GarciaBates2013 - Plasmablast Response and Dengue Severity

Full citation: Garcia-Bates TM, Cordeiro MT, Nascimento EJM, Smith AP, Soares de Melo KM, McBurney SP, Evans JD, Marques ETA Jr, Barratt-Boyes SM. Association between magnitude of the virus-specific plasmablast response and disease severity in dengue patients. J Immunol. 2013;190(1):80–87. doi:10.4049/jimmunol.1103350

Raw file: [[raw/garcia-bates2013.pdf]]

Summary

This study characterised the acute B cell response in a hospital-based dengue cohort in Recife, northeast Brazil (2004–2006), comparing primary and secondary DENV infections across mild (DF) and severe (DFC) clinical presentations. The authors performed multi-color flow cytometry on PBMC from 84 laboratory-confirmed dengue patients, plus OFI and healthy controls, to define B cell subset distributions, proliferation (Ki-67), apoptosis (active caspase-3), and activation (CD69, CD95).

The central finding is an extraordinary plasmablast expansion in severe secondary dengue (DFC): plasmablasts averaged 46% of total B cells (peak 87%) between days 4–7 of illness, significantly exceeding all other groups including primary DFC, secondary DF, and OFI. ELISpot confirmed that >70% of IgG-secreting cells in secondary DFC were DENV-specific, with 3-fold greater reactivity to the infecting serotype (DENV-3) than to heterotypic serotypes (DENV-1, DENV-2). Despite this massive virus-specific plasmablast expansion, there was no correlation between plasmablast frequency and serum PRNT₅₀ neutralizing Ab titers to any DENV serotype.

B cells in dengue patients showed increased proliferation (Ki-67⁺), activation (CD69⁺, CD95⁺), and apoptosis (active caspase-3⁺) relative to OFI controls, with apoptotic B cells approaching 60% in secondary DFC. The positive correlation between Ki-67 and caspase-3 suggests activation-induced cell death, paralleling the well-described T cell activation/apoptosis pattern in dengue.

Study Design

• Type: Prospective hospital-based cohort, cross-sectional analysis
• Sample size: n=84 laboratory-confirmed dengue (15 primary DF, 23 secondary DF, 18 primary DFC, 28 secondary DFC), n=15 OFI, n=10 healthy naive, n=13 healthy recovered
• Setting: Hospital cohort in Recife, northeast Brazil, 2004–2006; DENV-3 was the circulating serotype
• Population: Ages ≥5 years; mixed pediatric and adult (primary DF median age 7 years; secondary DFC median age 37 years); PBMC sampled at 1–9 days post-symptom onset; peak plasmablast analyses restricted to days 4–7
• Classification: DF vs. DFC (complicated dengue fever per Brazilian Ministry of Health criteria — includes neurologic signs, cardiopulmonary dysfunction, liver insufficiency, GI bleeding, plasma leakage, thrombocytopenia, leukopenia, or death); primary vs. secondary by acute/convalescent IgG/IgM serology

Key Findings

• Massive plasmablast expansion scales with severity in secondary infection: Plasmablasts (CD27⁺CD21⁻CD20⁻CD38⁺) averaged 46% of B cells in secondary DFC at days 4–7 (peak individual: 87%), vs. ~5% in OFI, ~10–15% in DF groups, and ~15% in primary DFC. Secondary DFC was significantly higher than all other dengue groups (p<0.05 to p<0.001).
• Plasmablast expansion window: Peak expansion occurred between days 4–7 post-symptom onset in both primary and secondary infections, with secondary responses substantially greater — consistent with anamnestic recall.
• >70% of plasmablasts are DENV-specific: ELISpot on secondary DFC samples (n=9) showed ~40,000 DENV-reactive IgG-secreting cells per 10⁶ PBMC, representing 72% of all IgG-secreting cells. Healthy naive controls had negligible IgG-secreting cells and zero DENV-reactive cells.
• Serotype cross-reactivity with infecting-serotype preference: Plasmablasts reacted with all three tested DENV serotypes (DENV-1, -2, -3) but had 3-fold greater reactivity to DENV-3 (the infecting serotype) than DENV-1 or DENV-2 (p<0.01, n=14 secondary DFC).
• No correlation between plasmablast frequency and neutralizing Ab: PRNT₅₀ titers to DENV-1, DENV-2, DENV-3, or DENV-4 did not correlate with plasmablast percentage at days 4–7, regardless of severity or infection history.
• B cell activation, proliferation, and apoptosis elevated in dengue: Ki-67⁺ B cells significantly increased in dengue vs. OFI (p<0.01 to p<0.001). Apoptotic (caspase-3⁺) B cells approached 60% in secondary DFC (p<0.01 vs. OFI and primary DF). CD69⁺ and CD95⁺ B cells elevated in secondary DF and both DFC groups. CD95 expression positively correlated with caspase-3 (r=0.60, p<0.0001).
• Naive B cell contraction in secondary DFC: Naive B cells (CD27⁻CD21⁺) fell to ~30% in secondary DFC (from ~50% in healthy and OFI), reflecting displacement by the massive plasmablast wave.
• B and T cell absolute counts preserved despite leukopenia: Leukopenia in DFC was primarily due to granulocyte loss; B and T cell absolute numbers were maintained.
• Minor influenza virus cross-reactivity: 0.7% of IgG-secreting cells in secondary DFC reacted with influenza virus (520 spots/10⁶ PBMC), possibly reflecting concurrent infection or bystander activation.
• Secondary infections were predominantly DENV-3 following prior DENV-1: PRNT profiles indicated primary infections with DENV-1 (constant DENV-1 titers), with DENV-3 as the infecting serotype in the current episode. Median interval from primary to secondary infection could extend >20 years (DENV-1 introduced to Brazil in 1986).

Methods Used

• Conventional Flow Cytometry — LSRII flow cytometer; panels for CD19, CD20, CD27, CD21, CD38, CD10, CD69, CD95, CD3, Ki-67 (intracellular), active caspase-3 (intracellular), LIVE/DEAD viability dye
• ELISpot — IgG capture; DENV-1 (Hawaii), DENV-2 (16681), DENV-3 (H87) whole virus antigen; influenza vaccine and BSA controls; no prior stimulation (spontaneous secretion)
• PRNT — plaque reduction neutralization test; DENV-1–4 Brazilian isolates; PRNT₅₀ on Vero cells

Entities Mentioned

• Plasmablast — CD27⁺CD21⁻CD20⁻CD38⁺; massive expansion in severe secondary dengue
• CD19 — B cell identification gate
• CD20 — lost on plasmablasts; used with CD38 to distinguish PB from activated memory
• CD27 — memory marker; CD27⁺CD21⁻ gate for plasmablast/activated memory identification
• CD21 — naive B cell marker; CD21⁺ on naive, CD21⁻ on plasmablasts and activated memory
• CD38 — CD38⁺ on plasmablasts; used with CD20 to distinguish PB (CD20⁻CD38⁺) from activated memory (CD20⁺CD38⁻/lo)
• CD10 — used to exclude immature/transitional B cells (CD10⁻ gate for mature B cells)
• IgG — DENV-specific IgG secretion by plasmablasts; serotype cross-reactive
• Double-Negative B Cell — “atypical memory” (CD27⁻CD21⁻) identified in gating scheme

Concepts Addressed

• Extrafollicular Response — massive plasmablast expansion without corresponding neutralizing Ab; severity association; cross-reactive antibody output
• Germinal Center — implied by memory B cell subsets; not directly assessed
• Somatic Hypermutation (implied — cross-serotype reactivity and anamnestic kinetics imply affinity-matured repertoire; SHM not directly measured)
• Class Switch Recombination — IgG-dominant plasmablast output in secondary infection
• Memory B Cell — naive B cell contraction, atypical memory identified

Relevance & Notes

This paper is a landmark for the wiki because it provides the first quantitative severity-stratified data on plasmablast frequencies in dengue, directly demonstrating that the magnitude of the plasmablast response scales with both secondary infection status and clinical severity. The 46% mean (87% peak) plasmablast frequency in secondary DFC is among the highest reported for any human infection.

Key connections to existing wiki content:

1. Confirms the neutralizing Ab paradox in dengue (cf. Woodruff2020, Ansari2025): Massive DENV-specific plasmablast expansion does not translate to higher neutralizing Ab titers. This is now demonstrated independently by three sources in the wiki: GarciaBates2013 (PRNT₅₀), Woodruff2020 (anti-RBD neutralization in COVID-19), and Ansari2025 (FRNT₅₀ in dengue). The consistency across infections strengthens the conclusion that EF-derived plasmablasts produce antibodies with non-neutralizing (or poorly neutralizing) specificities.

2. Gating strategy comparison: The GarciaBates2013 B cell panel (CD19⁺CD10⁻, then CD27/CD21 for naive/memory/atypical/PB, then CD20/CD38 for PB vs. activated memory) is the earliest dengue paper in the wiki to use a multi-subset B cell gating approach. It lacks CD11c, CXCR5, T-bet, and IgD — meaning DN1/DN2 distinction is impossible and the “atypical memory” (CD27⁻CD21⁻) population cannot be resolved by the Sanz2025 criteria.

3. Serotype cross-reactivity: The 3-fold preference for the infecting serotype (DENV-3) over heterotypic serotypes is notable because it differs from the Nicaraguan pediatric cohort (Zompi2012, already in the wiki as a raw PDF) which found higher cross-reactivity to the previous infecting serotype (consistent with original antigenic sin). The authors attribute this to the longer interval between primary and secondary infection in Brazilian adults (potentially >20 years) vs. Nicaraguan children (median age 10.5, shorter intervals). This age/interval effect on cross-reactivity is a novel observation.

4. Apoptosis data: The activation-induced apoptosis of B cells (Ki-67/caspase-3 correlation, CD95 expression) has not been reported in other dengue B cell studies in the wiki and may be relevant to the DN2 apoptosis resistance finding from Scharer2019.

Limitations noted by the authors:

• No direct measurement of ADE-relevant antibody function
• No cytokine correlation with plasmablast magnitude (IL-6, IL-10, IL-21, BAFF, APRIL tested but no association found)
• Cryopreserved PBMC (validated for ELISpot by Kyu et al. 2009)
• DENV-4 not tested in ELISpot (only DENV-1, -2, -3)

Questions Raised

• What is the functional quality (neutralizing, enhancing, autoreactive) of the antibodies produced by the massive plasmablast wave in secondary DFC? The disconnect between plasmablast magnitude and neutralizing titer suggests these may be predominantly non-neutralizing or poorly neutralizing — potentially ADE-competent.
• Why does the Brazilian adult cohort show infecting-serotype-dominant cross-reactivity while the Nicaraguan pediatric cohort shows previous-serotype-dominant reactivity? Is this an age effect, interval effect, or serotype-specific?
• Are the plasmablasts in secondary DFC derived from pre-existing cross-reactive memory B cells (recall) or from newly recruited naive B cells (de novo EF activation)? The anamnestic kinetics (days 4–7 peak in secondary > primary) favor memory recall, consistent with the Ansari2025 finding that Tph preferentially drive memory B cell → PB differentiation.
• Does the high apoptosis rate (60% caspase-3⁺ B cells) in secondary DFC represent a homeostatic mechanism limiting the plasmablast wave, or does it contribute to pathology through release of intracellular contents?
• What drives the massive expansion? The authors found no correlation with IL-10, TNF-α, IL-6, IL-21, BAFF, or APRIL serum levels. The Ansari2025 finding of Tph-derived IL-21 as the driver was published 12 years later — did the bulk serum assay miss the locally concentrated paracrine IL-21 from Tph cells?