis that complex immune phenotypes can be defined by stable gene expression profiles we are using a RNAseq-based transcriptome analysis to relate phenotypes of activated peripheral blood CD4+ cells from healthy subjects and stage IV melanoma patients to their freshly-isolated counterparts. We have defined molecular immune response signatures (MIRS) containing genes that consistently differentiate individuals from one another. Two different phenotypic groups are defined using this approach, both of which are mathematically distinct from randomly-selected groups. First, a MIRS generated using healthy donor T cells stimulated with α-CD3/CD28 beads defines a distinct response phenotype and successfully predicts 17 independent healthy subject and melanoma patient responses, demonstrating the predictive capability of the approach. Second, our comparison of melanoma patient signatures with those of healthy subjects revealed a group of healthy individuals with distinct expression signatures that did not cluster with those of any of the melanoma patients. These results indicate that our RNAseq-based transcriptome analysis has the potential to identify phenotypic groups as well as gene sets that have response-predictive capabilities. Our work has both discovery and translational implications, as our approach can be used to investigate complex genetic contributions to immune responses and predict how individual patients will respond to therapeutic interventions.

Abstract The complexity of molecular networks regulating the immune response presents challenges to measuring immune response potential in human populations. We measured gene expression in CD4+ peripheral blood cells from healthy subjects using RNAseq. RNA expression was assessed twice, one month apart to identify the subset of genes expressed stably within individuals over time yet discriminating individuals in a cohort. This immune response signature gene set then was used to identify and successfully order the immune response phenotypes of independent, unrelated subjects. Following a brief 4 hr stimulation of freshly isolated cells with anti-CD3/C28 antibody-coated beads, changes in gene expression were measured using a similar approach, defining individual response profiles and clustering response patterns. Examination of the RNA expression signatures of unstimulated CD4 cells revealed a gene subset which when analyzed for gene expression levels correctly aligned independent test subjects into the defined response-phenotype clusters following stimulation. Therefore, there exists a subset of genes stably expressed over time in circulating CD4+ cells which discriminate among individuals in a manner associated with immune response potential. Thus a systems approach can be used to describe the complex immune phenotypes of individuals in ways that might be useful in determining how individuals will respond to complex stimuli or in disease and treatment settings.

Abstract Breast tumors develop under environmental pressures with phenotypically variant cells generated by mutation and epigenetic changes providing the substrate for clonal selection. Chromosomal mutations are a feature of spontaneous breast tumors in the BALB-neuT mouse model. However, there is little evidence that specific changes in chromosomal structure or ploidy confer selective advantage in these spontaneous tumors. The prominent exception is a consistent loss of chromosome 4, a little understood feature common to several mouse tumors. We previously measured heterozygosity in spontaneous breast tumors from FVB X BALB-neuT F1 mice using Illumina’s Golden Gate assay. In addition to the expected gain and loss of heterozygosity (LOH) evident throughout the genomes of a cohort of spontaneous tumors, the assay also detected wide-spread stochastic pseudo-“LOH” at 600 discrete loci spread throughout the genome in unique patterns among the tumors. Several of these positions were examined by sequence analysis revealing no deviations from the WT sequences, suggesting the detected “LOH” may have been generated by epigenetic modification of DNA which altered sequence detection. While, epigenetically modified DNA templates recapitulated the observed “LOH” signals, no canonical CpG motifs were present in the majority of the 600 loci probed, suggesting that an unusual DNA modification could be responsible for the unexpected wide-spread stochastic structural changes in the breast tumor DNA. We next assessed whether these putative epigenetic changes in DNA structure might impact gene regulation, and have reported that the stochastic pattern observed as “LOH” in DNA was recapitulated in the transcriptomes in unique patterns among the tumors. A direct correlation between the number of “LOH” variants and the down regulation of hundreds of non-polymorphic genes in the transcriptome also was noted. Furthermore, pathway analyses of the genes exhibiting changes in allelic ratios in these tumors revealed significant enrichments within gene networks regulating tissue homeostasis and antigen presentation, providing strong evidence that the perturbations in gene expression translated into selectable tumor cell phenotypes. We now have extended this study to examine the relationship between the magnitude of change in the expression of genes mapping within the pathways regulating tissue homeostasis (Molecular Mechanisms of Cancer). A remarkable feature of the flagged genes is that the magnitude of change in gene expression was not great in each case, yet the biological consequences were strongly reflected in the evolutionary history of the tumors. Importantly, the polymorphisms marking the parental alleles are mostly silent, not altering the structure of the encoded products. Therefore magnitude and timing of gene expression are the likely determinants of phenotypic variation. Each tumor contained several outliers within the pathways regulating tissue homeostasis, suggesting that the integration of multiple small perturbations in the expression of genes comprising functional networks could influence the biology properties of the tumor cells. Overall expression of the loci marked by allelic outliers was significantly below the average expression found among tumors in the cohort highlighting the importance of down regulation of one allele in the establishment of selectable traits. We find a similar direct correlation of multiple small changes in the transcriptome of normal lymphocytes with immune response phenotypes, suggesting this principle of integration of multiple small deviations in gene expression applies widely to the phenotypes of normal cells, tumors, and by extension to organismic traits. Citation Format: Larry R Pease, Sara J Felts, Adam D Scheid, Xiaojia Tang, Krishna R Kalari. Phenotypes of breast tumor cells and normal lymphocytes are determined by the integration of minor changes in expression of multiple genes: A new dimension in quantitative inheritance [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P4-09-03.

Abstract Molecular and cellular diversity in cancer cells enables evolutionary processes that govern tumor progression and contributes to treatment failures, but the mechanisms governing this variability remain incompletely understood. We previously described tumor-specific epigenetic alterations in DNA devoid of canonical CpG methylation sites throughout the genomes of spontaneous breast tumors which arise in genetically uniform, yet heterozygous, [FVB X BALB/c MMTV-NeuT+] F1 mice. Those changes were detected in purified tumor DNA using a template-assisted quantitative ligation assay employing locus-specific and SNP-specific oligonucleotide genotyping probes, and the assay was subsequently shown to be sensitive to epigenetic modifications of the templates. We now describe losses of heterozygosity similarly distributed throughout the transcriptomes of tumors in this same model. We quantified expressed alleles across a population of clonal tumors and tested each expressed variant for the occurrence of allelic ratio outliers. The results revealed variegated patterns of expression in hundreds of genes, highly significantly enriched in pathways typically co-opted by tumors (Molecular Mechanisms of Cancer, P < 10-10, IPA analysis). An astounding 87% (2634 of 3044) of expressed genes that could be measured using this approach were represented in at least one tumor. Furthermore, the frequency of these outliers in any one individual tumor was found to correlate strongly with the transcriptional repression of an additional large set of non-polymorphic genes. These findings reveal underappreciated, latent mechanisms driving sporadic errors in epigenetic programming that promote repression of a multiple cis-linked transcripts. This genome-wide molecular chaos presents as dysregulated cellular homeostasis and is the foundation for heritable diversity as tumors evolve. Citation Format: Sara J. Felts, Xiaojia Tang, Virginia P. Van Keulen, Krishna R. Kalari, Larry R. Pease. Stochastic variation in gene expression is selected during clonal evolution of spontaneous mouse mammary tumors [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 2898.

Abstract Genome-wide identification and characterization of long noncoding RNAs (lncRNA) in individual immune cell lineages helps us better understand the driving mechanisms behind melanoma and advance personalized patient treatment. To elucidate the transcriptional landscape in diverse immune cell types of peripheral blood cells (PBC) in stage IV melanoma, we used whole transcriptome RNA sequencing to profile lncRNAs in CD4+, CD8+, and CD14+ PBC from 132 patient samples. Our integrative computational approach identified 27,625 expressed lncRNAs, 2,744 of which were novel. Both T cells (i.e., CD4+ and CD8+ PBC) and monocytes (i.e., CD14+ PBC) exhibited differential transcriptional expression profiles between patients with melanoma and healthy subjects. Cis- and trans-level coexpression analysis suggested that lncRNAs are potentially involved in many important immune-related pathways and the programmed cell death receptor 1 checkpoint pathways. We also identified nine gene coexpression modules significantly associated with melanoma status, all of which were significantly enriched for three mRNA translation processes. Age and melanoma traits closely correlated with each other, implying that melanoma contains age-associated immune changes. Our computational prediction analysis suggests that many cis- and trans-regulatory lncRNAs could interact with multiple transcriptional and posttranscriptional regulatory elements in CD4+, CD8+, and CD14+ PBC, respectively. These results provide novel insights into the regulatory mechanisms involving lncRNAs in individual immune cell types in melanoma and can help expedite cell type-specific immunotherapy treatments for such diseases. Significance: These findings elucidate melanoma-associated changes to the noncoding transcriptional landscape of distinct immune cell classes, thus providing cell type-specific guidance to targeted immunotherapy regimens. Cancer Res; 78(15); 4411–23. ©2018 AACR.

Abstract Human immunity exhibits remarkable heterogeneity among individuals, which engenders variable responses to immune perturbations in human populations. Population studies reveal that, in addition to interindividual heterogeneity, systemic immune signatures display longitudinal stability within individuals, and these signatures may reliably dictate how given individuals respond to immune perturbations. We hypothesize that analyzing relationships among these signatures at the population level may uncover baseline immune phenotypes that correspond with response outcomes to immune stimuli. To test this, we quantified global gene expression in peripheral blood CD4+ cells from healthy individuals at baseline and following CD3/CD28 stimulation at two time points 1 mo apart. Systemic CD4+ cell baseline and poststimulation molecular immune response signatures (MIRS) were defined by identifying genes expressed at levels that were stable between time points within individuals and differential among individuals in each state. Iterative differential gene expression analyses between all possible phenotypic groupings of at least three individuals using the baseline and stimulated MIRS gene sets revealed shared baseline and response phenotypic groupings, indicating the baseline MIRS contained determinants of immune responsiveness. Furthermore, significant numbers of shared phenotype-defining sets of determinants were identified in baseline data across independent healthy cohorts. Combining the cohorts and repeating the analyses resulted in identification of over 6000 baseline immune phenotypic groups, implying that the MIRS concept may be useful in many immune perturbation contexts. These findings demonstrate that patterns in complex gene expression variability can be used to define immune phenotypes and discover determinants of immune responsiveness.