CDR3 Amino Acid Research Articles

Combination of oncolytic Maraba virus with immune checkpoint blockade overcomes therapy resistance in an immunologically cold model of advanced melanoma with dysfunctional T-cell receptor signalling

BackgroundOver the past decade, cancer immunotherapies have revolutionized the treatment of melanoma; however, responses vary across patient populations. Recently, baseline tumor size has been identified as an independent prognostic factor...

VNARs as Neutralizing Intracellular Therapeutic Agents: Glioblastoma as a Target.

Glioblastoma is the most prevalent and fatal form of primary brain tumors. New targeted therapeutic strategies for this type of tumor are imperative given the dire prognosis for glioblastoma patients and the poor results of current multimodal therapy. Previously reported drawbacks of antibody-based therapeutics include the inability to translocate across the blood-brain barrier and reach intracellular targets due to their molecular weight. These disadvantages translate into poor target neutralization and cancer maintenance. Unlike conventional antibodies, vNARs can permeate tissues and recognize conformational or cryptic epitopes due to their stability, CDR3 amino acid sequence, and smaller molecular weight. Thus, vNARs represent a potential antibody format to use as intrabodies or soluble immunocarriers. This review comprehensively summarizes key intracellular pathways in glioblastoma cells that induce proliferation, progression, and cancer survival to determine a new potential targeted glioblastoma therapy based on previously reported vNARs. The results seek to support the next application of vNARs as single-domain antibody drug-conjugated therapies, which could overcome the disadvantages of conventional monoclonal antibodies and provide an innovative approach for glioblastoma treatment.

Single-Cell Transcriptomes and Immune Repertoires Reveal the Cell State and Molecular Changes in Pemphigus Vulgaris.

The etiology and pathogenesis of pemphigus vulgaris (PV) entail intricate interactions between immune cells and epithelial cells. However, the specific subtypes of immune cells involved in PV, along with their respective roles, remain elusive. Likewise, the precise functions and mechanisms by which glucocorticoids affect cell types within the disease context require further elucidation. To address these knowledge gaps, we performed 5' single-cell RNA sequencing, combined with V(D)J enrichment on buccal mucosal lesions and peripheral blood samples from treatment-naive patients with PV, in conjunction with post-treatment peripheral blood samples obtained after oral prednisone treatment. Our findings suggest that the IL-1α signaling pathway, myeloid APCs, inflammatory CD8+ resident memory T cells, and dysfunctional CD4+ regulatory T cells are involved in the pathogenesis of PV. Part of these findings were validated by immunohistochemical assays and multiplex immunofluorescence assays. Furthermore, our results highlight the significant impact of prednisone treatment on monocytes and mucosal-associated invariant T cells while revealing a limited effect on CD4+ regulatory T cells. Additionally, we present the CDR3 amino acid sequence of BCR related to PV disease and investigate the characteristics of TCR/BCR clonotypes. In conclusion, our study provides a comprehensive understanding of PV, particularly focusing on the mucosal-dominant type, and sheds light on the effects of glucocorticoids within the PV context. These insights hold promise for the development of new therapeutic strategies in this autoimmune disorder.

Aplastic Anemia Patients Have Diverse T Cell Repertoires and Flexible T Cell Responses Similar to Healthy Individuals

Acquired aplastic anemia (AA) is a bone marrow failure disease caused by T cell-mediated autoimmune attack on hematopoietic stem and progenitor cells (HSPCs). Identifying autoreactive T cells in AA has been challenging. Advances in T cell receptor (TCR) sequencing and the discovery of Human Leukocyte Antigen Class I (HLA-I) risk alleles linked to autoantigen presentation in AA have rekindled interest in exploring T cell repertoires in AA. We hypothesized that autoreactive CD8 + T cells in AA may have convergent TCR structures allowing them to recognize AA autoantigens presented by specific HLA-I risk alleles. To test our hypothesis, we analyzed TCR repertoires and HLA-restricted CD8 + T cell reactivity in AA patients. Using TCR Vβ (TRB) sequencing, we analyzed 83 patient samples obtained at AA diagnosis or relapse. High-risk alleles HLA-B*14:02 and B*40:02 were present in 26% and 10% of patients, respectively. HLA-DR*15 was present in 24%. Patient samples were largely sorted bone marrow (BM) CD8 + T lymphocytes. Patient TRB repertoires were compared to those in 709 control samples, largely of unsorted peripheral blood (PB). To account for sampling differences, we downsampled sequences to the minimum number of clones in the matching tissue. Normalized clone size distributions were visualized on rank plots (Fig 1). TRB repertoires in both patients and controls were diverse. Highly expanded clones (>5%) were rare in either group. Patients had 1,415,069 unique CDR3 amino acid sequences (CDR3AA), of which 106,446 (7.5%) were shared by ≥2 patients. Among the 106,446 shared CDR3AA, 95.4% were found in controls, and 4.3% TCRs had reported specificity to exogenous (mostly viral) or endogenous antigens in VDJdb. 4,920 (4.6%) of the shared CDR3AA were detected only in patients (median 2 patients (2-25)). Of these, 190 exhibited TCR convergence in ≥1 patient, such that the same CDR3AA was encoded by ≥2 distinct nucleotide sequences. Clustering 4,920 patient-associated CDR3AA using GLIPH (Grouping of lymphocyte interactions by paratope hotspots) identified 43 TCR antigen convergence groups. CDR3AA in each cluster were not significantly expanded in the majority of patients and were not enriched for the high-risk alleles HLA-B*14:02 or B*40:02. Our data suggest that most CDR3AA in AA are private (unique to individual patients). Shared autoreactive CDR3AA may exist in patients but are likely to also be present in healthy individuals. As the next step towards identifying autoreactive T cells, we focused on AA patients with known HLA-I risk alleles. We tested their CD8 + T cells for reactivity with endogenously expressed antigens by stimulating T cells with hematopoietic cell lines engineered to express single HLA-I alleles matching those of the patients (Fig 2). TRB sequencing showed increased clonality and significant expansion of a subset of clonotypes after T cell stimulation. Interferon γ enzyme-linked immunospot assays confirmed the expanded T cells' responsiveness to antigens presented by HLA alleles used for T cell stimulation. Additionally, these expanded T cells demonstrated variable reactivity to cells expressing other HLA alleles from the same patient. Among the 1% of CDR3AA clonotypes most highly expanded with single HLA-I alleles (63-3000-fold), we identified several that were expanded in an HLA-restricted manner. However, several highly expanded CDR3AA were highly expanded with multiple HLA alleles. Potential reasons for expansion of some CDR3AA after stimulation with cells expressing different HLA alleles include 1) binding of the same CDR3AA to different antigens presented by distinct HLA alleles, 2) expansion in response to the same antigens presented by distinct HLA alleles, or 3) expansion in response to non-HLA antigens. In conclusion, our data show diverse and mostly private TRB repertoires in AA patients. Our in vitro findings suggest that much like T cell responses in non-AA individuals, T cell responses in AA are diverse and exhibit flexibility with respect to activation in response to various endogenous antigens in the context of different self HLA-I alleles. Given the diversity of TCR responses observed in AA patients, the tissue specificity of autoimmune attack and established association with risk HLA-I alleles in AA remain a mystery and likely require HSPC-specific autoantigen expression and efficient autoantigen presentation by individual patients' HLA-I alleles.

257. Shared antibody sequences reveal novel public clonotypes of antibodies within and between HIV infected and healthy individuals

Abstract Background Public clonotypes, antibodies against specific antigens in unrelated individuals that have genetic similarities, have been shown in a variety of infections, including SARS-CoV-2 and HIV. These can be exploited as potential diagnostics and to identify past infections that may be related to post-infectious disorders, such as Kawasaki disease, Multiple Sclerosis or Alzheimer's disease. Methods Heavy chain variable sequences were retrieved from public biorepositories (Bioprojects PRJNA486667, PRJNA638224,PRJNA527941) from 270 subjects. These were derived from labeled cohorts: Healthy (116), HIV broadly neutralizing (46), HIV non-broadly-neutralizing (50), HIV-negative (43), and COVID-19 (15). We utilized the Immcantation package of software run on our SUNY Buffalo computational cluster. After PRESTo annotation, duplicate sequences were collapsed and sequences of only single counts were removed. Clonal groups were determined using SCOPer requiring IGHV, IGHJ, and CDR3 amino acid sequence to be perfectly matched. Figures and statistics were generated with immcantation, excel, pandas, and graphpad prism 8. Results There were 5.5 million unique heavy chain sequences from 32 million sequences. Defining public clonotypes as completely sharing CDR3 amino acid sequence and having predicted same germline variable, D, and J chains, 6861 (0.12%) shared clones were found. The distributions of CDR3 lengths showed skewing to shorter CDR3s (mean length of 10.5). There were no stringently defined clones noted in all members. Within HIV infected persons the two most shared clones both used VH1-18, JH4, had CDRaa 8, and were immature IGM/D. Three clones were used in near 10% of cases across both non-HIV and HIV datasets with prominent usage of VH 4 family members and were less mutated and IGM/D. There were eight clones solely shared > 10% within non-HIV/non-Covid infected individuals, with the top shared clones all consistent with matured IGG responses with prominent predicted use of VH3-7 and a number of novel CDR3s noted. Predicted CDR3 length using incantation was compared for cohort of clones that shared predicted V, D, J usage and the exact CDR3 amino acids between individuals. This distribution failed normality testing. Conclusion Identification of public clonotypes will show stochastic bias toward shorter CDR3 regions. There were differences between HIV and non-HIV cohorts and a number of novel public clonotypes were found. Disclosures Mark D. Hicar, MD/PhD, Pfizer: site investigator for 2 trial

Germline-encoded specificities and the predictability of the B cell response.

Antibodies result from the competition of B cell lineages evolving under selection for improved antigen recognition, a process known as affinity maturation. High-affinity antibodies to pathogens such as HIV, influenza, and SARS-CoV-2 are frequently reported to arise from B cells whose receptors, the precursors to antibodies, are encoded by particular immunoglobulin alleles. This raises the possibility that the presence of particular germline alleles in the B cell repertoire is a major determinant of the quality of the antibody response. Alternatively, initial differences in germline alleles' propensities to form high-affinity receptors might be overcome by chance events during affinity maturation. We first investigate these scenarios in simulations: when germline-encoded fitness differences are large relative to the rate and effect size variation of somatic mutations, the same germline alleles persistently dominate the response of different individuals. In contrast, if germline-encoded advantages can be easily overcome by subsequent mutations, allele usage becomes increasingly divergent over time, a pattern we then observe in mice experimentally infected with influenza virus. We investigated whether affinity maturation might nonetheless strongly select for particular amino acid motifs across diverse genetic backgrounds, but we found no evidence of convergence to similar CDR3 sequences or amino acid substitutions. These results suggest that although germline-encoded specificities can lead to similar immune responses between individuals, diverse evolutionary routes to high affinity limit the genetic predictability of responses to infection and vaccination.

Characterization of monoclonal antibodies against porcine epidemic diarrhea virus S1/S2 junction protein

Pig producers have faced considerable economic losses due to porcine epidemic diarrhea virus (PEDV) infection, emphasizing the need for PEDV antibody development. The S1/S2 junction (S1S2J) cleavage site of the S protein of PEDV is one of the major determinants of coronavirus infection success. In this study, we specifically selected the S1S2J protein of PEDV-AJ1102 (a representative strain of the G2 type) as a target protein to immunize mice and generated monoclonal antibodies (mAbs) using hybridoma technology. Three mAbs with high-binding activities to the S1S2J protein and were obtained and further analyzed. To reveal the characterization of these mAbs, variable region genes of antibodies were studied by using DNA sequencing, thereby revealing differences in their CDR3 amino acid sequences. We then developed a new method to identify the isotypes of these three mAbs. Results showed that these three antibodies were of the IgM type. As for the functions of these three mAbs, indirect immunofluorescence assay confirmed their good binding ability to Vero E6 cells infected with the PEDV-SP-C strain (G1 type). Epitope analysis showed linear epitopes for all three mAbs. These antibodies were also used to detect infected cells via flow cytometry analysis. In summary, we prepared and examined three mAbs against PEDV-S1S2J. These mAbs can be employed as detection antibodies for diagnostic reagents and further developed for other applications. We also designed a novel technique for easy and cost-saving identification of isotypes of mouse mAbs. Our results lay a good foundation for the development of research on PEDV.

A differential analysis of TCR repertoire between patients with EGFR-mutant and KRAS-mutant non–small-cell lung cancer.

e21012 Background: T-cell receptor (TCR) repertoire could partially represent tumor micro-environment. EGFR and KRAS are common driver genes mutated and often mutually exclusive in non-small cell lung cancer (NSCLC). There’s a hypothesis that the immune system responds differently according to different mutations. Here we investigated the TCR features, including diversity, clonality, evenness and sequence similarity in EGFR and KRAS mutated NSCLC patients, to better understand the immune response in carcinogenesis. Methods: A total of 71 EGFR-mutant and 31 KRAS-mutant NSCLC tissue samples underwent multiplex PCR based TCR-β sequencing. TCR-β cdr3 amino acid sequences were obtained from the sequencing data using MiXCR and VDJtools. To compare TCR repertoires in EGFR-mutant and KRAS-mutant groups, each assemble CDR3 amino acid sequence was taken as a word, and each repertoire was taken as a document. Each repertoire was compared with every other one as a sample pair. TF-IDF (term frequency-inverse document frequency) document vectors was built for each sample and cosine similarity scores were computed for sample pairs. Results: The total cosine similarity score within EGFR-mutant group is 11.07 and average cosine similarity score is 0.0044±0.01, while in the KRAS-group, the total cosine similarity score is 13.41 and average cosine similarity score is 0.029±0.04. Between these two groups, the total cosine similarity score is 8.76 and average cosine similarity score is 0.0019±0.0009. See table for summary. The intergroup similarity score is significantly less than that within each group respectively (p < 1E10). Conclusions: Our results demonstrated that the TCR repertoire was clearly associated with cancer mutational profile. The mechanism of how mutations affect the immune system and shape the TCR repertoire is yet to be investigated. [Table: see text]

Immunoaffinity biosensors for the detection of SARS-CoV-1 using screened Fv-antibodies from an autodisplayed Fv-antibody library

The detection of severe acute respiratory syndrome coronavirus (SARS-CoV-1) was demonstrated using screened Fv-antibodies for SPR biosensor and impedance spectrometry. The Fv-antibody library was first prepared on the outer membrane of E. coli using autodisplay technology and the Fv-variants (clones) with a specific affinity toward the SARS-CoV-1 spike protein (SP) were screened using magnetic beads immobilized with the SP. Upon screening the Fv-antibody library, two target Fv-variants (clones) with a specific binding affinity toward the SARS-CoV-1 SP were determined and the Fv-antibodies on two clones were named “Anti-SP1” (with CDR3 amino acid sequence: 1GRTTG5NDRPD11Y) and “Anti-SP2” (with CDR3 amino acid sequence: 1CLRQA5GTADD11V). The binding affinities of the two screened Fv-variants (clones) were analyzed using flow cytometry and the binding constants (KD) were estimated to be 80.5 ± 3.6 nM for Anti-SP1 and 45.6 ± 8.9 nM for Anti-SP2 (n = 3). In addition, the Fv-antibody including three CDR regions (CDR1, CDR2, and CDR3) and frame regions (FRs) between the CDR regions was expressed as a fusion protein (Mw. 40.6 kDa) with a green fluorescent protein (GFP) and the KD values of the expressed Fv-antibodies toward the SP estimated to be 15.3 ± 1.5 nM for Anti-SP1 (n = 3) and 16.3 ± 1.7 nM for Anti-SP2 (n = 3). Finally, the expressed Fv-antibodies screened against SARS-CoV-1 SP (Anti-SP1 and Anti-SP2) were applied for the detection of SARS-CoV-1. Consequently, the detection of SARS-CoV-1 was demonstrated to be feasible using the SPR biosensor and impedance spectrometry utilizing the immobilized Fv-antibodies against the SARS-CoV-1 SP.

The Longitudinal Analysis of Convergent Antibody VDJ Regions in SARS-CoV-2-Positive Patients Using RNA-Seq.

Severe acute respiratory syndrome-related coronavirus-2 (SARS-CoV-2) is an ongoing pandemic that continues to evolve and reinfect individuals. To understand the convergent antibody responses that evolved over the course of the pandemic, we evaluated the immunoglobulin repertoire of individuals infected by different SARS-CoV-2 variants for similarity between patients. We utilized four public RNA-seq data sets collected between March 2020 and March 2022 from the Gene Expression Omnibus (GEO) in our longitudinal analysis. This covered individuals infected with Alpha and Omicron variants. In total, from 269 SARS-CoV-2-positive patients and 26 negative patients, 629,133 immunoglobulin heavy-chain variable region V(D)J sequences were reconstructed from sequencing data. We grouped samples based on the SARS-CoV-2 variant type and/or the time they were collected from patients. Our comparison of patients within each SARS-CoV-2-positive group found 1011 common V(D)Js (same V gene, J gene and CDR3 amino acid sequence) shared by more than one patient and no common V(D)Js in the noninfected group. Taking convergence into account, we clustered based on similar CDR3 sequence and identified 129 convergent clusters from the SARS-CoV-2-positive groups. Within the top 15 clusters, 4 contain known anti-SARS-CoV-2 immunoglobulin sequences with 1 cluster confirmed to cross-neutralize variants from Alpha to Omicron. In our analysis of longitudinal groups that include Alpha and Omicron variants, we find that 2.7% of the common CDR3s found within groups were also present in more than one group. Our analysis reveals common and convergent antibodies, which include anti-SARS-CoV-2 antibodies, in patient groups over various stages of the pandemic.

T-Cell Repertoire Analysis in the Conjunctiva of Murine Dry Eye Model.

Dry eye is closely related to the activation and proliferation of immune cells, especially T cells. However, the determination of the preferential T-cell clonotypes is technically challenging. This study aimed to investigate the characterization of T-cell receptor (TCR) repertoire in the conjunctiva during dry eye. A desiccating stress animal model was established using C57/BL6 mice (8-10 weeks, female). After 7 days of stress stimulation, the slit-lamp image and Oregon-green-dextran staining were used to evaluate the ocular surface injury. Periodic acid-Schiff staining was used to measure the number of goblet cells. Flow cytometry was used to detect the activation and proliferation of T cells in the conjunctiva and cervical lymph nodes. Next-generation sequencing was used to detect the αβ TCR repertoire of the conjunctiva. The αβ TCR diversity increased significantly in the dry eye group, including the higher CDR3 amino acid length, marked gene usage on TCR V and J gene segments, extensive V(D)J recombination, and distinct CDR3 aa motifs. More important, several T-cell clonotypes were uniquely identified in dry eye. Furthermore, these perturbed rearrangements were reversed after glucocorticoid administration. A comprehensive analysis of the αβ TCR repertoire in the conjunctiva of the dry eye mouse model was performed. Data in this study contributed significantly to the research on dry eye pathogenesis by demonstrating the TCR gene distribution and disease-specific TCR signatures. This study further provided some potential predictive T-cell biomarkers for future studies.

493. Defining background shared antibody sequences between unrelated healthy individuals (public clonotypes) to support future studies on specific infectious disease related conditions.

Abstract Background Public clonotypes, antibodies against specific antigens in unrelated individuals that have genetic similarities, have been shown in a variety of infections, including SARS-CoV-2 and HIV. Likely, there are shared antibody responses between individuals for many infections. To explore antibody responses that would coincide with specific infectious diseases that may set off chronic illnesses, such as Multiple Sclerosis or Alzheimer's disease, defining the background shared clonotypes is needed to differentiate disease from normal background public clonotype responses. Methods Heavy chain variable sequences were retrieved from public biorepositories (Bioproject PRJNA486667) composed of 43 healthy persons, and two groups of HIV infected persons; 114 with broadly neutralizing antibodies and 91 without broadly neutralizing antibodies. We utilized the Immcantation package of software run on our SUNY Buffalo computational cluster. After PRESTo annotation, duplicate sequences were collapsed and sequences of only single counts were removed. Clonal groups were determined using ChangeO requiring IGHV, IGHJ, and CDR3 amino acid sequence to be perfectly matched. Figures and statistics were generated with immcantation, excel, and graphpad prism 8. Results 244850 heavy chain sequences from 43 healthy controls were compared for exact matches to predicted germline variable segment and CDR3 amino acid sequence and identified 0.23% as public clonotypes. Comparison to 205 HIV + individuals (a total of 1.4 million comparative sequences) showed that 2.35% of heavy chain sequences were seen in more than one individual. Generally, public clonotypes had shorter CDR3s (peak of 9 amino acids). VH 3-9, 3-30 and 4-34 were the most commonly used variable segments in public clonotypes. Common exact match CDR3 sequences using a variety of variable sequences, including an 11 amino acid CDR3 sequence motif, were also discovered. Conclusion This early work has identified several public clonotypes that are shared among subjects who are HIV positive and otherwise healthy people. Defining the sequences commonly seen between individuals can assist in specifying antibody responses specific to disease states from larger sequence databases. Disclosures All Authors: No reported disclosures.

Longitudinal Profiling of the T Cell Compartment in Patients with Chronic Lymphocytic Leukemia Treated with Venetoclax

Venetoclax (ven), an inhibitor of the BCL2 anti-apoptotic protein, has been proven remarkably effective in the treatment of patients with chronic lymphocytic leukemia (CLL), where the malignant B cells typically overexpress BCL2. BCL2 is also expressed by various bystander cells in the CLL microenvironment, including T cells. Hence, investigating the effects of ven treatment on T cells is highly relevant, especially considering that distinct T cell subpopulations exhibit different sensitivity to anti-apoptotic signaling blockade. Herein we interrogated the molecular and phenotypic profiles of the T cell compartment in 1 treatment-naïve and 15 relapsed/refractory CLL patients who received ven either as monotherapy (n=12) or in combination with an anti-CD20 antibody (obinutuzumab, n=1; rituximab, n=3). We studied peripheral blood samples collected prior to ven initiation and after 5 months of treatment, from which we isolated T cell subpopulations (CD3+ T cells n=8; CD4+ and CD8+ T cells, n=8). Longitudinal analysis of pre/post-treatment samples by flow cytometry (n=7) revealed a significant (p=0.01) decrease in the numbers of central memory (CD45RO+/CCR7+) CD4+ (21.8% vs 8.8%) and CD8+ cells (21.4% vs 9.7%) under ven; in contrast, a significant (p=0.01) increase of CD8+ effector memory cells (CD45RO+/CCR7-) was observed (27.8% vs 32.9%). Profiling of the T cell receptor beta (TRB) chain gene repertoire overtime (n=15) was performed by next generation sequencing (NGS). Clonotypes (i.e. TRBV-TRBD-TRBJ gene rearrangements with unique pairs of TRBV genes and identical CDR3 amino acid sequences) were calculated and clonality was estimated as the average cumulative frequency of the 10 most frequent clonotypes per sample (ACF-10). A significantly (p=0.01) more diverse repertoire was identified in CD4+ vs CD8+ cells at both pre- and post-treatment timepoints (average numbers of clonotypes pre-/post-treatment: CD4+ cells, 16,908/13,768; CD8+ cells, 4,437/3,633). Expanded clonotypes corresponding to oligoclonal expansions were documented in all examined subpopulations at both timepoints. However, significantly (p=0.01) more pronounced skewing was observed in CD8+ vs CD4+ cells (pre-/post-treatment ACF-10 values: 64.4%/70.6% vs 31.4%/42.2%, respectively). Repertoire comparisons overtime highlighted a fraction of T clonotypes that were retained after treatment in all examined subpopulations. Of note, repertoire conservation was significantly (p=0.01) greater in CD8+ vs CD4+ T cells [retained pre-treatment clonotypes: 7,102/31,063 (28%) vs 11,525/118,359 (12%), respectively]. Transcriptome profiling of CD4+ and CD8+ cells pre-/post-treatment was performed using the Oncomine Immune Response Research Assay (395 genes) in 5 cases attaining a complete response after treatment. While the transcriptomes of CD4+ cells remained essentially unaltered, significant differences were identified in CD8+ cells, where 19 genes were found to be downregulated (adj.p<0.05, fold change>2). These genes are implicated in functions such as immune response (e.g. TLR7, TLR9, CD40), regulation of immune system (e.g. HLA genes, CIITA, TNFAIP8) and lymphocyte activation pathways (e.g. ICOSLG, CEACAM1, CXCR5), with at least some of these having immunosuppressive functions (e.g. BTLA and CTLA4 encoding inhibitory receptors associated with T cell exhaustion). In conclusion, ven treatment led to increased numbers of CD8+ effector subpopulations along with prominent clonal expansions and transcriptional rewiring that could conceivably contribute to clinical response. Downregulation of immune-related genes implicated in core regulatory/activation pathways supports immune recovery by ven, offering a rationale for future combination strategies aiming to increase the depth of clinical response.

A Deep Learning Model for Accurate Diagnosis of Infection Using Antibody Repertoires

The adaptive immune receptor repertoire consists of the entire set of an individual's BCRs and TCRs and is believed to contain a record of prior immune responses and the potential for future immunity. Analyses of TCR repertoires via deep learning (DL) methods have successfully diagnosed cancers and infectious diseases, including coronavirus disease 2019. However, few studies have used DL to analyze BCR repertoires. In this study, we collected IgG H chain Ab repertoires from 276 healthy control subjects and 326 patients with various infections. We then extracted a comprehensive feature set consisting of 10 subsets of repertoire-level features and 160 sequence-level features and tested whether these features can distinguish between infected individuals and healthy control subjects. Finally, we developed an ensemble DL model, namely, DL method for infection diagnosis (https://github.com/chenyuan0510/DeepID), and used this model to differentiate between the infected and healthy individuals. Four subsets of repertoire-level features and four sequence-level features were selected because of their excellent predictive performance. The DL method for infection diagnosis outperformed traditional machine learning methods in distinguishing between healthy and infected samples (area under the curve = 0.9883) and achieved a multiclassification accuracy of 0.9104. We also observed differences between the healthy and infected groups in V genes usage, clonal expansion, the complexity of reads within clone, the physical properties in the α region, and the local flexibility of the CDR3 amino acid sequence. Our results suggest that the Ab repertoire is a promising biomarker for the diagnosis of various infections.

Multi-omics analysis of locally advanced esophageal cell squamous carcinoma treated with radio-chemotherapy.

e16106 Background: The incidence rate of esophageal cell squamous carcinoma (ESCC) increases every year worldwide. Although the molecular characteristics of locally advanced esophageal carcinoma have been extensively studied, TCR evolution dynamics before and after treatment in patients with ESCC undergoing radical radiotherapy and chemotherapy are unknown. Methods: In this study, we performed whole-exome sequencing (WES) and RNA sequencing (RNA-seq) on tumor tissues from 11 ESCC patients who received radiotherapy and chemotherapy. T cell receptor (TCR) sequencing was performed in 33 ESCC blood samples, which included three time points: pre-treatment, on-treatment and post-treatment. We compared TCR clonality and Shannon index among the pre-treatment, on-treatment and post-treatment ESCC patients using paired samples Wilcoxon test. The t test was uesd for statistical analysis comparing the clonality and clonal frequency between patients with mutant and wild-type gene. Results: The median TMB is 3.32 mutations/Mb (range 1-6 muts/Mb). We identified nine significantly mutant genes including TP53 (100%), MUC17 (55%), FLG (45%), FLG2 (36%), FOXN1 (27%), HIST1H1E (18%), VN1R1 (18%), RIPK3 (9%), and ZNF628 (9%). Clonality was significantly increased from pre-treatment to post-treatment (p = 0.008) and from on-treatment to post-treatment (p = 0.04). Two patients exhibited an continuous increase of TCR CDR3 (CHCLPAED_AGGGELFF) frequency in post-treatment compared to on-treatment patients. Three patients had elevated level of CDR3 amino acid sequence (CASSLDSNQPQHF) after treatment. To explore the effect of the mutant gene on the TCR clonal pattern, we found that the baseline clonality and clonal frequency in mutant FAT1 (MUT) were significantly lower than that in wild-type (WT) patients (p = 0.044 and p = 0.013, respectively), while they were not significant after treatment between FAT1-MUT and WT patients. Intriguily, in FAT1-MUT patients, the clonality and clonal frequency underwent therapy were higher than that at baseline. The results suggested that the FAT1-MUT patients might be sensitive to treatment and FAT1 might contribute to immune response upon radio-chemotherpay. During radio-chemotherapy, the clonality and clonal frequency in INTS1-MUT and LCT-MUT patients had the similar phenomenon to the FAT1-MUT patients. Interestingly, a higher clonal frequency was observed after radiotherapy and chemotherapy than that of pre-treatment (p = 0.0086) in patients with mutant KMT2D (an epigenetic modifier coding lysine methyl transferase 2D). Conclusions: TCRs is clonal expansion after radiotherapy and chemotherapy in ESCC, suggesting an immune-activated microenvironment after radio-chemotherapy. Our multi-omics analysis provided the basic TCRs dynamics and its genomic and epigenomic association during the radio-chemotherapy of esophageal cell squamous carcinoma.

Profiling the baseline performance and limits of machine learning models for adaptive immune receptor repertoire classification.

BackgroundMachine learning (ML) methodology development for the classification of immune states in adaptive immune receptor repertoires (AIRRs) has seen a recent surge of interest. However, so far, there does not exist a systematic evaluation of scenarios where classical ML methods (such as penalized logistic regression) already perform adequately for AIRR classification. This hinders investigative reorientation to those scenarios where method development of more sophisticated ML approaches may be required.ResultsTo identify those scenarios where a baseline ML method is able to perform well for AIRR classification, we generated a collection of synthetic AIRR benchmark data sets encompassing a wide range of data set architecture-associated and immune state–associated sequence patterns (signal) complexity. We trained ≈1,700 ML models with varying assumptions regarding immune signal on ≈1,000 data sets with a total of ≈250,000 AIRRs containing ≈46 billion TCRβ CDR3 amino acid sequences, thereby surpassing the sample sizes of current state-of-the-art AIRR-ML setups by two orders of magnitude. We found that L1-penalized logistic regression achieved high prediction accuracy even when the immune signal occurs only in 1 out of 50,000 AIR sequences.ConclusionsWe provide a reference benchmark to guide new AIRR-ML classification methodology by (i) identifying those scenarios characterized by immune signal and data set complexity, where baseline methods already achieve high prediction accuracy, and (ii) facilitating realistic expectations of the performance of AIRR-ML models given training data set properties and assumptions. Our study serves as a template for defining specialized AIRR benchmark data sets for comprehensive benchmarking of AIRR-ML methods.

Serial Analysis of the T-Cell Receptor β-Chain Repertoire in People Living With HIV Reveals Incomplete Recovery After Long-Term Antiretroviral Therapy.

Long-term antiretroviral therapy (ART) in people living with HIV (PLHIV) is associated with sustained increases in CD4+ T-cell count, but its effect on the peripheral blood T-cell repertoire has not been comprehensively evaluated. In this study, we performed serial profiling of the composition and diversity of the T-cell receptor β-chain (TRB) repertoire in 30 adults with HIV infection before and after the initiation of ART to define its long-term impact on the TRB repertoire. Serially acquired blood samples from 30 adults with HIV infection collected over a mean of 6 years (range, 1-12) years, with 1-4 samples collected before and 2-8 samples collected after the initiation of ART, were available for analysis. TRB repertoires were characterized via high-throughput sequencing of the TRB variable region performed on genomic DNA extracted from unsorted peripheral blood mononuclear cells. Additional laboratory and clinical metadata including serial measurements of HIV viral load and CD4 + T-cell count were available for all individuals in the cohort. A previously published control group of 189 TRB repertoires from peripheral blood samples of adult bone marrow transplant donors was evaluated for comparison. ART initiation in PLHIV was associated with a sustained reduction in viral load and a significant increase in TRB repertoire diversity. However, repertoire diversity in PLHIV remained significantly lower than in the control group even after long-term ART. The composition of TRB repertoires of PLHIV after ART also remained perturbed compared to the control cohort, as evidenced by large persistent private clonal expansions, reduced efficiency in the generation of TRB CDR3 amino acid sequences, and a narrower range of CDR3 lengths. Network analysis revealed an antigen-experienced structure in the TRB repertoire of PLHIV both before and after ART initiation that was quite distinct from the structure of control repertoires, with a slight shift toward a more naïve structure observed after ART initiation. Though we observe significant improvement in TRB repertoire diversity with durable viral suppression in PLHIV on long-term ART, the composition and structure of these repertoires remain significantly perturbed compared to the control cohort of adult bone marrow transplant donors.

Bruton's Tyrosine Kinase Supports Gut Mucosal Immunity and Commensal Microbiome Recognition in Autoimmune Arthritis.

Bruton’s tyrosine kinase (Btk) deficiency preferentially eliminates autoreactive B cells while sparing normal humoral responses, but has not been studied in mucosal immunity. Commensal microbes and intact BTK signaling have been independently shown to be essential for arthritis development in K/BxN mice. Here, we examine how BTK-mediated signaling interfaces with the gut microbiome. Btk-deficient K/BxN mice were found to have small Peyer’s Patches with reduced germinal center and IgA class-switched B cells. IgA-switched plasma cells in small intestines were reduced, especially in villi of Btk-deficient mice. IgH CDR3 sequencing showed similar V gene diversity and somatic hypermutation frequency despite Btk deficiency but showed reduced CDR3 amino acid polarity, suggesting potential qualitative differences in the gut plasma cell repertoire. Small intestinal IgA was low and IgA coating of commensal bacteria was reduced. IgA-seq showed a shift in small intestinal microbes that are normally IgA-coated into the uncoated fraction in Btk-deficient mice. Overall, this study shows that BTK supports normal intestinal IgA development in response to commensals.This manuscript was previously published as a preprint at: https://www.biorxiv.org/content/10.1101/2021.03.10.434762v2.

Stitchr: stitching coding TCR nucleotide sequences from V/J/CDR3 information.

The study and manipulation of T cell receptors (TCRs) is central to multiple fields across basic and translational immunology research. Produced by V(D)J recombination, TCRs are often only recorded in the literature and data repositories as a combination of their V and J gene symbols, plus their hypervariable CDR3 amino acid sequence. However, numerous applications require full-length coding nucleotide sequences. Here we present Stitchr, a software tool developed to specifically address this limitation. Given minimal V/J/CDR3 information, Stitchr produces complete coding sequences representing a fully spliced TCR cDNA. Due to its modular design, Stitchr can be used for TCR engineering using either published germline or novel/modified variable and constant region sequences. Sequences produced by Stitchr were validated by synthesizing and transducing TCR sequences into Jurkat cells, recapitulating the expected antigen specificity of the parental TCR. Using a companion script, Thimble, we demonstrate that Stitchr can process a million TCRs in under ten minutes using a standard desktop personal computer. By systematizing the production and modification of TCR sequences, we propose that Stitchr will increase the speed, repeatability, and reproducibility of TCR research. Stitchr is available on GitHub.

Comparison with Private and Public Clonotypes Reveals That, While CLL Type Stereotyped BCR Are Produced in 90% of Healthy Subjects from 18 to 78 Years, They Are Not Accumulated and Are Mostly Found in Immature and Naïve B-Cells

Introduction:One third of Chronic Lymphocytic Leukemia (CLL) patients harbor stereotyped B-Cell receptor (BCR). This is in contrast with the 1012 potential different BCRs that can be produced by the immune system. Some stereotyped BCR are also highly associated with prognosis. For example, the stereotyped subset #2 identifies aggressive CLLs with frequent SF3B1 and ATM alterations. This underlies an important role of BCR selection for emergence of CLL clones.Aim and methods:We raised the question of the existence and quantification of B-cells with stereotyped BCR of the CLL-type in a series of 69 healthy subjects (HS) from 18 to 78 years old, without any past of hematological history nor any detectable monotypic B-cells by flow cytometry.HS were divided in age three groups, adults (18 - 55 years, n= 31), senior-1 (55 - 65 years, n = 14) and senior-2 (> 65 years, n = 24).The variable region of the immunoglobulin heavy chain gene (IGHV) was amplified from genomic DNA of purified B-cells sequenced on an Illumina MiSeq and annotated with the online IMGT HighV-Quest tool.Clonotypes were defined as IGHV genes with the same rearranged V and J segment and with the same CDR3 amino acid sequence with one mismatch tolerance, and if a minimal threshold read number was reached (ranging 5 to 15 according to the DNA input and sequencing depth).Public clonotypes were those shared in between at least 4 (5%) HS.Detection of the 19 major stereotyped CLL's BCRs was performed with an in house R script. And was confirmed with the ARResT/AssignSubsets online tool (V Bistry et al, Bioinformatics, 2015).Exploration of the IGHV repertoire of the different B-cell subpopulations was done from public data (Mitsunaga, Mol Cell Proteomics, 2020).Results:To be independent of the number of reads and to compare HS to each other, clonotypic frequencies were ranked in 150 increasing intervals. In each case, the mean rank of clonotypic frequency ranged 3 to 7 with an SD in between 2 and 6 and was independent of the age category. A Z-score was calculated for each clonotype. Less than 1% of private clonotypes had a Z-score>1.96, while 10% to 30% of public clonotypes had a significant Z-score>1.96 (figure 1). This indicates that public clonotypes are often accumulated when compared to their private counterpart.A stereotyped CLL's BCR was found in 90% HS all stereotypes together. Subsets #2 and #5 were found in more than 50% of cases (figure 2). Frequencies of subsets #3, #59, #7C, #202, #14, #64B, decreased from 23% to 6% cases. Subsets #1, #4, #6, #8, #12, #16, #99 and #201 were never found.Frequencies of stereotyped BCRs ranged 5.2x10 -5 to 1.2x10 -4, being similar in the different HS age categories. Ranks of stereotyped BCR frequencies were almost always below the mean rank of private clonotypes, that indicates a very low abundance (figure 3). Junction analyses suggest that B-cells with stereotyped BCRs were polyclonal since they were very likely to derive from different B-cell precursors. Exploring the longitudinal repertoires of Mitsunaga and Snyder(Mol Cell Proteomics 2020), showed that stereotyped BCRs were almost restricted to immature and naïve B-cells and were fugace while public clonotypes were recurrently and frequently found in memory B-cells and could be stable over time for many of them.Conclusion:B-cells with stereotyped CLL BCRs were easily detected in healthy people whatever their age category. Remarkably, subsets #2 and #5 that are associated with aggressive CLL were found in more than 50% healthy donors at any age. Comparison with private and public clonotypic repertoire reveals that abundance of stereotyped BCR was almost always very low and was restricted to immature and naïve B-cells.Therefore, B-cells with stereotyped BCRs are produced throughout life in almost everybody. But these B-cells do not accumulate and do not seem to undergo to B-cell maturation.This suggests that stereotyped BCRs are not intrinsically prone to transformation and that clonal selection of CLL cells would involve additional events such as antigenic encounter or oncogenic alterations. [Display omitted] DisclosuresNo relevant conflicts of interest to declare.