GodoyLozano2016 - Lower IgG SHM Rates in Acute Dengue

Full citation: Godoy-Lozano EE, Téllez-Sosa J, Sánchez-González G, et al. Lower IgG somatic hypermutation rates during acute dengue virus infection is compatible with a germinal center-independent B cell response. Genome Medicine. 2016;8:23. doi:10.1186/s13073-016-0276-1

Raw file: [[raw/godoy-lozano2016.pdf]]

Summary

This study used high-throughput VH region cDNA sequencing (454-Roche pyrosequencing) of peripheral blood IgG B cells from 19 Mexican adults with acute laboratory-confirmed dengue (10 DWS−, 9 DWS+; 3 primary, 16 secondary). For 11 patients, paired post-convalescent samples (6 months later) were analysed for comparison. The authors developed a custom bioinformatics pipeline (ImmunediveRsity) to reconstruct heavy chain clonotypes and lineages, and used Monte Carlo simulation to estimate the probability of sampling memory B cells vs. antibody-secreting cells (ASCs) from unsorted IgG⁺ B cell RNA.

The central finding was paradoxically low somatic hypermutation (SHM) rates in acute-phase IgG B cells compared to post-convalescence — lower in DWS+ than DWS−, and lower in secondary than primary infections. Biased usage of IGHV1-2, IGHV1-18, and IGHV1-69 during acute infection was also identified, with polyclonal CDRH3 diversity within biased IGHV expansions. Convergent CDRH3 signatures (particularly 10-residue CDRH3s with the consensus ARQxGNWFDx) were shared across up to 52% of individuals during acute infection, at 8.3-fold higher prevalence than post-convalescence, and these convergent clones had even lower SHM than non-convergent clones.

The authors propose a dual-pathway model: concurrent GC-dependent and GC-independent (extrafollicular) B cell differentiation during acute dengue. The GC-independent pathway produces class-switched (IgG) but poorly mutated antibodies via an innate-like recognition system using germline-coded IGHV segments, potentially driven by TLR7-mediated BCR crosslinking and T-independent CSR. These hypomutated IgG antibodies are implicated as candidates for cross-reactive, poly-reactive, and ADE-competent antibodies.

Study Design

• Type: Prospective cohort with paired acute/convalescent samples
• Sample size: n=19 acute dengue (DWS− n=10, DWS+ n=9); n=11 paired post-convalescent (6 months); n=10 healthy influenza vaccination controls
• Setting: Veracruz, Mexico (endemic zone), 2010–2011 dengue seasons; acute samples days 1–9 post-symptom onset (median 4 DWS−, 6 DWS+); post-convalescent 128–323 days
• Population: Adults 16–50 years; predominantly secondary infections (84.2%); all primary cases were DWS−; predominant serotype DENV-2 (52.6%), DENV-1 (36.8%); 78.9% had cross-neutralizing antibodies to all 4 serotypes

Key Findings

• Overall lower SHM during acute dengue: Proportion of VH mutations (pM-VH%) was significantly lower during the acute phase than post-convalescence across all comparisons (p<0.001, Kruskal-Wallis with Dunn’s correction). This applied to total, non-synonymous, and synonymous mutations.
• SHM lower in DWS+ than DWS−: Among acute-phase patients, DWS+ had significantly lower SHM than DWS− (p<0.001), and among secondary cases specifically, DWS+ SHM was lower than DWS− (p<0.001).
• SHM lower in secondary than primary infections: Paradoxically, secondary infections showed lower SHM than primary infections during the acute phase (p<0.001) — the opposite of what memory recall through GC-matured cells would predict.
• Contrast with influenza vaccination: TIV vaccination at day 7 post-immunisation showed increased SHM relative to baseline — the opposite pattern to acute dengue, confirming that the dengue finding is not a general feature of acute immune activation.
• IGHV1-2 overrepresented in acute DWS+: Significant increase in IGHV1-2 usage during acute DWS+ vs. post-convalescence (p<0.01 by lineage frequency). IGHV1-2-using B cells had the most pronounced SHM reduction in DWS+ (p<0.001 vs. DWS− and vs. post-convalescent).
• IGHV1-69 overrepresented in acute DWS+: Significant increase in lineage frequency (p<0.001) and clonotype frequency (p<0.001). IGHV1-69 is recurrently implicated in innate-like “natural antibody” recognition across influenza A, hepatitis C, and HIV.
• IGHV1-18 overrepresented in acute DWS−: Significant increase in clonotype frequency in DWS− but not DWS+ (p<0.05).
• Polyclonal expansions within biased IGHV segments: Digital CDRH3 spectratyping confirmed that the IGHV1-2, IGHV1-18, and IGHV1-69 expansions were polyclonal (diverse CDRH3 repertoires), consistent with germline-encoded VH region recognition of DENV antigens rather than single-clone selection.
• Convergent CDRH3 signatures at 8.3-fold higher prevalence in acute infection: 3.6% of clonotypes shared similar CDRH3 in ≥3 individuals during acute infection vs. 0.44% post-convalescence. Two 10-residue CDRH3s (ARQFGNWFDS, ARQWGNWFDL) were shared in 10/19 individuals (52%). These convergent clones had significantly lower SHM than non-convergent clones (p<0.001).
• Monte Carlo simulation validates ASC origin of acute-phase reads: When plasmablasts constitute 56% of IgG⁺ B cells (per Wrammert2012), the probability of sampling a memory B cell lineage was 0.015 per single read and decreased for larger clonotypes. This supports that the SHM reduction reflects the ASC (plasmablast) repertoire specifically.
• Increased B cell diversity during acute infection: Higher lineage numbers, rarefaction richness, and Shannon-Weaver entropy during the acute phase, consistent with massive plasmablast mobilisation.
• No allelic variation bias: IGHV genotyping for IGHV1-2, IGHV1-18, and IGHV1-69 showed no correlation between allele type and segment expansion.
• Clonotype/lineage ratio reduced in acute dengue: Fewer lineages per clonotype during acute infection (p<0.001), consistent with lower SHM-driven diversification.
• PCA separates acute from post-convalescent and healthy controls: PC1 (76.3% variance) driven by mutation parameters clustered DWS+ acute, DWS− acute, post-convalescent, and healthy controls into four groups.

Methods Used

BCR Sequencing

Entities Mentioned

Plasmablast, IgG, TLR7

Concepts Addressed

Somatic Hypermutation, Extrafollicular Response, Germinal Center, Class Switch Recombination, Memory B Cell

Relevance & Notes

This is the first study to directly measure SHM rates in acute dengue IgG B cells using high-throughput BCR sequencing, filling the critical gap identified in wiki state.md (“SHM distribution of dengue acute-phase plasmablasts — partially addressed”). The finding of paradoxically low SHM — lower in secondary infections and in more severe disease — is the strongest molecular evidence to date for a GC-independent B cell response in dengue.

The data directly connect to multiple wiki themes:

1. Resolves the SHM benchmark question: Parameswaran2013 reported 4.4–6.9% V mutation in convergent CDR3-bearing cells from unsorted PBMCs (intermediate between EF and GC benchmarks). GodoyLozano2016 uses IgG-specific cDNA (enriching for plasmablasts during acute phase per Monte Carlo simulation) and finds a global SHM reduction — consistent with a substantial EF contribution. The low SHM in convergent CDRH3s from this paper corroborates and extends Parameswaran2013.
2. IGHV1-69 connection: Appanna2016 found VH1-69 enriched in dengue PB-derived mAbs (not MBCs). GodoyLozano2016 independently confirms IGHV1-69 overrepresentation at the repertoire level in DWS+ acute infection. IGHV1-69 is a canonical “innate-like” antibody segment with known cross-reactive specificities across flaviviruses and other enveloped viruses.
3. Supports the dual EF+GC model: The concurrent low-SHM (EF) and normal-SHM (GC-derived memory) populations align with the Ansari2025 finding of concurrent Tph (EF) and CXCL13 (GC) activity in acute dengue.
4. CSR decoupled from SHM: The observation that IgG class-switched B cells can carry near-germline mutation levels directly supports the extrafollicular CSR model where AID catalyses CSR independently of the iterative SHM cycles typical of GC reactions.
5. ADE implications: The authors explicitly propose that germline-coded, cross-reactive, poly-reactive IgG antibodies produced by the GC-independent pathway are candidates for ADE-mediating antibodies. This connects to the Tph→ADE link watch item.

Limitations:

• Uses unsorted IgG⁺ B cells (RNA), not flow-sorted plasmablasts — the Monte Carlo simulation infers ASC dominance but does not formally demonstrate it
• The 5ʹ RACE-PCR approach sequences IgG cDNA (all IgG subclasses together); isotype subclass resolution is not available
• No antigen-specific sorting — cannot confirm that low-SHM clones are DENV-specific
• Small sample (n=19), heavily skewed to secondary infections (only 3 primary, all DWS−)
• Mexican cohort only; generalisability to other populations unknown
• 454-Roche sequencing (read length ~450–500 bp) — now superseded by Illumina and 10x single-cell platforms
• No paired heavy-light chain data

Questions Raised

• Are the low-SHM IgG B cells in acute dengue DENV-specific? The polyclonal CDRH3 diversity within biased IGHV segments suggests germline-coded antigen recognition, but antigen specificity was not tested.
• Do the convergent CDRH3s (ARQFGNWFDS, ARQWGNWFDL) identified across 52% of individuals target the same DENV epitope? Their antigen specificity is unknown — the same gap as the Parameswaran2013 convergent CDR3s.
• Why is SHM lower in secondary than primary infections? The authors propose that the “innate-like” response is stronger (or less diluted by GC output) in secondary infection. An alternative: original antigenic sin rapidly activates cross-reactive memory clones with lower SHM, outcompeting higher-SHM, serotype-specific clones.
• Does the TLR7-mediated T-independent CSR proposed in Figure 9 actually occur in dengue? Endosomal DENV recognition by TLR7 is proposed but not experimentally demonstrated in this study.
• Are IGHV1-2- and IGHV1-69-using antibodies preferentially involved in ADE? The association between IGHV1-2 low SHM and disease severity (DWS+) is correlative, not causal.