The partial success of tumor immunotherapy induced by checkpoint blockade, which is not antigen-specific, suggests that the immune system of some patients contain antigen receptors able to specifically identify tumor cells. Here we focused on T-cell receptor (TCR) repertoires associated with spontaneous breast cancer. We studied the alpha and beta chain CDR3 domains of TCR repertoires of CD4 T cells using deep sequencing of cell populations in mice and applied the results to published TCR sequence data obtained from human patients. We screened peripheral blood T cells obtained monthly from individual mice spontaneously developing breast tumors by 5 months. We then looked at identical TCR sequences in published human studies; we used TCGA data from tumors and healthy tissues of 1,256 breast cancer resections and from 4 focused studies including sequences from tumors, lymph nodes, blood and healthy tissues, and from single cell dataset of 3 breast cancer subjects. We now report that mice spontaneously developing breast cancer manifest shared, Public CDR3 regions in both their alpha and beta and that a significant number of women with early breast cancer manifest identical CDR3 sequences. These findings suggest that the development of breast cancer is associated, across species, with biomarker, exclusive TCR repertoires.

The B cell response to Ehrlichia muris is dominated by plasmablasts (PBs), with few-if any-germinal centers (GCs), yet it generates protective immunoglobulin M (IgM) memory B cells (MBCs) that express the transcription factor T-bet and harbor V-region mutations. Because Ehrlichia prominently infects the liver, we investigated the nature of liver B cell response and that of the spleen. B cells within infected livers proliferated and underwent somatic hypermutation (SHM). Vh-region sequencing revealed trafficking of clones between the spleen and liver and often subsequent local clonal expansion and intraparenchymal localization of T-bet+ MBCs. T-bet+ MBCs expressed MBC subset markers CD80 and PD-L2. Many T-bet+ MBCs lacked CD11b or CD11c expression but had marginal zone (MZ) B cell phenotypes and colonized the splenic MZ, revealing T-bet+ MBC plasticity. Hence, liver and spleen are generative sites of B cell responses, and they include V-region mutation and result in liver MBC localization.

Although structural studies of individual T cell receptors (TCRs) have revealed important roles for both the α and β chain in directing MHC and antigen recognition, repertoire-level immunogenomic analyses have historically examined the β chain alone. To determine the amount of useful information about TCR repertoire function encoded within αβ pairings, we analyzed paired TCR sequences from nearly 100,000 unique CD4+ and CD8+ T cells captured using two different high-throughput, single-cell sequencing approaches. Our results demonstrate little overlap in the healthy CD4+ and CD8+ repertoires, with shared TCR sequences possessing significantly shorter CDR3 sequences corresponding to higher generation probabilities. We further utilized tools from information theory and machine learning to show that while α and β chains are only weakly associated with lineage, αβ pairings appear to synergistically drive TCR-MHC interactions. Vαβ gene pairings were found to be the TCR feature most informative of T cell lineage, supporting the existence of germline-encoded paired αβ TCR-MHC interaction motifs. Finally, annotating our TCR pairs using a database of sequences with known antigen specificities, we demonstrate that approximately a third of the T cells possess α and β chains that each recognize different known antigens, suggesting that αβ pairing is critical for the accurate inference of repertoire functionality. Together, these findings provide biological insight into the functional implications of αβ pairing and highlight the utility of single-cell sequencing in immunogenomics.

Analysis of antibody repertoires by high-throughput sequencing is of major importance in understanding adaptive immune responses. Our knowledge of variations in the genomic loci encoding immunoglobulin genes is incomplete, resulting in conflicting VDJ gene assignments and biased genotype and haplotype inference. Haplotypes can be inferred using IGHJ6 heterozygosity, observed in one third of the people. Here, we propose a robust novel method for determining VDJ haplotypes by adapting a Bayesian framework. Our method extends haplotype inference to IGHD - and IGHV -based analysis, enabling inference of deletions and copy number variations in the entire population. To test this method, we generated a multi-individual data set of naive B-cell repertoires, and found allele usage bias, as well as a mosaic, tiled pattern of deleted IGHD and IGHV genes. The inferred haplotypes may have clinical implications for genetic disease predispositions. Our findings expand the knowledge that can be extracted from antibody repertoire sequencing data.

While a highly diverse T-cell receptor (TCR) repertoire is the hallmark of a healthy adaptive immune system, relatively little is understood about how the CD4 + and CD8 + TCR repertoires differ from one another. We here utilize high-throughput single T-cell sequencing to obtain approximately 100,000 TCR αβ chain pairs from human subjects, stratified into CD4 + and CD8 + lineages. We reveal that substantial information about T-cell lineage is encoded by V αβ gene pairs and, to a lesser extent, by several other TCR features such as CDR3 length and charge. We further find that the strength of association between the β chain and T-cell lineage is surprisingly weak, similar in strength to that of the α chain. Using machine learning classifiers to predict T-cell lineage from TCR features, we demon-strate that αβ chain pairs are significantly more informative than individual chains alone. These findings provide unprecedented insight into the CD4 + and CD8 + TCR repertoires and highlight the importance of αβ chain pairing in TCR function and specificity.

Cancer immunotherapy by checkpoint blockade proves that an effective immune response to a tumor can be induced clinically. However, little is known about the evolution of tumor-associated T-cell receptor (TCR) repertoires without intervention. Here we studied TCR repertoire evolution in mice spontaneously developing mammary tumors; we sequenced peripheral blood alpha and beta TCRs of CD4+CD62L+CD44− T cells monthly for 8 months in 10 FVB/NJ mice transgenic at the Erbb2 locus, all developing tumors; 5 FVB/NJ mice without the transgene were age-matched controls. Sequences were either private (restricted to one mouse) or public (shared among mice); public sequences were either exclusive to the tumor group or inclusive among different groups. We now report that 1), public AA sequences were each encoded by many different nucleotide sequences (NT) recombinations (convergent recombination; CR); 2) mice developing tumors evolved tumor-exclusive public sequences, derived initially from private or from inclusive public sequences; and 3) tumor-exclusive public sequences in mice were also present among published public TCR sequences from human breast cancer patients. These cross-species tumor-exclusive TCR sequences manifested high CR; but the AA sequences shared by mice and humans did not share NT sequences. Thus, tumor-exclusive TCR AA sequences across species are selected from different NT recombination events. The roles of tumor-exclusive TCR repertoires in advancing or inhibiting tumor development and the effects of tumor immunotherapy on these T cells remain to be seen.

During manufacturing of an exemplary CD19-directed CAR-T cell product, patient T cells were transduced with a viral vector encoding the CD19-directed, 41BBζ endodomain-containing chimeric antigen receptor (CAR) construct, enabling permanent genetic modification of the target T cells. Lentiviral transduction with a self-inactivating (SIN) vector can offer certain advantages in safety and consistency over other DNA delivery methods. Although the integration patterns of wildtype HIV-1 have been well-characterized and shown to have a low risk of insertional transformation, the addition of an artificial construct may have the potential to affect integration patterns. Existing techniques for viral integration mapping, such as ligation-mediated polymerase chain reaction (LM-PCR), have limitations precluding comprehensive, unbiased, high-throughput insertion site mapping. To characterize insertional events in engineered CAR-T cells, we developed an unbiased NGS assay and computational pipeline capable of characterizing viral integration events genome-wide. The method leverages Tn5 transposase, coupled with a long terminal repeat-specific PCR, to map integration events down to nulceotide resolution. This assay identified over 150,000 high-confidence integration events across 23 patient-derived CAR-T cell products. Analysis of these sites revealed a similar integration preference to lentiviruses described in the literature: driven by GC content, local nucleotide content, chromatin status, and transcriptional activity, with no indication of preferential integration at oncogenic loci. In conclusion, in this assay a CD19-directed, CAR-containing lentivirus exhibited a low risk integration profile with no significant evidence of construct-specific risk factors. Furthermore, the findings suggest that the NGS assay may be useful for more efficient, cost-effective, and sensitive measurement of CAR-T integration events for a range of constructs. Citation Format: Timothy G. Johnstone, Rajagopal Chari, David Koppstein, Ronald J. Hause, Rafael Ponce, Adrian W. Briggs, Francois Vigneault. Directed, unbiased mapping of lentiviral integrations with next generation sequencing in CAR-T cells [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 1541.

Hepatitis C virus (HCV) is a major public health concern, with over 70 million people infected worldwide, who are at risk for developing life-threatening liver disease. No vaccine is available, and immunity against the virus is not well-understood. Following the acute stage, HCV usually causes chronic infections. However, ~30% of infected individuals spontaneously clear the virus. Therefore, using HCV as a model for comparing immune responses between spontaneous clearer (SC) and chronically infected (CI) individuals may empower the identification of mechanisms governing viral infection outcomes. Here, we provide the first in-depth analysis of adaptive immune receptor repertoires in individuals with current or past HCV infection. We demonstrate that SC individuals, in contrast to CI patients, develop clusters of antibodies with distinct properties. These antibodies' characteristics were used in a machine learning framework to accurately predict infection outcome. Using combinatorial antibody phage display library technology, we identified HCV-specific antibody sequences. By integrating these data with the repertoire analysis, we constructed two antibodies characterized by high neutralization breadth, which are associated with clearance. This study provides insight into the nature of effective immune response against HCV and demonstrates an innovative approach for constructing antibodies correlating with successful infection clearance. It may have clinical implications for prognosis of the future status of infection, and the design of effective immunotherapies and a vaccine for HCV.

This paper reports on the use of a digital microfluidic platform to perform multiplex automated genetic engineering (MAGE) cycles on droplets containing Escherichia coli cells. Bioactivated magnetic beads were employed for cell binding, washing, and media exchange in the preparation of electrocompetent cells in the electrowetting-on-dieletric (EWoD) platform. On-cartridge electroporation was used to deliver oligonucleotides into the cells. In addition to the optimization of a magnetic bead-based benchtop protocol for generating and transforming electrocompetent E. coli cells, we report on the implementation of this protocol in a fully automated digital microfluidic platform. Bead-based media exchange and electroporation pulse conditions were optimized on benchtop for transformation frequency to provide initial parameters for microfluidic device trials. Benchtop experiments comparing electrotransformation of free and bead-bound cells are presented. Our results suggest that dielectric shielding intrinsic to bead-bound cells significantly reduces electroporation field exposure efficiency. However, high transformation frequency can be maintained in the presence of magnetic beads through the application of more intense electroporation pulses. As a proof of concept, MAGE cycles were successfully performed on a commercial EWoD cartridge using variations of the optimal magnetic bead-based preparation procedure and pulse conditions determined by the benchtop results. Transformation frequencies up to 22% were achieved on benchtop; this frequency was matched within 1% (21%) by MAGE cycles on the microfluidic device. However, typical frequencies on the device remain lower, averaging 9% with a standard deviation of 9%. The presented results demonstrate the potential of digital microfluidics to perform complex and automated genetic engineering protocols.

Tumor-infiltrating lymphocytes (TILs) are critical to anti-cancer immune responses, but their diverse phenotypes and functions remain poorly understood and challenging to study. We therefore developed a single-cell barcoding technology for deep characterization of TILs without the need for cell-sorting or culture. Our emulsion-based method captures full-length, natively paired B-cell and T-cell receptor (BCR and TCR) sequences from lymphocytes among millions of input cells. We validated the method with 3 million B-cells from healthy human blood and 350,000 B-cells from an HIV elite controller, before processing 400,000 cells from an unsorted dissociated ovarian adenocarcinoma and recovering paired BCRs and TCRs from over 11,000 TILs. We then extended the barcoding method to detect DNA-labeled antibodies, allowing ultra-high throughput, simultaneous protein detection and RNA sequencing from single cells.

Myasthenia gravis (MG) is a prototypical B cell-mediated autoimmune disease affecting 20–50 people per 100,000. The majority of patients fall into two clinically distinguishable types based on whether they produce autoantibodies targeting the acetylcholine receptor (AChR-MG) or muscle specific kinase (MuSK-MG). The autoantibodies are pathogenic, but whether their generation is associated with broader defects in the B cell repertoire is unknown. To address this question, we performed deep sequencing of the BCR repertoire of AChR-MG, MuSK-MG, and healthy subjects to generate ∼518,000 unique VH and VL sequences from sorted naive and memory B cell populations. AChR-MG and MuSK-MG subjects displayed distinct gene segment usage biases in both VH and VL sequences within the naive and memory compartments. The memory compartment of AChR-MG was further characterized by reduced positive selection of somatic mutations in the VH CDR and altered VH CDR3 physicochemical properties. The VL repertoire of MuSK-MG was specifically characterized by reduced V-J segment distance in recombined sequences, suggesting diminished VL receptor editing during B cell development. Our results identify large-scale abnormalities in both the naive and memory B cell repertoires. Particular abnormalities were unique to either AChR-MG or MuSK-MG, indicating that the repertoires reflect the distinct properties of the subtypes. These repertoire abnormalities are consistent with previously observed defects in B cell tolerance checkpoints in MG, thereby offering additional insight regarding the impact of tolerance defects on peripheral autoimmune repertoires. These collective findings point toward a deformed B cell repertoire as a fundamental component of MG.

ABSTRACT A diverse T cell repertoire is a critical component of the adaptive immune system, providing protection against invading pathogens and neoplastic changes, relying on the recognition of foreign antigens and neoantigen peptides by T cell receptors (TCRs). However, the statistical properties and function of the T cell pool in an individual, under normal physiological conditions, are poorly understood. In this study, we report a comprehensive, quantitative characterization of the T cell repertoire from over 1.9 million cells, yielding over 200,000 high quality paired αβ sequences in 5 healthy human subjects. The dataset was obtained by leveraging recent biotechnology developments in deep RNA sequencing of lymphocytes via single-cell barcoding in emulsion. We report non-random associations and non-monogamous pairing between the α and β chains, lowering the theoretical diversity of the T cell repertoire, and increasing the frequency of public clones shared among individuals. T cell clone size distributions closely followed a power law, with markedly longer tails for CD8 + cytotoxic T cells than CD4 + helper T cells. Furthermore, clonality estimates based on paired chains from single T cells were lower than that from single chain data. Taken together, these results highlight the importance of sequencing αβ pairs to accurately quantify lymphocyte receptor diversity.

Adaptive immunity is driven by the expansion, somatic hypermutation, and selection of B cell clones. Each clone is the progeny of a single B cell responding to Ag, with diversified Ig receptors. These receptors can now be profiled on a large scale by next-generation sequencing. Such data provide a window into the microevolutionary dynamics that drive successful immune responses and the dysregulation that occurs with aging or disease. Clonal relationships are not directly measured, but they must be computationally inferred from these sequencing data. Although several hierarchical clustering-based methods have been proposed, they vary in distance and linkage methods and have not yet been rigorously compared. In this study, we use a combination of human experimental and simulated data to characterize the performance of hierarchical clustering-based methods for partitioning sequences into clones. We find that single linkage clustering has high performance, with specificity, sensitivity, and positive predictive value all >99%, whereas other linkages result in a significant loss of sensitivity. Surprisingly, distance metrics that incorporate the biases of somatic hypermutation do not outperform simple Hamming distance. Although errors were more likely in sequences with short junctions, using the entire dataset to choose a single distance threshold for clustering is near optimal. Our results suggest that hierarchical clustering using single linkage with Hamming distance identifies clones with high confidence and provides a fully automated method for clonal grouping. The performance estimates we develop provide important context to interpret clonal analysis of repertoire sequencing data and allow for rigorous testing of other clonal grouping algorithms.

Snakebite envenomings represent a neglected public health issue in many parts of the rural tropical world. Animal-derived antivenoms have existed for more than a hundred years and are effective in neutralizing snake venom toxins when timely administered. However, the low immunogenicity of many small but potent snake venom toxins represents a challenge for obtaining a balanced immune response against the medically relevant components of the venom. Here, we employ high-throughput sequencing of the immunoglobulin (Ig) transcriptome of mice immunized with a three-finger toxin and a phospholipase A2 from the venom of the Central American coral snake, Micrurus nigrocinctus. Although exploratory in nature, our indicate results showed that only low frequencies of mRNA encoding IgG isotypes, the most relevant isotype for therapeutic purposes, were present in splenocytes of five mice immunized with 6 doses of the two types of toxins over 90 days. Furthermore, analysis of Ig heavy chain transcripts showed that no particular combination of variable (V) and joining (J) gene segments had been selected in the immunization process, as would be expected after a strong humoral immune response to a single antigen. Combined with the titration of toxin-specific antibodies in the sera of immunized mice, these data support the low immunogenicity of three-finger toxins and phospholipases A2 found in M. nigrocinctusvenoms, and highlight the need for future studies analyzing the complexity of antibody responses to toxins at the molecular level.

Snakebite envenomings represent a neglected public health issue in many parts of the rural tropical world.Animal-derived antivenoms have existed for more than a hundred years and are effective in neutralizing snake venom toxins when timely administered.However, the low immunogenicity of many small but potent snake venom toxins represents a challenge for obtaining a balanced immune response against the medically relevant components of the venom.Here, we employ high-throughput sequencing of the immunoglobulin (Ig) transcriptome of mice immunized with a three-finger toxin and a phospholipase A 2 from the venom of the Central American coral snake, Micrurus nigrocinctus.Although exploratory in nature, our indicate results showed that only low frequencies of mRNA encoding IgG isotypes, the most relevant isotype for therapeutic purposes, were present in splenocytes of five mice immunized with 6 doses of the two types of toxins over 90 days.Furthermore, analysis of Ig heavy chain transcripts showed that no particular combination of variable (V) and joining (J) gene segments had been selected in the immunization process, as would be expected after a strong humoral immune response to a single antigen.Combined with the titration of toxin-specific antibodies in the sera of immunized mice, these data support the low immunogenicity of three-finger toxins and phospholipases A 2 found in M. nigrocinctus venoms, and highlight the need for future studies analyzing the complexity of antibody responses to toxins at the molecular level.

Abstract We present high resolution single cell profiling of tumor infiltrating lymphocytes (TILs) for immune research and cancer target discovery. TILs hold the key to understanding anti-cancer immune responses but remain challenging to study due to their vast range of phenotypes, functions and abundance. Current approaches to TIL analysis such as immune receptor sequencing, flow cytometry, and RNA expression profiling are limited, since each approach gives only partial insights into a TIL repertoire and their results cannot be connected at the single cell level. We therefore built a technology for simultaneous determination of immune receptor sequences, surface protein phenotypes and RNA expression profiles from single TILs at unprecedented scale. Able to process millions of cells per experiment, the throughput of our emulsion-based DNA barcoding method allows deep and unbiased TIL profiling directly from tumor tissue without the need for cell sorting, stimulation or culture. Briefly, single-cell dissociated biopsy tissue samples are stained with DNA-labeled surface marker antibodies and isolated into individual ∼65-pL microdroplets by a microfluidic emulsion chip. Within the droplets, cellular RNAs are reverse transcribed into cDNA, and a barcoding step attaches droplet-specific DNA barcodes to all of the cDNAs as well as to the DNA labels carried by the marker antibodies. At the end of the reaction, all products associated with a single input cell carry the same DNA barcode sequence, allowing subsequent high throughput sequencing to assign the products to their cell of origin. During analysis the full-length immune receptor pairs for each B- and T-cell are reconstructed and the marker antibody tags are identified and quantified along with dozens of additional mRNA markers, resulting in a high-dimensional dataset containing integrated immune receptor clone, protein marker and RNA expression profiles of thousands to hundreds of thousands of individual lymphocytes in the input sample. We have successfully applied our method to a variety of healthy blood and tissue samples including human cancers such as pancreatic, ovarian and lung cancer, profiling millions of B-cells and T-cells present in the samples. We identify likely tumor-reactive TILs by finding BCR or TCR clonal lineages associated with particular protein and RNA signatures, including expression of immunosuppression markers such as PD-1 and CTLA-4. Following candidate selection, full-length BCR and TCR pairs are synthesized and expressed for functional screening and target antigen identification. By providing an unprecedented view into TIL diversity across a wide range of disease and sample types, our platform also has widespread potential for basic and applied research into the patient immune response to cancer and the dynamics of immunotherapy. Citation Format: Katherine L. Connor, Adrian W. Briggs, Stephen J. Goldfless, Sonia Timberlake, Brian J. Belmont, Christopher R. Clouser, David Koppstein, Francois Vigneault. Comprehensive TIL profiling by simultaneous DNA barcoding of proteins, RNA and natively paired immune receptors from millions of single cells. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 1442.

Abstract Immune sequencing allows for the study of complex immunological diseases by sequencing millions of V(D)J combinations from B-cell antibody and T-cell receptors. The popularity of this technique has increased due to recent throughput and read length improvements in next-generation sequencing technologies. However, structural and sequence complexities of antibody genes have made reliable targeting approaches challenging. We have developed and optimized a method for accurate sequencing of full-length immune gene repertoires of B-cells and T-cells. The method uses a unique barcoding scheme specifically designed to tag every mRNA molecule with a unique identifier (UID) so that all PCR copies of each mRNA fragment can be collapsed into a single consensus sequence. This makes the assay extremely accurate, by resolving PCR bias and sequencing errors as well as allowing quantitative digital molecule counting. Immune sequencing libraries were generated from total RNA extracted from Peripheral Blood Mononuclear Cells in duplicate from a single patient. The use of UIDs enabled absolute quantification of starting RNA molecules present in the original sample and therefore accurate ranking of the antibody clone abundance, by avoiding the bias incorporated by PCR or sequencing when total reads only were measured. Using the same sequencing method, tumor samples were analyzed for abundance of expanded clones via grouping clones by V gene, J gene and CDR3 similarity and ranking by mRNA abundance. Additionally, the use of isotype-specific primers (IgM, IgD, IgG, IgA and IgE) enabled measurement of the heavy chain isotype proportions within the samples. Further, alignment of full-length heavy chain antibody sequences generated using this method to germline genes from reference databases enabled quantitation of the mutation level of each antibody sequence, thereby providing information on the overall maturity and mutational profile of the sample repertoire. Citation Format: Fiona J. Stewart, Mehmet Karaca, Kris Adams, Chris Clouser, Bonny Patel, Sonia Timberlake, William Donahue, Lynne Apone, Salvatore Russello, Eileen T. Dimalanta, Theodore B. Davis, Francois Vigneault. Sequencing the B-cell and T-cell repertoire. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 3998.