EFB Dengue Literature Review

BCR Sequencing

10 min read

BCR Sequencing

Overview

BCR (B cell receptor) sequencing refers to the sequencing of rearranged immunoglobulin heavy and/or light chain variable region genes to assess somatic hypermutation (SHM), CDR3 diversity, clonal relationships, and selection pressure (replacement:silent mutation ratios in framework vs. CDR regions). In 2007-era studies, this was performed by PCR amplification of VH family genes from sorted B cell populations followed by Sanger sequencing of individual clones. Modern approaches use high-throughput next-generation sequencing (NGS) of bulk or single-cell BCR repertoires.

Key Points from Literature

  • Wei et al. used VH3 family-specific PCR primers with a Cγ constant region primer to amplify IgG⁺ sequences from sorted CD27⁺ and CD27⁻ (DN) B cell fractions; PCR products were cloned into pCR4Blunt-TOPO and individual colonies sequenced by BigDye Terminator cycle sequencing (see Wei2007 - DN Memory B Cells in SLE).

  • Analysis of ~28 sequences per group revealed that DN IgG⁺ cells carry ~3.2% nucleotide mutation rate (healthy) and ~2.6% (SLE), compared with ~5.4% and ~5.1% for CD27⁺ IgG⁺ memory cells — supporting antigen experience in DN cells but at a lower level than conventional memory (see Wei2007 - DN Memory B Cells in SLE).

  • Replacement:silent (R:S) mutation ratios in CDR1 were substantially higher in DN cells (13 in healthy, 3.2 in SLE) than CD27⁺ cells (3.9, 9.6), while FR region R:S ratios were broadly similar — broadly consistent with antigen-driven selection but with possible differences in selection stringency (see Wei2007 - DN Memory B Cells in SLE).

  • Limitation: analysis restricted to VH3 family sequences only; small sample size (n=28 per group); Sanger-based single-clone approach lacks the clonal diversity information available from modern NGS repertoire sequencing.

  • Tipton2015 NGS approach (Illumina MiSeq, VH1–7): RNA isolated from sorted B cell populations (naive, IgD⁻ memory, CD138⁻ ASC, CD138⁺ ASC, and acN cells); ~2 ng RNA converted to cDNA; PCR amplified with VH1-VH7 FR1 primers and constant-region primers (Cα, Cµ, Cγ) carrying Illumina Nextera tags. Sequenced on MiSeq (500-cycle v2 or 600-cycle v3). After paired-end joining, quality/length filtering, and IMGT/HighV-QUEST alignment, 50,000 sequences per dataset retained for downstream analysis. This approach covers all VH families (vs. VH3-only Sanger), enabling VH segment usage analysis (e.g., VH4-34 enrichment) (see Tipton2015 - ASC Diversity and Origin in SLE).

  • Clonality metrics (D20 and D50): Sequences are assigned to clones using shared VH/JH rearrangements + identical HCDR3 length + Hamming identity >85% in HCDR3. D20 = number of clones accounting for 20% of total sequences (lower = more oligoclonal); D50 = same for 50%. SLE CD138⁻ ASCs: D20 ~199, D50 ~1,722 vs. vaccination: D20 ~21, D50 ~159. acN cells have D20 index 28.7-fold higher than healthy naive B cells — consistent with antigen-driven clonal expansion (see Tipton2015 - ASC Diversity and Origin in SLE).

  • Clonal connectivity analysis between sorted populations: By applying the same clonal assignment metric across simultaneously sorted populations (naive, acN, memory, ASC), the fraction of sequences shared between populations can be quantified as a “connectivity” score. This approach revealed that acN cells contribute up to 32.5% of sequences clonally connected to co-circulating ASCs in SLE (see Tipton2015 - ASC Diversity and Origin in SLE).

  • Phylogenetic tree construction with IgTree: Clonal sequences aligned to germline with Clustal X; genealogical trees built with IgTree (distance-method algorithm). This allows identification of putative common ancestors and ordering of diversification steps within complex clones containing cells from multiple populations (see Tipton2015 - ASC Diversity and Origin in SLE).

  • Integration with serum proteomics: NGS databases from sorted populations serve as reference databases for LC-MS/MS identification of serum antibody sequences, enabling direct linkage of serum proteins to specific B cell clonotypes (see Tipton2015 - ASC Diversity and Origin in SLE; see also Serum Proteomics).

  • Jenks2018 clonal connectivity across aNAV, DN2, and PC: BCR sequencing of FACS-sorted aNAV, DN2, and PC populations demonstrated clonal sharing between all three — the in vivo repertoire evidence for the rNAV → aNAV → DN2 → plasmablast developmental pathway. IgG mutation rates in DN2 cells are similar to co-circulating PCs but lower than SWM, arguing against DN2 derivation from memory cells and supporting de novo EF differentiation from naive precursors (see Jenks2018 - DN2 B Cells and EF Pathway in SLE, BCR sequencing of sorted populations).

  • Single-cell V(D)J via 10x Genomics Chromium in COVID-19: ASCs from an ICU patient were captured at 10,000 cells per sample on a 10x Chromium device (Next GEM Single Cell V(D)J Reagent Kits v1.1). Sequencing on Illumina NovaSeq (2×150 bp, >70% saturation at 5,000 reads/cell) yielded 5,338 cells and 2,017 clonotypes. Combined with bulk V(D)J from CD138-enriched ASCs of 2 additional patients, this revealed germline-dominant repertoire (>50% unmutated VH), balanced IgM/IgG1/IgA1, ongoing CSR, and oligoclonal expansions — the first single-cell ASC repertoire analysis in an acute viral infection through the EF pathway lens (see Woodruff2020 - EF B Cell Responses in COVID-19, 10x Chromium + Illumina NovaSeq and MiSEQ).

  • Lineage analysis tools: GLaMST (Grow Lineages along Minimum Spanning Tree) for lineage tree construction from single-cell V(D)J data. Circos plots for visualising contemporaneous IgM↔IgG1/IgA1 class switching connections. Both approaches complement the IgTree phylogenetics used in Tipton2015 (see Woodruff2020 - EF B Cell Responses in COVID-19).

  • FIRST DENGUE BCR REPERTOIRE DATA — 454 pyrosequencing of VH from unsorted PBMC gDNA: Parameswaran2013 amplified VH rearrangements from 100 ng PBMC genomic DNA using BIOMED-2 barcoded multiplex primers in 6 independent PCR reactions per sample. Sequencing on GS FLX (454/Roche) with dual runs; iHMMune-align for V/D/J assignment; median ~2,000 reads per sample (range 450–9,000). The use of gDNA ensures each B cell contributes a single copy regardless of Ig expression level, unlike RNA-based approaches where plasmablasts are overrepresented. This makes clonality estimates unbiased by expression level but precludes isotype resolution (see Parameswaran2013 - Convergent Antibody Signatures in Dengue, n=60 dengue + 16 controls, 454 pyrosequencing).

  • Replicate library sequencing for clonality estimation (P(collision)): By sequencing 6 independent PCR libraries from separate DNA aliquots of the same sample, sequences observed in >1 replicate are inferred to derive from sufficiently expanded clones. P(collision) — the probability that two randomly chosen B cells share clonal origin — is estimated from inter-replicate coincidences. This metric is independent of sequencing depth (validated by subsampling) and captures global B cell expansion without requiring antigen-specific pre-selection. P(collision) was significantly higher in acute vs. convalescent (p=0.0004) and acute vs. post-convalescent (p<0.0001) dengue, and higher in secondary vs. primary acute dengue (p=0.0409) (see Parameswaran2013 - Convergent Antibody Signatures in Dengue).

  • Convergent CDR3 identification by cross-validation: The 44-patient dataset was split into training and test sets; CDR3 amino acid sequences and their one-mismatch derivatives were screened for high prevalence in acute samples and low prevalence in post-convalescent/healthy samples. Six CDR3s (10-mers and 13-mers) were validated across training set, test set, an independent 16-patient cohort, and confirmed absent in >1,000 datasets from 640 non-dengue individuals. Four independent analytical approaches (mismatch cross-validation, BLOSUM62 similarity, association testing with FDR correction, L1-regularized logistic regression) yielded overlapping CDR3 candidates, with the cluster-based logistic regression achieving AUC 0.834 for dengue classification (see Parameswaran2013 - Convergent Antibody Signatures in Dengue).

  • Convergent CDR3s encoded by multiple V gene families — evidence of true convergent evolution: The most prevalent CDR3 (ARLD(Y)₅GMDL) was encoded by 6 distinct V genes from V1, V3, and V5 families across individuals. Nucleotide sequences underlying the same CDR3 amino acid sequence differed across individuals, with synonymous codon variation and diverse V/D/J usage. This rules out PCR contamination or shared germline alleles and establishes independent derivation through convergent evolution under antigen selection (see Parameswaran2013 - Convergent Antibody Signatures in Dengue).

  • Combined Sanger single-cell and 454 pool sequencing of sorted B cells in dengue: Appanna2016 used two complementary BCR sequencing approaches on sorted populations: (1) Sanger sequencing of single-cell VH/VL from plasmablasts and MBCs (RT-PCR from 96-well plates), enabling paired heavy/light chain cloning and mAb expression; (2) 454 GS-FLX Titanium pyrosequencing (250,000 reads/sample) of pool-sorted populations using iRepertoire primers covering V and C regions. CDR3 clones were defined by same V/D/J usage, same length, and ≥85% sequence identity. Clonal overlap analysis (Venn diagrams) revealed very few shared CDR3s between PBs and DENV-binding MBCs — the first direct clonal comparison of these two compartments in dengue. MiXCR v1.8.2 was used for V/D/J assignment and isotype extraction from 454 data (see Appanna2016 - Plasmablasts as Subset of Memory B Cell Pool, n=12 dengue, FACSAria sorting → Sanger + 454).

  • 454-Roche IgG VH cDNA sequencing with ImmunediveRsity pipeline in dengue: GodoyLozano2016 used 5ʹ RACE-PCR with an IgHG CH1 exon-specific antisense primer to generate VH libraries from total peripheral blood RNA (unsorted IgG⁺ B cells). Sequencing on GS FLX Titanium (454-Roche, ~450–500 bp reads, 3ʹ→5ʹ direction to prioritise CDRH3 quality). 2.36 million raw reads yielded 2.04 million pass-filter sequences, reconstructed into 146,565 clonotypes and 385,206 lineages across 30 libraries (19 acute + 11 post-convalescent). The custom ImmunediveRsity pipeline performs IgBLAST V(D)J assignment, CDRH3-based clonotype clustering (97% identity), lineage consensus (99.5% identity), Shannon entropy/rarefaction diversity estimation, and SHM quantification via IMGT/HighVQuest. IGHV usage was expressed at three levels: raw read proportion (relative transcription), clonotype frequency, and lineage frequency — enabling discrimination of transcription-level bias (high Ig expression per cell) from clonal expansion. Hierarchical clustering of ΔA−Pc IGHV usage identified patient groups correlated with clinical (DWS−/DWS+) and immune (primary/secondary) status (see GodoyLozano2016 - Lower IgG SHM Rates in Acute Dengue, n=19 acute + n=11 post-convalescent, 454-Roche).

  • Monte Carlo simulation for B cell subpopulation deconvolution: A second-order Monte Carlo simulation estimated the probability of sampling memory B cells vs. ASCs in unsorted IgG⁺ repertoire data, modelling individual variation in mB:ASC ratios and differential Ig transcription (ASCs ~12-fold higher than mB cells, gamma-distributed). At the expected acute dengue plasmablast proportion (56% of IgG⁺ per Wrammert2012), the probability of sampling a single read from a memory B cell was 0.015. During post-convalescence (mB ~95.8%), lineages ≥30 reads were estimated to derive exclusively from ASCs (p = 2.06E-06 for mB origin). This approach enables in silico sorting of repertoire data by Ig expression level, partially compensating for the lack of physical cell sorting (see GodoyLozano2016 - Lower IgG SHM Rates in Acute Dengue).

  • De novo convergent CDRH3 identification: Beyond testing against previously published signatures (Parameswaran2013), GodoyLozano2016 identified de novo convergent CDRH3s by screening for identical amino acid sequences shared in ≥3 individuals. Found 1,365 clonotypes representing 269 identical CDRH3s; 70% had a predominant 10-residue length. Two CDRH3s (ARQFGNWFDS, ARQWGNWFDL) were shared in 10/19 (52%) of individuals. Convergent clones carried significantly lower SHM than non-convergent clones (p<0.001) (see GodoyLozano2016 - Lower IgG SHM Rates in Acute Dengue).

Contradictions & Debates

None documented in current wiki sources.

FACS Sorting, Somatic Hypermutation, Double-Negative B Cell, DN2 B Cell, Activated Naive B Cell, Memory B Cell, Plasmablast

Sources

Graph View

Backlinks