Kwissa2014 - Monocytes Drive Plasmablast Differentiation in Dengue

Full citation: Kwissa, M., Nakaya, H. I., Onlamoon, N., Wrammert, J., Villinger, F., Perng, G. C., Yoksan, S., Pattanapanyasat, K., Chokephaibulkit, K., Ahmed, R., & Pulendran, B. (2014). Dengue virus infection induces expansion of a CD14⁺CD16⁺ monocyte population that stimulates plasmablast differentiation. Cell Host & Microbe, 16(1), 115–127. https://doi.org/10.1016/j.chom.2014.06.001

Raw file: [[raw/kwissa2014.pdf]]

Summary

This is the wiki’s first source to supply an extrinsic, receptor-level mechanism for the dengue plasmablast response — an answer to what drives the expansion that Wrammert2012 - Plasmablast Responses in Acute Dengue and others documented but did not explain. Using whole-blood transcriptomics on 28 acute secondary dengue patients, a rhesus macaque model, and human in vitro coculture with blocking antibodies, the authors show that DENV infection expands an intermediate CD14⁺CD16⁺ monocyte population that drives resting B cells to become plasmablasts.

The mechanism is nailed down by blockade: anti-BAFF, TACI-Fc and anti-IL-10 each reduced plasmablast differentiation, whereas blocking IL-6 or IP-10 (CXCL10) did not. This is the wiki’s first functional demonstration that the BAFF/APRIL axis contributes to a human antibody response in an infection — and it sits in productive tension with GarciaBates2013 - Plasmablast Response and Dengue Severity, which found no serum BAFF/APRIL correlation with plasmablast magnitude.

A second finding matters for the wiki’s severity axis: the acute dengue whole-blood transcriptome tracks viral load and duration of illness but does not discriminate DF from DHF at all.

Study Design

• Type: Systems-biology cohort study + non-human primate model + in vitro mechanistic/blockade experiments
• Sample size: 28 acute dengue patients (DF n=18, DHF n=10); 19 of them re-sampled at convalescence (DF n=13, DHF n=6); 9 healthy local young adults as controls. NHP: 5 rhesus macaques. In vitro: 4 independent experiments with 4 different healthy blood donors
• Setting: Siriraj Hospital, Bangkok, Thailand, 2009 season. Single acute blood collection between days 2–9 of symptoms; convalescence ≥4 weeks after discharge. Confirmed by serotype-specific RT-PCR, NS-1 test, IgG/IgM ELISA and DENV antigen-specific ELISpot
• Population: All secondary dengue. Uncomplicated DF and DHF — no DSS cases
• NHP model: 5 adult Indian rhesus macaques, DENV-2 strain 16681, 2×10⁷ pfu i.v.; blood at −7 d, 5 hr, and days 1, 3, 5, 7, 10, 14, 28; inguinal and axillary lymph nodes sampled
• Data availability: GEO accession GSE51808

Key Findings

Transcriptional response tracks virus, not severity

• Viral load ranged <100 to 9.37×10⁹ copies/ml; VL and plasma NS-1 antigen correlated positively with each other (p=0.0001, r²=0.4424)
• Both correlated inversely with duration of acute illness (VL p=0.0002, r²=0.4272; NS-1 p=0.0021, r²=0.3109)
• No defined clusters between acute DF and DHF could be detected by hierarchical clustering or PCA. Patients instead separated into high-VL (>10⁷) and low-VL (<10⁵) groups
• Type I IFN signalling was the top predicted upstream regulator of genes correlated with high VL and early disease. Top high-VL pathways: PRR recognition of bacteria and viruses, TREM1, NF-κB, RIG-I-like receptor antiviral innate immunity, IRF activation by cytosolic PRR, Toll-like receptor signalling, PKR in IFN induction, interferon signalling, IL-10 signalling, p38 MAPK, IL-6 signalling
• XBP-1 target genes were significantly enriched in patients with low VL (late illness). The authors read this as the unfolded protein response driving plasma cell differentiation, consistent with increased plasmablast numbers correlating with duration of illness
• Cell-type GSEA: neutrophil, monocyte, DC and macrophage signatures positively correlated with high VL; NK, B cell, CD8⁺ and CD4⁺ T signatures negatively correlated (i.e. associated with late, low-VL disease)

CD14⁺CD16⁺ monocytes expand; classical mDC-1 collapse

• Increased frequency of blood monocytes in high-VL patients, though no significant increase in absolute monocyte numbers
• Striking reduction in proportion and absolute number of BDCA-1⁺ mDC-1 at early illness; low mDC-1 correlated with duration of disease. No change in BDCA-1⁻ mDC-2, pDC or NK cells
• Monocyte-subset GSEA: CD14⁺CD16⁺ gene set enriched in dengue (NES=1.60, FDR q=0.001); CD14⁺CD16⁻ depleted (NES=−1.79, q<0.001); CD14^dim^CD16⁺⁺ depleted (NES=−1.54, q=0.005)
• Plasma cytokines elevated in acute dengue: IP-10 (CXCL10), MCP-1 (CCL2), MIP-1β (CCL4), IL-1ra, IL-10, eotaxin (CCL11), IL-6, IL-8
• MIP-1β correlated with the proportion of CD14⁺CD16⁺ monocytes (p=0.0008, r²=0.3650); IP-10 correlated with their absolute count (p=0.0037, r²=0.3013)

In vitro: DENV converts monocytes into a plasmablast-driving APC

• DENV-2 (MOI 1) drove >70% of monocytes to a CD14⁺CD16⁺ phenotype at 48 h. The TLR7/8 ligand R-848 did the same; the TLR4 ligand LPS did not
• DENV-infected monocytes upregulated CD206 (mannose receptor), CD115 (M-CSFR), CCR5, and high CD163 and CD169; electron microscopy showed dendrites and large cytoplasmic vacuoles resembling M-CSF-cultured cells
• They secreted MCP-1, IP-10, IL-6, IL-8 and IL-10 — but no IL-1β
• In 6-day coculture with allogeneic resting CD19⁺ B cells (+IL-2 +CpG), DENV-infected monocytes drove robust B cell proliferation; dividing cells were CD19⁺CD20^lo^, the differentiating-plasmablast pattern
• CD27⁺⁺CD38⁺⁺ plasmablasts rose to 22.8% of total B cells with DENV-infected monocytes vs 5.21% with MDDCs and 3.83% with B cells alone
• IgG and IgM secretion (but not IgA) was significantly higher than with MDDCs or controls

★ The blockade experiments — what is actually required

• Genes encoding BAFF and APRIL were increased in blood of high-viremia patients at early illness and correlated with the magnitude of the CD14⁺CD16⁺ population; APRIL protein trended higher in high-VL plasma; monocytes secreted both after DENV infection in vitro
• Anti-BAFF and TACI-Fc modestly diminished B cell proliferation and plasmablast differentiation; TACI-Fc significantly reduced IgM production
• Anti-IL-10 and anti-IP-10 both reduced B cell proliferation, but only anti-IL-10 significantly blocked plasmablast differentiation and IgM secretion
• Blocking IL-6 did not reproduce the effect
• Authors’ conclusion: DENV-infected monocytes drive plasmablast differentiation via BAFF/APRIL and IL-10

NHP model confirms the expansion, and localises it

• All 5 macaques developed viremia (5 hr–14 days), DENV-2-specific IgM, and skin haemorrhagic manifestations on days 5–7
• CD14⁺CD16⁺ monocytes expanded in blood at days 1–3 post-challenge
• In axillary lymph nodes, a 13-fold increase in absolute CD14⁺CD16⁺ cell number (0.32–0.44% of all LN cells in the three animals with the highest proportions; up to 11.1% of all LN monocytes in the other two)
• LN monocytes upregulated CD163 and CD169 (siglec-1) at day 3 — markers of subcapsular sinus macrophages, which sit adjacent to migratory B cells and can present viral particles across the subcapsular sinus floor

Methods Used

Conventional Flow Cytometry, In Vitro B Cell Stimulation, RNA Sequencing (whole-blood transcriptome on an Affymetrix Human U133 Plus 2.0 microarray — no sequencing was performed anywhere in the paper — read by GSEA/IPA cell-type deconvolution against monocyte-subset-specific gene sets)

Re-propagated 2026-09-06. RNA Sequencing was trimmed from this triad on 2026-09-04 because the paper is a microarray study and that page was framed as sequencing-only — which left real content homeless while the same page was quietly carrying microarray-derived entries from Stone2019 and Cancro2020. The page has now been rescoped to bulk transcriptome profiling and names each study’s platform, so this content has a home and the link is restored.

Also discussed but not separately updated: Serum Proteomics (plasma proteins were measured by Bio-Plex Pro 27-plex bead assay and by BAFF/APRIL/NS-1 ELISA — not the LC-MS/MS antibody-sequence identification that page describes), ELISpot (named once, in a list of confirmatory diagnostics, with the protocol referenced to Wrammert2012 - Plasmablast Responses in Acute Dengue; no ELISpot result is reported — the page could carry nothing from it), T-B Coculture Assay (the coculture is monocyte–B cell; no T cells are present in any culture in this study), FACS Sorting (no cells were sorted — monocytes and B cells were isolated by Miltenyi CD14/CD19 magnetic positive selection to ≥95% purity, and the FACSAria was used only as an analyser).

Entities Mentioned

Plasmablast, CD27, CD38, CD19, CD20, BAFF, APRIL, TACI, Type I Interferon, TLR7, Inflammatory Monocyte, IgG, IgM, IgA, XBP1

Discussed but not separately updated: CXCR3 (the receptor is never named in the paper — only its ligand CXCL10/IP-10 is measured and blocked; no chemokine receptor was stained on B cells), TNF-alpha (two mentions, both in the Introduction as background citing Cros 2010 / Wong 2011 on monocyte-subset cytokine output; no TNF-α result of the authors’ own is reported, and it is not among the analytes reported as elevated in the 27-plex plasma panel).

Concepts Addressed

Extrafollicular Response, Antibody-Dependent Enhancement, Dengue Severity Classification, Atypical B Cell Effector Output, Toll-like Receptor Signaling in B Cells

Relevance & Notes

What it adds to the wiki. Two things nothing else supplies.

1. An extrinsic driver for the dengue plasmablast response, demonstrated by blockade. The wiki previously had the magnitude of the dengue plasmablast response (Wrammert2012 - Plasmablast Responses in Acute Dengue, GarciaBates2013 - Plasmablast Response and Dengue Severity) and the T cell arm (Ansari2025 - Peripheral T Helper Subset Drives B Cell Response in Dengue, IL-21R-Fc blockade). Kwissa2014 supplies the innate/myeloid arm with its own blockade data. Together these are the wiki’s only two receptor-blockade experiments in dengue.
2. The first functional BAFF/APRIL evidence in the corpus. Before this ingest the BAFF axis had no entity page and its only mechanism was review-carried and murine (Cancro2020 - Age-Associated B Cells).

★ A three-way tension the wiki must hold open — see BAFF Contradictions.

• Cancro2020 - Age-Associated B Cells (review, mouse): ABCs express BAFFR and TACI yet are largely BAFF-independent, letting them outcompete follicular B cells for BAFF-regulated space.
• GarciaBates2013 - Plasmablast Response and Dengue Severity (dengue primary): serum BAFF/APRIL/IL-6/IL-10/IL-21 showed no correlation with plasmablast magnitude.
• Kwissa2014 (dengue primary): BAFF/APRIL transcripts correlated with the CD14⁺CD16⁺ population, and blocking BAFF/APRIL functionally reduced plasmablast differentiation — though only “modestly.”

These are not straightforwardly contradictory: GarciaBates measured serum protein against plasmablast frequency in vivo, Kwissa measured blood transcript plus in vitro blockade. A reasonable synthesis is that BAFF/APRIL are contributory but not obligate — which is exactly what Zumaquero2019 - IFN-gamma Programs T-bet-hi B Cells for ASC Differentiation independently found (removing BAFF decreased ASC numbers but was not required). Recorded as a contradiction rather than resolved, per Rule 4.

A contradiction the authors themselves flag. Xu et al. 2012 reported that CD163⁺ human macrophages stimulate CD138⁺⁺CD38⁺⁺ plasma cells through an IP-10- and IL-6-dependent mechanism. Kwissa2014 explicitly did not observe IL-6- or IP-10-dependent plasmablast stimulation. Xu 2012 is not ingested; this is recorded as the ingested source’s own reported disagreement, not as a wiki adjudication.

Bearing on the severity axis. That the whole-blood transcriptome separates high- from low-VL patients but produces no DF/DHF signal at all is a meaningful negative for anyone hoping severity has a clean blood-transcriptomic correlate. It sits alongside Narvaez2011 - Evaluating WHO Dengue Severity Classifications (κ=0.25 between schemes) and Morra2018 - Defining Warning Signs and Severe Dengue (within-scheme definitional heterogeneity) as a third, independent reason to distrust severity as a clean outcome variable. The authors note their cohort contained no DSS cases, which they acknowledge may explain the absence of a severity signature.

Relevance to ADE. The discussion notes monocytes are “broadly decorated with Fc-γ receptors,” the recognised substrate for Antibody-Dependent Enhancement in secondary infection, and that both CD14⁺ and CD14^dim^CD16⁺ subsets are susceptible to DENV infection with comparable efficiency. The paper does not itself test ADE.

Sensing pathway. The authors attribute DENV sensing in monocytes to RIG-I and MDA-5, noting monocytes weakly express TLR3 and do not respond to TLR3 agonists, while distinct monocyte subsets express TLR7 and TLR8 as ssRNA sensors. That R-848 (TLR7/8) but not LPS (TLR4) reproduced the CD14⁺CD16⁺ conversion is consistent.

Limitations.

• Single acute time point per patient across a wide day-2–9 window, so all kinetics are inferred cross-sectionally. This is the same day-of-sampling confound the council raised against Ansari2025 - Peripheral T Helper Subset Drives B Cell Response in Dengue.
• All patients were secondary infections — no primary-infection comparison.
• No DSS cases, so the severity null is a null for DF-vs-DHF only.
• The coculture is allogeneic (monocytes and B cells from different donors) with exogenous IL-2 and CpG, so an MLR component and a TLR9 signal are present alongside the monocyte effect.
• CD14⁺CD16⁺ numbers did not correlate with antibody titres, and plasmablasts were higher in low-VL late-stage patients — the authors attribute this to differing kinetics of the two cell types rather than to absence of a relationship.
• This paper measures plasmablasts, not DN/atypical B cells. Nothing here phenotypes the IgD⁻CD27⁻ compartment. Its relevance to the wiki’s spine is as an upstream driver of the same extrafollicular output.

Questions Raised

• Do DENV-infected CD14⁺CD16⁺ monocytes drive DN2 formation, or only plasmablast differentiation? BAFF/APRIL/IL-10 is a different signal set from the IFN-γ/TLR7/IL-21 triad that generates DN2 cells (Jenks2018 - DN2 B Cells and EF Pathway in SLE, Zumaquero2019 - IFN-gamma Programs T-bet-hi B Cells for ASC Differentiation). Nobody has run the coculture with the DN2 readout.
• IL-10 is required here for plasmablast differentiation, yet Ansari2025 - Peripheral T Helper Subset Drives B Cell Response in Dengue found anti-IL-10 gave only a −25% effect against IL-21R-Fc’s −60%. How do the myeloid and Tph arms weight against each other in the same patient?
• The type I IFN signature dominates early high-VL disease, and type I IFN is a known driver of T-bet⁺ B cells. Does the dengue IFN environment prime the DN2 pathway even though this paper did not look?
• If CD14⁺CD16⁺ monocytes accumulate 13-fold in lymph nodes and acquire CD169⁺ subcapsular-sinus-macrophage markers, is the relevant B cell interaction happening in the node rather than in blood — where all human dengue B cell sampling actually occurs?
• Why does serum BAFF/APRIL protein fail to correlate with plasmablast magnitude (GarciaBates2013 - Plasmablast Response and Dengue Severity) when blocking it functionally reduces differentiation here? Is serum concentration simply the wrong measurement for a locally delivered signal?