EFB Dengue Literature Review

Why DN B Cells Matter - Disease Relevance and Infectious Disease Case

25 min read

Why DN B Cells Matter — Disease Relevance and the Infectious Disease Case

Research Question

What does the wiki’s ingested literature actually establish about why DN (IgD⁻CD27⁻) / atypical B cells matter in disease — and specifically, what is their demonstrated usefulness in an infectious disease context?

Sources Used

Tier A — ingested primaries with original DN/atypical-phenotype data

Tier B — zero-original-data reviews/opinion; the wiki holds these claims only through the review’s characterization of papers it has never read

Supporting — antibody-output and clinical-classification layer (not DN-phenotype papers)

Coverage note — partially stale; read the dates. Beckers2023 - Origins and Functions of DN B Cells was added 2026-08-26 and is cited, bringing cited sources to 22. Two sources ingested since the original sweep have NOT been re-assessed against this question — Posadas-Mondragon2020 - TLR Polymorphisms and Clinical Forms of Dengue and Wang2006 - Flavivirus Activation of pDCs and TLR7 Signaling — both TLR papers with no DN-compartment measurement, so they are unlikely to change the argument, but neither has been checked. The corpus now stands at 31 sources. Original note (2026-08-18) follows: All 24 wiki sources were assessed; 21 are cited above. The three assessed-and-not-cited are Wrammert2012 - Plasmablast Responses in Acute Dengue, Parameswaran2013 - Convergent Antibody Signatures in Dengue and Bhattacharya2016 - Memory B Cell Subset Selection in Secondary Dengue — all plasmablast/BCR-repertoire work with no DN-compartment measurement and no bearing on this question. The sources: frontmatter counts cited sources, not corpus coverage.

Synthesis

0. Definition guard — fix the referent before aggregating anything

Claims about “DN cells” travel badly between papers. Three hazards must be cleared first.

  1. DN here means IgD⁻CD27⁻ (Double-Negative B Cell), subdivided DN1/DN2/DN3 (+DN4 per Lamprinou2026 - ABCs and DN B Cells). The functionally interesting cell is DN2 B Cell (CXCR5⁻CD21⁻CD11c⁺T-bet⁺FCRL5⁺).
  2. ABC ⊃ DN2 asymmetrically. The ABC population contains CD27⁺ and IgD⁺ cells; only its IgD⁻CD27⁻ fraction maps to DN2, and even there ABC ≠ DN2 transcriptomically (Lamprinou2026 - ABCs and DN B Cells, opinion, citing Maul 2021). “ABC evidence” is not automatically “DN evidence.”
  3. ⚠ The “DN MBC” false friend. In the murine memory literature “DN” = CD80⁻PD-L2⁻ — an origin proxy describing a quiescent, low-SHM, largely unswitched IgM⁺ cell biased toward secondary GC re-entry. That is close to the inverse of the activated CD21⁻CD11c⁺ effector tracked here (Glaros2025 - Multilayered Identity of B Cell Memory, review). Functional claims must not be transplanted across the two axes.

Beyond nomenclature, Sanz2025 - Human Atypical B Cells Overview argues the “atypical” label conflates at least five distinct populations, and Cancro2020 - Age-Associated B Cells supplies independent murine confirmation: within the CD21⁻CD23⁻ splenic gate only ~2/3 of cells are T-bet⁺ and ~half of those CD11c⁺ — ≥3 populations inside one “ABC” gate.

1. The shape of the evidence — read this before the arguments

SourceContextDN-phenotype original data?Study type / nTier
Woodruff2020 - EF B Cell Responses in COVID-19COVID-19Yes — DN1/DN2/DN3, DN2:DN1 ratioprospective cohort; 17 COVID (10 ICU/7 outpt), 17 HD, 24 retrospective HD, 7 SLE; 24-marker spectral FCM; scV(D)J n=1A
Kaneko2020 - GC Loss and TFH Block in COVID-19COVID-19Yes — DN2/DN3 in blood + DN localised in tissuepost-mortem tissue n=11 + controls; blood n=68; 13-colorA
Sutton2021 - Alternative Lineage B Cells in Vaccination and Infectionmalaria / vaccinationYes — atBC1–3 + MBC1 (scRNA-seq/CITE-seq)core 10x n=4; Smart-seq2 n=11 (163 cells); flow n=18; PfSPZ n=15; influenza n=9A
Ansari2025 - Peripheral T Helper Subset Drives B Cell Response in DenguedengueYes — CD21⁻CD11c⁺ within IgD⁻CD27⁻prospective cohort; n=170 acute, 94 HD, 32 conv; scRNA-seq n=3A
Singh2026 - DENV-Specific Memory B Cell SubsetsdengueYes — DENV-specific CD27⁻CD21⁻ atypical MBCs58 samples / 18 pediatric; longitudinal arm n=4/group; bioRxiv preprintA⁻
Jenks2018 - DN2 B Cells and EF Pathway in SLESLE (+8 HIV comparators)Yes — the DN2 definition itselfSLE-1 n=40, SLE-2 n=50, HCD n=21, RA n=15, HIV n=8; FCM + in vitroA
Wei2007 - DN Memory B Cells in SLESLEYescross-sectional; 36 SLE, 29 HC, 45 RA, 7 chronic HCVA
Tipton2015 - ASC Diversity and Origin in SLESLEYes5 SLE flare (NGS) + vaccine comparatorsA
Scharer2019 - Epigenetic Programming in SLE B CellsSLEYesRRBS/ATAC/RNA; 9 SLE + 12 HCA
Anolik2004 - Rituximab and B Cell Abnormalities in SLESLEYesphase I/II trial; n=17 analysableA
William2002 - Extrafollicular Somatic Hypermutation in Autoimmune Micemurine autoimmunityEF site, not DN gate8 mice, 305 sequencesA
Sanz2025 - Human Atypical B Cells Overviewcross-diseaseNo — reviewzero original dataB
Cancro2020 - Age-Associated B Cellscross-diseaseNo — review, predominantly murinezero original data; 185 refsB
Glaros2025 - Multilayered Identity of B Cell Memorycross-diseaseNo — reviewzero original data; 285 refsB
Lamprinou2026 - ABCs and DN B Cellscross-diseaseNo — opinion, self-citedzero original dataB

What this table says, plainly:

  • Only three non-dengue infection primaries exist in the wiki — Woodruff2020, Kaneko2020, Sutton2021. Two are COVID-19; the malaria one has a core n=4.
  • In dengue the entire DN evidence base is two papers, one a preprint using a CD27⁻CD21⁻ gate rather than IgD⁻CD27⁻ — so the two dengue measurements are not quite of the same cell. The other six dengue primaries characterise plasmablasts and BCR repertoire, not DN.
  • ★ The most quotable infectious-disease exemplars are Tier B. Everything the wiki holds on HIV (Moir’s “tissue-like memory”), malaria atypical memory (Weiss 2009), LCMV, γHV68, Ehrlichia, influenza, TB, HCV arrives through the four zero-data reviews. No HIV primary, no TB source, no influenza primary, no Ebola source has been ingested — and per External Citation Audit, only 1 of ~89 external papers has been independently verified corpus-wide. Grant text leaning on “atypical B cells are established in HIV and malaria” is currently leaning on a review’s paraphrase.
  • ⚠ Ebola appears exactly once in the entire wiki — an Overview line on Double-Negative B Cell with no source attribution. Treat as a bare assertion pending correction.

2. Eight arguments the wiki supports for “why DN cells matter”

(1) Magnitude — a major compartment, and one that pre-exists disease

SettingDN / atypical burdenSource
Healthy donors (flow)4.6 ± 1.8% of CD19⁺, always <10%Wei2007 - DN Memory B Cells in SLE, n=29 healthy controls, cross-sectional
Healthy donors (transcriptomic)~20% of B cells are alternative lineageSutton2021 - Alternative Lineage B Cells in Vaccination and Infection, n=4 CITE-seq
ICU COVID-19DN 19.3% of CD19⁺ vs 3.0% HD; DN2 = 80.3% of DN in a representative ICU patient vs 9.5% HDWoodruff2020 - EF B Cell Responses in COVID-19
Active SLEDN >10% of CD19⁺ in half of patients; DN2 up to 70% of blood CD19⁺Wei2007 - DN Memory B Cells in SLE, n=36; Jenks2018 - DN2 B Cells and EF Pathway in SLE
Acute dengueCD21⁻CD11c⁺ within IgD⁻CD27⁻ significantly expanded vs HD and convalescenceAnsari2025 - Peripheral T Helper Subset Drives B Cell Response in Dengue, n=170

The healthy-donor row is the one that changes the argument: infection amplifies this compartment, it does not create it. DN2/DN3 expansion is a shift in an existing population’s activation state, not the appearance of an aberrant cell.

(2) ★ They are antigen-driven participants, not inflammatory bystanders

This is the argument that licenses everything downstream, and it has direct proof:

  • Dual-fluorophore RBD probes show the DN2 and DN3 gates contain SARS-CoV-2-specific cells (Kaneko2020 - GC Loss and TFH Block in COVID-19, blood n=68, 13-color). The coordinated EF expansion is an antigen-directed response, not a non-specific inflammatory epiphenomenon.
  • ABC frequencies fall in patients with CD40/CD40L mutations, ABCs do not develop in fixed-BCR mice, and they sit within the antigen-specific pool after immunization (Glaros2025 - Multilayered Identity of B Cell Memory, review) — antigen-experienced and at least partly T-help-dependent.
  • Dengue: DENV-specific atypical MBCs at acute/3M temporally correlate with later class-switched and activated MBC levels, “implying functional responsiveness rather than exhaustion” (Singh2026 - DENV-Specific Memory B Cell Subsets, n=18 pediatric).

(3) ★ They are not exhausted — the reversal that reframed the infection literature

Worth stating explicitly, because the field’s first infectious-disease framing of these cells (HIV, malaria) was exhaustion — and the wiki carries three independent demolitions of it:

  1. SLE DN2 are FCRL4⁻FCRL5⁺ with intact proximal BCR signalling (BLNK phosphorylation after anti-IgG), unlike FCRL4⁺ HIV exhausted memory (Jenks2018 - DN2 B Cells and EF Pathway in SLE; see FcRH4).
  2. HIV “tissue-like memory” cells are BCR-hyporesponsive but proliferate robustly to TLR9 — “initially read as exhausted” (Cancro2020 - Age-Associated B Cells, review).
  3. The hyporesponsiveness is antigen-form-specific: atypical MBCs respond robustly to membrane-associated/immune-complexed antigen but not soluble antigen, because FcγRIIB (and per the review FCRL5) must be physically excluded from the immune synapse. The foundational “poor PC differentiation” results (Portugal 2015, Sullivan 2015 — both malaria) used soluble anti-Ig (Glaros2025 - Multilayered Identity of B Cell Memory, review, citing Ambegaonkar 2020 — the one externally verified reference in the wiki).

Sanz2025 - Human Atypical B Cells Overview states it flatly: DN2/ABC are not exhausted or anergic.

⚠ Assay-design consequence for dengue: soluble DENV E/NS1 restimulation would under-report exactly this population’s capacity.

(4) They are effectors — antibody, cytokine, and antigen presentation

  • DN2 → plasma cell without BCR stimulation or extensive division on TLR7+IL-21+IFN-γ; IgG output per cell equivalent to switched memory; produces anti-Sm, anti-RNP, anti-Ro at titres comparable to switched memory (Jenks2018 - DN2 B Cells and EF Pathway in SLE, in vitro n=5, LIPS). Clonal connectivity aNAV→DN2→PC confirmed by BCR sequencing.
  • Epigenetically, DN2 is the closest non-ASC subset to ASC by DNA methylation phylogeny — in both SLE and healthy controls (Scharer2019 - Epigenetic Programming in SLE B Cells, RRBS, 9 SLE + 12 HC). DN2 also uniquely lacks G2/M-checkpoint and apoptosis enrichment by GSEA — a proposed expansion mechanism.
  • Powerful antigen-presenting cells, potentially inducing TFH and sustaining GC responses; excessive APC activity is proposed to drive abnormal TFH regulation (Sanz2025 - Human Atypical B Cells Overview, review).
  • Cytokine effectors: on TLR7/9 activation ABCs produce high IFN-γ and exceptionally high IL-10 plus IL-6, and skew naive CD4 toward Th17; ABC-derived TNF-α suppresses B lymphopoiesis (Cancro2020 - Age-Associated B Cells, review, mouse).

(5) They mark GC failure and an alternative route to humoral immunity

  • Post-mortem COVID-19 LN/spleen: Bcl-6⁺ GC B cells and GC-TFH near-absent (LN p<0.001, spleen p<0.01) with AID⁺ B cells preserved and FDC networks intact — a lymphocyte differentiation block, not stromal destruction; TNF-α implicated (Kaneko2020 - GC Loss and TFH Block in COVID-19, n=11 tissue). IgD⁻CD27⁻ DN cells sit at both follicular and extrafollicular sites with T–B conjugates.
  • Murine proof that EF sites hypermutate at GC-comparable rates (~0.3 mut/gene/generation) (William2002 - Extrafollicular Somatic Hypermutation in Autoimmune Mice, Science).
  • Repertoire corroboration: >50% of COVID-19 EF ASC clonotypes carry entirely germline VH with contemporaneous IgM→switched connections (Woodruff2020 - EF B Cell Responses in COVID-19, scV(D)J n=1 + bulk n=2). Dengue analogue: paradoxically low IgG SHM in acute infection, lower in secondary than primary and lower in DWS+ than DWS− (GodoyLozano2016 - Lower IgG SHM Rates in Acute Dengue, n=19).
  • ★ The endotype has prognostic content: ~75% of SLE patients cluster as EF or memory endotype, with higher severity and nephritis concentrated in the EF cluster; the EF endotype predicts reduced affinity maturation and reduced neutralizing activity of SARS-CoV-2 vaccine responses (Sanz2025 - Human Atypical B Cells Overview, review, citing Jenks 2021 / Faliti 2024). This is the clearest statement in the wiki that EF/DN2 dominance costs you antibody quality.

⚠ Origin proxies have weakened: CSR happens mostly pre-GC and SHM loads overlap in both directions between GC-derived and GC-independent memory (Glaros2025 - Multilayered Identity of B Cell Memory) — isotype or mutation load alone no longer infers pathway.

(6) They are the substrate of low-fidelity antibody — the wiki’s own thesis

Scope note. The autoantibody ↔ neutralizing-antibody synthesis is canonical to ../bridge-wiki/; this page records the cellular side only and does not build standalone autoantibody content (CLAUDE.md Domain Context). The cells→autoantibody link is SLE-imported and unproven in dengue.

(7) They form durable memory, not just transient effectors

  • Antigen-specific DN2 and DN3 persist >1 year after SARS-CoV-2 mRNA vaccination, accounting for >50% of all spike/RBD⁺ cells (Sanz2025 - Human Atypical B Cells Overview, review, citing Faliti 2024). If this holds, the compartment is a major fraction of durable antiviral memory — not a disposable effector burst.
  • MBC1 sits at the base of the alternative-lineage pseudotime branch as quiescent memory — transcriptomic support for “memory DN2” (Sutton2021 - Alternative Lineage B Cells in Vaccination and Infection).
  • Dengue: DENV-specific atypical MBCs are durable to 18 months and higher in secondary infection (p<0.05 at 18M), with a significant infection-history effect by RM-ANOVA (p<0.01) — while resting MBCs (CD27⁺CD21⁺) do not differ (Singh2026 - DENV-Specific Memory B Cell Subsets). The atypical compartment is where repeat-exposure history is written.

(8) They are therapeutically tractable

FCRL5 and SLAMF7 are preferred selective-depletion surface targets (SLAMF7-directed therapy already approved in myeloma); indirect routes include TLR7 inhibition and IFN-γ blockade, with the stated attraction that selective ABC depletion could spare protective memory and long-lived plasma cells (Sanz2025 - Human Atypical B Cells Overview). Rituximab-mediated depletion and reconstitution resolves DN expansion (P=0.05) — in effective depletors only (Anolik2004 - Rituximab and B Cell Abnormalities in SLE, n=17); belimumab-associated loss of ABC-phenotype cells correlates with therapeutic response (Cancro2020 - Age-Associated B Cells, review, citing Wang 2018). Dengue-specific: IL-21R-Fc blockade reduces plasmablast output ~60% in coculture (Ansari2025 - Peripheral T Helper Subset Drives B Cell Response in Dengue).

⚠ Murine ABCs resist anti-CD20/anti-BLyS while human SLE ABCs are sensitive — an unresolved species-vs-context contradiction (Age-Associated B Cell, Contradictions).

3. ★ The counterweight the “why they matter” case usually omits — they can be protective

Framing DN/atypical cells purely as pathology drivers is not what the wiki’s evidence supports. Under usefulness in infectious disease, the pro-immunity reading has to be stated in the open:

  • LCMV: T-bet⁺ B cells are required for control of chronic infection (IgG2a only partially accounts for it) (Cancro2020 - Age-Associated B Cells, review, mouse).
  • γHV68: CD11c⁺T-bet⁺ ABC expansion is partly responsible for reduced viral load, secreting virus-specific IgG2a ex vivo (same review).
  • Ehrlichia muris: protective T-bet⁺ memory B cells, multipotential on serial adoptive transfer; IgM⁺CD11c⁺ EF splenic plasmablasts drive CD4-T-independent antibody (same review, citing Racine 2008). See Tissue-Resident Memory B Cell.
  • Naive-derived DN2 in primary SARS-CoV-2 show “significant participation in generating neutralizing antibodies” — alongside, not instead of, autoreactivity (Sanz2025 - Human Atypical B Cells Overview, citing Woodruff 2020/2022).
  • Influenza: T-bet⁺FcRL5⁺CD27⁺ memory ABCs are poised for ASC differentiation and correlate with long-lived antibody responses (Sanz2025 - Human Atypical B Cells Overview, citing Nellore 2023).
  • The alternative lineage is a normal vaccine response: both PfSPZ (n=15) and influenza (n=9) vaccination prime it, and repeated boosting shifts cells toward CD21⁻CD27⁻ (Sutton2021 - Alternative Lineage B Cells in Vaccination and Infection).
  • ★ ZEB2 is dual-edged. B-cell-specific Zeb2 deletion improved disease in a lupus mouse model but decreased GC B cell numbers during persistent Plasmodium infection (Glaros2025 - Multilayered Identity of B Cell Memory, review, citing Gao 2024 Sci Immunol). The therapeutic read is on record: depleting these cells is not unidirectionally beneficial — it may cost humoral immunity in infection. Logged as an open contradiction on ZEB2.

4. And the measurement is treacherous

Each item below changes what a blood DN number can be claimed to mean.

  • ★ Blood may not measure the pool. Murine blood and splenic ABC pools are not in equilibrium, and Cancro explicitly cautions against blood-only tracking; in humans, blood ABC frequencies fall on antiretroviral therapy in HIV, suggesting circulating ABCs are an activated/mobilised state rather than a pool-size readout (Cancro2020 - Age-Associated B Cells, review; murine/splenic + human observational). This cuts both ways: “DN expansion” is not defensible phrasing for a blood-only study, but blood is arguably the right compartment for an acute mobilisation question.
  • ★ The standard gate misses most of the population. CD21⁻CD27⁻ captures only 44.7% of transcriptomically defined atBC1; CD11c is the best single surface marker (Sutton2021 - Alternative Lineage B Cells in Vaccination and Infection, n=4 CITE-seq). Every dengue DN number in the wiki is an undercount of unknown size.
  • ★ And the gate over-calls on a different axis. A CD21⁻-anchored ABC gate over-calls by roughly 50%, since only ~2/3 of CD21⁻CD23⁻ cells are T-bet⁺ and ~half of those CD11c⁺ (Cancro2020 - Age-Associated B Cells, murine splenic). Undercounting by transcriptome and over-calling by marker are two different errors on two different axes — they do not cancel.
  • Functional priors are shakier than they read. Sutton2021 - Alternative Lineage B Cells in Vaccination and Infection finds no PC-programme genes (XBP1, IRF4, PRDM1) in any atBC cluster at rest, while the “can’t make PCs” result is substantially a soluble-antigen artefact (Glaros2025 - Multilayered Identity of B Cell Memory). Defensible position: conditional capacity — capable, often primed, not committed. The Plasmablast contradiction stays open.

5. ★ The asymmetry that has to be named

The biomarker / disease-activity case is far stronger in autoimmunity than in infection.

AutoimmunityInfection
Activity correlationDN–SLEDAI; nephritis p=0.025; anti-dsDNA p=0.001; anti-RNP/Sm p=0.009; SLAM p=0.02 (Wei2007 - DN Memory B Cells in SLE, n=36)DN2-within-DN vs CRP r²=0.39, P=0.022 (Woodruff2020 - EF B Cell Responses in COVID-19, n=10 ICU)
Autoantibody correlationDN vs VH4.34 IgG R²=0.8, P<0.05 (Anolik2004 - Rituximab and B Cell Abnormalities in SLE, n=17)none
Treatment-response correlationrituximab resolves DN expansion (P=0.05); belimumab ABC loss tracks responsenone
Index metricDN2:DN1 informative even when total DN is normal (Jenks2018 - DN2 B Cells and EF Pathway in SLE)DN2:DN1 in critical COVID indistinguishable from active SLE (P≤0.0001 vs HD) — the wiki’s most robust cross-disease EF metric
Dengue—severity association council-downgraded for day-of-sampling confounding (severe sampled 8±4 vs 5±2 days)

Anyone arguing “DN cells matter in infectious disease” from this wiki is, to a substantial degree, arguing by transfer from autoimmunity. The transfer is licensed by the conserved cross-disease transcriptional programme — ABCs from malaria, HIV and autoimmune disease share transcriptional profiles (Glaros2025 - Multilayered Identity of B Cell Memory, review, citing Holla 2021) — but that licence is Tier B, transcriptomic, drawn from chronic settings, and dengue-untested. It is also internally contested: even within the shared T-bet⁺CD11c⁺ phenotype, ABCs are transcriptomically distinct from DN2 (Lamprinou2026 - ABCs and DN B Cells, citing Maul 2021), and functionally SLE DN2 (BCR-competent, FCRL4⁻) separate from HIV/malaria atypical memory (hyporesponsive, FCRL4⁺).

6. ★ The gap is documented, not merely inferred — dengue is absent from the field’s landmark disease survey

Everything in §5 is an argument about evidence the wiki holds. This section is about evidence the field does not hold, which is a different and unusually citable kind of claim.

Beckers2023 - Origins and Functions of DN B Cells (Immunology Letters 2023; 86 refs) is the standard comprehensive reference for DN B cells in health and disease. Its Table 1 surveys human peripheral blood DN B cells across ~25 conditions:

BlockConditions covered
Health / aginghealthy donors (young <60, aged >60)
Autoimmune (13)SLE, pSS, SSc, RA, JIA, axSpA, MS, NMOSD, myasthenia gravis, Guillain–Barré, Hashimoto’s, Graves’, IBD
Infections (6)meningitis/encephalitis, acute sepsis, malaria, rotavirus, HIV, COVID-19
Vaccinationinfluenza, tick-borne encephalitis virus
Other (5)Alzheimer’s, CVID, ALS, obesity, NSCLC

Dengue appears nowhere — not in the table, not in the text, not in the reference list.

Why this is worth a section rather than a footnote. The wiki’s premise that DN/EF B cells in dengue are understudied has until now been an inference from its own reading list, which is the weakest form of a negative claim. Beckers converts it into an observation about the field, made by an independent group with no dengue interest, in a review broad enough to include rotavirus and lung tumour tissue. Dengue’s absence is therefore not a scope decision.

And it is a missing category, not just a missing row. The conditions Beckers covers are predominantly chronic — HIV, malaria, autoimmunity, normal aging — with COVID-19 as the single acute-infection exemplar. Dengue would occupy an under-populated cell: an acute, self-limiting, serotype-structured febrile illness with a well-defined secondary-infection immunopathology. This is the same asymmetry §5 names from the other direction: the transfer argument runs from chronic and autoimmune settings, and dengue is neither.

⚠ Three limits on how hard this can be pushed.

  1. It is a 2023 snapshot. Both dengue DN measurements in this wiki — Ansari2025 - Peripheral T Helper Subset Drives B Cell Response in Dengue (n=170) and Singh2026 - DENV-Specific Memory B Cell Subsets (preprint) — postdate the review. The defensible claim is “as of early 2023, dengue had not entered the DN disease literature,” not “nobody has studied DN cells in dengue.” Writing the second where the first is true would be a real overreach — and it is the version that reads better, so it needs guarding against.
  2. It is a narrative review with no stated search strategy. An omission from a non-systematic review is weaker evidence of a field-level gap than an omission from a systematic one. It remains the best available evidence, and it is the reference the field actually uses.
  3. The absence may be partly methodological rather than purely a gap in attention. Beckers reports that the severe-COVID DN2/DN3 expansion is transient, with frequencies in recovered patients matching those who were never severe. If dengue’s DN response decays comparably, cohorts sampled at convalescence would record nothing — so some fraction of “nobody has found it” may be “nobody has sampled early enough.” That is a constraint on cohort design, and it is actionable rather than merely cautionary.

What Beckers additionally hands this analysis, beyond the absence. Three points that bear on §§1–5:

  • The exhausted-vs-activated split now has a phenotypic axis, and it splits by disease class. DN cells carry multiple inhibitory receptors (FcRH3-5, CD22, CD85j) in HIV, malaria, vaccinees and healthy donors, but lack them in SLE and Hashimoto’s thyroiditis. Beckers’ reading: exhaustion in chronic infection and aging, activation in autoimmunity. This is the axis §5 was reaching for when it separated “SLE DN2 (BCR-competent, FCRL4⁻)” from “HIV/malaria atypical memory (hyporesponsive, FCRL4⁺).” Dengue’s placement is undetermined — it is acute like COVID-19 (activated side) but infectious like HIV/malaria (exhausted side), and Singh2026 - DENV-Specific Memory B Cell Subsets’s temporal-correlation result arguing against exhaustion is currently the only dengue-specific evidence either way. ⚠ Note the internal inconsistency flagged on FcRH4: the same review elsewhere states that circulating DN cells are FcRH4⁻ in HD and SLE, which cannot be squared with “FcRH3-5 in healthy donors” as written.
  • An expanded DN compartment does not imply an expanded DN2 compartment. In MS, only a minority of DN cells could be retraced to CD21⁻CD11c⁺ (DN2-like) cells, and Beckers concludes the role of DN cells in MS “could differ from that in SLE.” Since §7 already flags that no dengue study has resolved DN1/DN2/DN3, MS is a worked example of the failure mode: total-DN and DN2 can move independently, and the SLE model does not transfer automatically.
  • The DN1-memory / DN2-effector dichotomy is softer than §§2–3 assume. DN1 cells also differentiate into ASCs in vitro under the same TLR7 + BCR + IFN-γ + IL-21 cocktail used for DN2, and scRNA velocity has been reported flowing both DN1→SM and SM→DN1. The dichotomy stays as the best organising frame — DN2 alone carries the ASC transcriptional programme (IRF4⁺, Ets-1⁻, BACH2⁻) and shares clones with plasma cells — but it should be stated as a difference in propensity, not in capability.

7. What this licenses saying about dengue

Supportable now. DN2-phenotype cells are expanded in acute dengue (Ansari2025 - Peripheral T Helper Subset Drives B Cell Response in Dengue, n=170). DENV-specific atypical MBCs are higher in secondary than primary infection at early convalescence (p<0.01) and remain higher at 18 months (p<0.05) while resting MBCs do not differ — they accumulate with repeat exposure and behave as a bona fide memory compartment (Singh2026 - DENV-Specific Memory B Cell Subsets, preprint). Dengue’s antibody output carries the low-fidelity signature the compartment predicts (GodoyLozano2016 - Lower IgG SHM Rates in Acute Dengue, Priyamvada2016 - Cross-Reactive Memory Plasmablasts in Secondary Dengue).

Not supportable yet. That dengue DN cells are true DN2 — no dengue study has stained T-bet, CXCR5, FCRL5 or ZEB2 within the DN gate, so DN1/DN2/DN3 has never been resolved in dengue. That DN frequency tracks severity independently of sampling day. That DN cells are the source of dengue autoantibodies or ADE-competent IgG. That “DN expansion” describes pool size rather than mobilisation.

Open Questions

  • ★ No functional antibody output has ever been measured from sorted DN/atypical cells in any infection. Sutton2021 - Alternative Lineage B Cells in Vaccination and Infection states this as its own limitation (no neutralization, no ADE). The autoantibody evidence is SLE-only and in vitro (Jenks2018 - DN2 B Cells and EF Pathway in SLE, LIPS). The wiki’s central claim — that this compartment makes the harmful antibodies — has no direct infection evidence anywhere in the corpus.
  • No ingested HIV primary, no malaria-infection primary beyond Sutton2021 (core n=4), no TB source, no influenza primary, no Ebola source. The canonical human atypical-memory primaries Moir 2008 and Weiss 2009 — with Portugal 2015 (audit #29) and Holla 2021 (audit #61, the cross-disease transcriptomics underpinning the whole transfer argument) — are catalogued as un-ingested in External Citation Audit. Until they are in raw/, the two most-quoted infectious-disease exemplars are Tier B. → routed to state.md Queue.
  • No source in the wiki tests DN frequency as a prospective biomarker in any infection. Every association is cross-sectional or post-hoc; none reports sensitivity, specificity, predictive value, or a pre-specified threshold.
  • Are the two dengue DN measurements measuring the same cell? Ansari2025 gates IgD⁻CD27⁻ then CD21⁻CD11c⁺; Singh2026 gates CD27⁻CD21⁻ on DENV-specific cells; neither matches Sutton2021’s CD11c-primary definition. Cross-gate reconciliation is unattempted.
  • Does the protective (GC-supporting) reading apply to acute infection at all? The Gao 2024 Zeb2 result (via Glaros2025 - Multilayered Identity of B Cell Memory) comes from persistent Plasmodium, and the LCMV/γHV68 data (via Cancro2020 - Age-Associated B Cells) are chronic murine models. Whether atypical cells support or subvert humoral immunity across a ~7-day febrile illness is untested in any wiki source.
  • Does the >1-year DN2/DN3 vaccine-memory persistence (>50% of spike/RBD⁺ cells) hold for infection-induced dengue memory? It is currently a Tier B claim from one review, and it would substantially change how the compartment is framed if true.
  • ⚠ Correct the unattributed Ebola mention on Double-Negative B Cell — either source it or remove it.

Already-tracked, not re-filed here: dengue DN2 identity unconfirmed (Watch Item 2026-05-08); atypical-lineage undercounting by CD21⁻CD27⁻ gating in dengue studies (2026-05-22); IgG3/complement angle (2026-05-15, 2026-05-22); Sutton PC-fate context dependence (2026-05-22); blood-vs-tissue compartment caveat not yet swept corpus-wide.

Entities: Double-Negative B Cell, DN2 B Cell, DN3 B Cell, Atypical B Cell, Age-Associated B Cell, Activated Naive B Cell, Plasmablast, Tissue-Resident Memory B Cell, CD11c, CD21, CD27, T-bet, ZEB2, FCRL5, FcRH4, SLAMF7, TLR7, IL-21, TNF-alpha, IRF4

Concepts: Extrafollicular Response, Germinal Center, Memory B Cell, Somatic Hypermutation, Class Switch Recombination, Antibody-Dependent Enhancement, Original Antigenic Sin, Dengue Severity Classification

Methods: Conventional Flow Cytometry, Spectral Flow Cytometry, CITE-seq, BCR Sequencing

Analyses: Mechanistic Case for DN and DN2 Cells in Dengue, Thesis Objectives and Grant Pitch, Research Plan - DN B Cell Expansion in Dengue, DN2 Gating Strategy, B Cell Panel Variant 1, External Citation Audit, Notable Findings

Graph View

Backlinks