Cancro2020 - Age-Associated B Cells

Full citation: Cancro, M. P. (2020). Age-Associated B Cells. Annual Review of Immunology, 38, 315–340. https://doi.org/10.1146/annurev-immunol-092419-031130

Raw file: [[raw/cancro2020.pdf]]

Note on evidence type: This is a narrative review with zero original data, written by one of the two investigators who first defined the ABC subset (Hao et al. 2011, from the Cancro group). Every claim below is a characterisation of someone else’s published work; none of it is independently verified here except where noted. The great majority of the evidence is murine — where a claim is human, this page says so explicitly. Treat it as the field’s canonical framing of the ABC subset circa 2020, not as a data source.

Summary

Cancro reviews the age-associated B cell (ABC) subset a decade after its first description, covering its phenotype, the signals that generate it, its origins in vivo, and its roles in aging, microbe-specific immunity, and autoimmunity. The subset was defined in companion 2011 papers from the Cancro and Marrack groups: Hao et al. defined ABCs as B220⁺CD19⁺ splenic cells lacking CD21, CD23, CD95 and CD43; Rubtsov et al. used CD11c expression on B220⁺CD19⁺ splenocytes. The marker sets differ but the populations largely overlap. ABCs accumulate continuously across the murine lifespan — nearly undetectable in juveniles, a distinguishable pool by 12–18 months, and as much as half of all splenic B cells by 24–30 months.

The central mechanistic argument is a two-signal model of ABC specification: endosomal nucleic-acid-sensing TLR7 or TLR9 signals are necessary to poise a B cell for the ABC fate but are not sufficient; they must be followed by IFN-γ or IL-21. BCR ligation — alone or with CD40 costimulation — does not enable ABC fate, though it synergises with TLR signals for proliferation. IL-4 actively blocks the T-bet⁺ ABC fate in the context of IL-21 but does not block IFN-γ-driven ABC formation, which fits the Th1 association of ABCs and was confirmed in vivo with Th1-skewed (influenza) versus Th2-skewed (Heligmosomoides polygyrus) infections in cytokine-knockout mice.

On origins, Cancro is markedly more cautious than the literature that cites him. The inference that ABCs are GC-experienced rests on somatic mutation in sorted ABC Ig genes plus a requirement for MHC-II, CD40 and CD154 — but he states plainly that cognate T cell interaction and GC entry “are based on inference, but have not yet been directly demonstrated,” that anatomic localisation of ABCs in GCs has never been shown, and that SHM can occur independent of GC formation. He concludes that ABCs may arise and hypermutate in extrafollicular or other GC-independent niches, and that the ABC pool most likely reflects a combination of formative routes (antigen-driven memory, homeostatic expansion) rather than one. The review closes on aging (ABCs displace follicular B cells, produce TNF-α that suppresses B lymphopoiesis, and contribute to inflammaging), autoimmunity (ABCs are enriched for anti-chromatin/anti-Sm/anti-DNA specificities and track with SLEDAI in human SLE), and a tolerance model in which TLR9 normally triggers programmed death of B cells that internalise nucleic-acid-containing antigen — ABCs being what emerges when survival cytokines rescue those cells instead.

Study Design

• Type: Narrative review (Annual Reviews), no original data, no systematic search protocol, no PRISMA
• Sample size: N/A — 185 cited references
• Setting: Predominantly murine (splenic ABC biology, knockout and adoptive-transfer studies); human data drawn in for HIV, malaria, SLE, RA, Sjögren, scleroderma, CVID
• Population: N/A. Species coverage is the key caveat — the mechanistic core (§3, §4) is almost entirely mouse; the human sections (§5.2, §7.2) are descriptive
• Author standpoint: Cancro is co-author of the founding Hao et al. 2011 paper and of much of the signalling work reviewed (Naradikian 2016, Sindhava 2017, Russell Knode 2017). This is an insider synthesis, which makes it authoritative on the mouse work and worth reading sceptically where it adjudicates between competing phenotyping schemes.

Key Findings

Phenotype and heterogeneity

• Two founding definitions, different markers: Hao et al. — B220⁺CD19⁺ splenic B cells lacking CD21, CD23, CD95, CD43. Rubtsov et al. — CD11c⁺ on B220⁺CD19⁺ splenocytes. “Largely overlapping populations,” not identical (mouse).
• ★ At least three populations sit inside the ABC gate. Within the murine CD21⁻CD23⁻ splenic B cell pool, only ~2/3 are T-bet⁺, and among those, roughly half are CD11c⁺. Whether these are stable unrelated pools or differentiation/activation stages within one lineage is unresolved. A CD21⁻-only gate therefore over-calls ABC by ~50%.
• Age accumulation: nearly undetectable in juvenile/young adult spleens; low frequency at 3–6 months; a readily distinguishable, steadily enlarging pool by 12–18 months; up to half of all splenic B cells at 24–30 months. Considerable individual variation even in age-matched, co-housed cohorts (mouse).
• ★ Blood and splenic ABC pools are not in equilibrium. ABCs are consistently spleen-enriched, scarce in most lymph nodes and lymphatics (distinguishing them from FO B cells), present in bone marrow and blood but with substantial variability between mice and within a mouse over time. Cancro states explicitly that the proportional representation of ABCs in blood “does not necessarily parallel that observed in the spleen” and sounds “a cautionary interpretive note for studies that track ABCs only in peripheral blood” (mouse).

Activation requirements

• ABCs do not divide in response to BCR cross-linking — distinguishing them from FO B cells — yet remain viable under those conditions, distinguishing them from MZ and transitional B cells, which die rapidly after BCR cross-linking (mouse).
• They proliferate robustly to TLR7 or TLR9 ligands. BCR cross-linking has no effect alone but synergises with either TLR signal, yielding more rounds of division. The BCR remains an active signalling system in ABCs, uncoupled from direct mitogenic activity.
• Lack CD43 and CD5, and are absent in neonates/young adults — both opposite to the B1 compartment.

★ §3 — Signalling requisites (two-signal model)

• TLR7 or TLR9 signals are necessary to poise naive B cells for ABC fate (Naradikian et al. 2016). Regardless of subsequent cytokine exposure, BCR ligation — alone or with CD40 costimulation — does not enable ABC fate. Poising must be followed by IFN-γ or IL-21. Coculture experiments established that both the TLR and the cytokine requisites are cell-intrinsic (mouse, in vitro + in vivo).
• IL-4 is an active antagonist, conditionally. IL-4 negatively regulates T-bet⁺ ABC fate in the context of IL-21, but does not block ABC fate driven by IFN-γ. In vivo: ABCs formed in WT mice during influenza (Th1) but failed in IFN-γ-deficient mice — yet arose in IFN-γ/IL-4 double-deficient mice. Conversely ABCs do not arise during H. polygyrus (Th2) in WT mice but emerge robustly in IL-4-deficient mice. So IL-4 blocks ABC differentiation in the absence of IFN-γ; remove IL-4 and ABCs can be generated IFN-γ-independently, presumably via IL-21 (mouse, knockout).
• Kinetics: T-bet expression begins within 12 h of TLR ligand + cytokine exposure — well before the first division. Cancro suggests this is too fast for a purely epigenetic mechanism and more consistent with immediate shifts in intracellular signalling or metabolic status.
• ★ CD11c induction may be cytokine-direct, not T-bet-downstream. Transcriptional analyses of IFN-γ- or IL-21-treated WT versus T-bet-deficient cells showed that while some aspects of the ABC phenotype rely strongly on T-bet per se, others — such as CD11c expression — were largely direct effects of each cytokine rather than downstream targets of T-bet. Cancro notes this “remains somewhat controversial” in vivo, since T-bet has been implicated as necessary for CD11c induction in some systems but not others, and proposes that the discrepancies reflect differing routes of ABC formation (mouse).
• ★ A human exception to the two-signal rule. Naradikian et al. also examined human peripheral blood B cells. The same general relationships held, except that some activated human CD27⁻ B cells expressed T-bet directly induced by IFN-γ without concomitant TLR ligands. Cancro’s preferred explanation: some B cells within the human CD27⁻ blood pool have already received ABC-poising signals in vivo (human, in vitro).

★ §4 — Origins in vivo (the section most load-bearing for this wiki)

• Not de novo from aged bone marrow. After sublethal irradiation (5 Gy) ablates peripheral B cell pools, FO and MZ compartments fully reconstitute but the ABC pool does not return rapidly. Naturally arising ABCs are therefore a slowly accumulating population derived from preimmune peripheral B cell compartments, not the output of an ABC-skewed progenitor (mouse).
• FO B cells are competent progenitors. Adoptive transfer: ABCs arise from splenic FO B cells within 30 days; donor-derived ABCs were found only among the most extensively divided cells. Neither donor nor recipient age influenced the result — young and aged donor FO B cells were equally efficient in young or aged recipients (mouse).
• Repertoire argues against clonal expansion. Russell Knode et al. sequenced heavy and light chains from sorted ABCs: a diverse array of germline V_H and V_κ genes, largely congruent with the FO pool — ruling out the age-associated clonal expansions described for T cells. Many ABC V regions were somatically mutated (mouse).
• T help requirement: neither MHC-II-deficient nor CD40-deficient FO B cells yielded ABCs, and CD154-deficient mice fail to develop natural ABCs with age (mouse).
• ★ But GC origin is inference, not demonstration. Cancro states directly that “both the requisite for cognate T cell interactions and the assertion that ABCs have entered GCs are based on inference, but have not yet been directly demonstrated.” Three specific caveats: (1) the ABC phenotype can be achieved without CD40 ligation — bystander cytokine, particularly IFN-γ, may suffice (Naradikian 2016; Zumaquero et al. 2019, eLife, human T-bet^hi B cells); (2) although T-bet⁺ B cells express GC-associated markers (PNA, CD95) during the first weeks of a response, anatomic localisation of ABCs in GCs has not been directly demonstrated; (3) SHM can occur independent of GC formation (citing Di Niro et al. 2015, Immunity — Salmonella drives promiscuous B cell activation followed by extrafollicular affinity maturation). Conclusion in his words: “it remains possible that ABCs arise and undergo somatic hypermutation in extrafollicular or other GC-independent niches.”
• Alternative route — homeostatic expansion. The Swain group reported ABC-characteristic cells emerging under limited exogenous antigen availability, and that with advancing age an increasing proportion of primary influenza responses involves ABCs. Cancro proposes a T-cell-analogous route in which endogenous ligands drive gradual proliferation yielding cells bearing hallmarks of prior activation. He notes these two paths are “neither mutually exclusive nor mechanistically disparate.”
• Synthesis: the ABC pool reflects a combination of formative routes whose relative contributions vary with age, antigenic load and other variables. Consistent with this, Hao-defined ABCs include IgM⁻ and IgM⁺ cells, and within the IgM⁺ pool a range of sIgD — plausibly switched (IgM⁻IgD⁻) and unswitched (IgM⁺IgD⁻) memory from antigen-driven responses versus homeostatically expanded primary cells (IgM⁺IgD⁺). ABCs include both somatically mutated and germline-configuration Ig genes.

★ §5 — Microbe-specific immunity

• Mouse infection models: Ehrlichia muris — Winslow group described IgM⁺CD11c⁺ extrafollicular splenic plasmablasts responsible for CD4-T-independent antibody responses (Racine et al. 2008) and protective T-bet⁺ memory B cells; T-bet⁺ Bmem are multipotential, giving rise to multiple effector B cell lineages on serial adoptive transfer. Gammaherpesvirus 68 — CD11c⁺T-bet⁺ ABC expansion, secreting virus-specific IgG2a ex vivo and partly responsible for reduced viral load. LCMV — T-bet⁺ B cells required for control of chronic infection, with IgG2a only partially accounting for the effect. Also influenza.
• Human infections: HIV “tissue-like memory” B cells (Moir et al.) lack CD21 and CD27, express inhibitory receptors, are hyporesponsive to BCR cross-linking — initially read as exhausted — yet proliferate robustly to TLR9, like murine ABCs; T-bet expression demonstrated subsequently. Malaria “atypical memory” B cells (Weiss et al. 2009) are IFN-γ-driven and display diverse V_H gene usage, consistent with the Th1 requirement and broad repertoire of murine ABCs.
• Acute versus chronic: transient ABC increases follow live virus vaccination or infection — influenza, yellow fever, vaccinia. Sustained ABC pools occur in chronic infection — HIV, hepatitis C, tuberculosis.
• ★ Blood ABC frequency may be a mobilisation readout, not a pool-size readout. Peripheral blood ABC frequencies in HIV-infected individuals are reduced during antiretroviral therapy. Cancro raises the possibility that T-bet⁺ ABCs found in blood “may represent an activated or mobilized differentiation state that wanes upon viral clearance, despite their retention in the spleen or other tissues as T-bet⁺ ABC memory cells.” He notes this would explain the blood/spleen disequilibrium and is consistent with autoimmune data where blood ABC frequency tracks clinical disease activity (human, observational).
• ★ The intracellular-infection commonality. “In nearly all cases, these involve intracellular infections.” Cancro speculates this is necessary to meet the criteria for ABC generation — abundant IFN-γ plus internalisation of TLR7 or TLR9 ligands, the latter available from apoptotic debris associated with cytolytic cell death or from pathogen components internalised via BCR ligation.

Function and antibody output

• Isotype skewing toward IgG2a/c in mice and IgG1 in humans, favouring inflammatory effector functions and ADCC as a major effector mechanism. Adoptive transfer of virus-specific IgG2a only partially restored viral control, so additional functional distinctions exist.
• Cytokines: on TLR7/9 activation ABCs produce higher IFN-γ and exceptionally high IL-10 than FO B cells (transcriptional array + ELISPOT). They are effective antigen-presenting cells with high MHC-II and skew naive CD4 T cells toward Th17 in vitro.
• ★ T-bet represses Blimp-1 — so ABC→plasma cell differentiation likely requires losing T-bet. Several studies show ABCs differentiate rapidly to antibody-secreting plasmablasts on TLR7/TLR9 stimulation plus IL-21 (citing Wang et al. 2018 and Jenks2018 - DN2 B Cells and EF Pathway in SLE). However, “few if any plasma cells express T-bet, and there is evidence that T-bet represses Blimp-1, suggesting that the formation of plasma cells from ABCs likely involves the loss of T-bet expression.” ⚠ Cancro asserts this without a numbered reference at that sentence — see Relevance & Notes.
• The ABC-as-depot hypothesis: the splenic ABC Bmem pool may be a continuously replenishing source of precursors feeding plasma cell pools, potentially resolving how the bone marrow niche sustains lifelong PC production without continuous enlargement or loss of prior specificities. Consistent with Allman group data showing a significant fraction of BM plasma cells turn over rapidly. Cancro flags that testing this requires fate-mapping tools and turnover measurement in the ABC pool.
• Unresolved progenitor–successor relationships: are unswitched, unmutated IgM⁺ ABCs progenitors of class-switched mutated ABCs? Are CD11c⁺ ABCs a differentiation state derived from CD11c⁻ ABCs? Is there appreciable interchange between T-bet⁺ ABC and T-bet⁻ Bmem pools? All open.
• Tissue residency: growing evidence that T-bet⁺ ABCs differ in recirculation and tissue-residency properties, implied by homing/trafficking molecule expression and direct localisation studies — an “intriguing parallel with T-bet-expressing T cells,” suggesting the T-bet programme is partly generic to effector memory cells with proscribed tissue residency.

§6 — Aging

• ★ ABC-derived TNF-α suppresses B lymphopoiesis. The Riley laboratory showed ABCs impede early B cell developmental steps through TNF-α production — both directly, by inducing apoptosis in pre-B cells, and indirectly, through systemic inflammatory effects on the bone marrow microenvironment (mouse).
• ABCs displace follicular B cells. The sum of ABCs and FO B cells tends to remain constant — as ABC numbers rise, FO numbers drop correspondingly. Mechanism: ABCs express BAFFR and TACI and can consume/sequester BAFF, yet are themselves largely BAFF-independent (BAFF-blocking antibody eliminated FO and MZ pools but left ABC numbers unchanged), making them exceptional competitors in BAFF-regulated homeostatic space. Their BAFF independence resembles Bmem.
• Inflammaging: ABC propensity for IL-6 and IFN-γ production plus Th17-skewing APC function implicates them in heightened basal inflammatory states with age (not yet tested by gain/loss-of-function).
• Visceral adipose tissue: Frasca and Blomberg groups report T-bet⁺CD21⁻ “inflammatory B cells” / “late memory B cells” congruent with ABCs in VAT; VAT-resident ABCs increase with age, VAT correlates with ABC numbers, and VAT adipocyte-conditioned medium enriches for ABCs.
• ★ ABCs and original antigenic sin (dedicated sidebar). As the ABC:FO ratio rises with age, antigenic challenges become progressively less likely to draw responses from the naive FO pool and increasingly recruit ABCs. Consequence: “the repertoire participating in primary or recall responses will be increasingly colored by previous antigenic exposures, suggesting the ABC pool may be involved in antigenic imprinting characteristic of some sequential viral infections.” The sidebar notes the influenza case is particularly promising because the dominant anti-hemagglutinin isotypes — IgG2a/c in mice, IgG1 in humans — are the T-bet-driven ones.

§7 — Autoimmunity

• ABCs are elevated in nearly all murine humoral autoimmunity models interrogated. In NZB/WF1 and Mer⁻/⁻ SLE models ABCs expanded as early as 3 months and comprised 15% of splenic B cells by 6 months — levels rarely seen in healthy C57BL/6 before 12–18 months. Implicated as sources of anti-chromatin antibodies; TLR7-dependent.
• Hybridomas derived from autoimmune-model ABCs are enriched for anti-Sm and anti-DNA specificities. CD11c⁺T-bet⁺ ABCs are required for anti-chromatin antibodies in the bm12 chronic-GVH SLE model. In B6.Sle1, ABC-phenotype cells emerge in IFN-γ-dependent spontaneous GCs.
• IL-21 → IRF5 axis: in the SWEF double-knockout model, ABC expansion was strongly IL-21-driven — IL-21-deficient SWEF DKOs developed neither expanded ABCs nor autoantibodies — and the IL-21-driven expansion relied on IRF5, an established human SLE risk factor (plain-text citation; no IRF5 wiki page).
• Human: Cancro treats “ABC-like cells termed double negative (DN)” from the Ettinger and Sanz groups as the human counterpart, noting both TLR7 and IL-21 are key, that transcriptional profiles differ from other mature peripheral cells, and that many unique transcriptional characteristics are shared between murine and human ABCs. Wang et al. showed elevated blood ABCs correlate strongly with SLEDAI and that the ABC pool is highly enriched for autoantibody specificities. In a belimumab trial, loss of ABC-phenotype cells correlated with therapeutic response. Also RA (sometimes in affected joint tissue), Crohn disease, Sjögren syndrome, scleroderma. In CVID, the CVIDc subset (autoimmune cytopenia + interstitial lung disease) but not CVIDio (infectious complications only) shows clear elevation of an atypical T-bet⁺ population with ABC characteristics, in a Th1/IFN-γ-skewed milieu (human, small cohorts).

★ §8 — TLR9 as a tolerance checkpoint

• TLR7 and TLR9 act in opposite directions on autoimmunity. TLR7 duplications foster disease and TLR7 deficiency ameliorates it, but TLR9 knockouts exacerbate autoimmune manifestations, including earlier disease onset and more severe glomerulonephritis. TLR9 therefore has a role in maintaining peripheral B cell tolerance (mouse).
• The mechanism (Sindhava et al. 2017): when B cells are stimulated with antigens that deliver a TLR9 ligand via the BCR, they undergo cell cycle arrest and mitochondrial cell death after an initial proliferative burst. This TLR9-dependent programmed death is circumvented by survival cytokines or CD40 costimulation — and in the presence of IFN-γ or IL-21, the rescued cells assume the ABC phenotype.
• Interpretation: the frequent association of ABCs with humoral autoimmunity may reflect failure of, or rescue from, this peripheral tolerance mechanism. More broadly, it implies that molecular pattern recognition systems parsing internalised antigen components — rather than BCR epitope specificity per se — underlie peripheral B cell tolerance.

Methods Used

No original experiments. Methods discussed in the reviewed literature and relevant to this wiki: Conventional Flow Cytometry (the two divergent founding phenotyping schemes), FACS Sorting (ABC isolation by Hao criteria for Ig sequencing), BCR Sequencing (V_H/V_κ repertoire and SHM in sorted ABCs), In Vitro B Cell Stimulation (the TLR-poising/cytokine two-signal system; T-bet readout at 12 h), ELISpot (ABC cytokine output ex vivo), RNA Sequencing (transcriptional arrays of T-bet-deficient versus WT cytokine-treated cells).

Entities Mentioned

Age-Associated B Cell, Atypical B Cell, Double-Negative B Cell, DN2 B Cell, Plasmablast, T-bet, CD11c, CD21, CD23, CD27, CD19, B220, CD40L, TLR7, IL-21, TNF-alpha, BLIMP-1, IgG, IgM, IgD

Concepts Addressed

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

Relevance & Notes

Why this paper matters here. It is the canonical review of the wiki’s spine entity, by the investigator who defined it. Three things it supplies that the wiki did not have:

1. A quantitative heterogeneity prior. Only ~2/3 of murine CD21⁻CD23⁻ B cells are T-bet⁺, and only ~half of those are CD11c⁺. This puts numbers on the T-bet-as-confirmation question tracked on B Cell Panel Variant 1 and warns that a CD21⁻-anchored gate over-calls the population.
2. An explicitly agnostic position on ABC origin, from 2020. Cancro’s §4 caveats — GC entry never directly demonstrated, PNA/CD95 expression ≠ GC localisation, SHM occurs GC-independently — line up with the fate-mapping result the wiki now carries via Glaros2025 - Multilayered Identity of B Cell Memory (Song 2022) and cut against the “ABCs are at least partly GC-experienced” position from Lamprinou2026 - ABCs and DN B Cells. Notably, Cancro reaches this scepticism from the same somatic-mutation evidence Lamprinou uses to argue the opposite, and he names the reason: mutation is not a GC-origin proxy.
3. A mechanism for the tracked Sutton/Jenks contradiction (wiki-generated synthesis). Sutton2021 - Alternative Lineage B Cells in Vaccination and Infection finds no PC-programme genes (XBP1, IRF4, PRDM1) in atBC clusters at rest; Jenks2018 - DN2 B Cells and EF Pathway in SLE finds DN2 cells are efficient plasmablast precursors on stimulation. If T-bet represses Blimp-1 and PC formation from ABCs requires T-bet loss, both observations follow: a T-bet^hi atBC at rest should be PRDM1-low because T-bet is holding it there, and the transition to ASC should require dismantling that programme. ⚠ This reconciliation is the wiki’s synthesis, not Cancro’s — he never mentions Sutton2021 (published a year later) and does not frame the repression claim as resolving anything.

What this paper does not settle. It is a review, and where it is most useful to this wiki it is also least verified:

• ⚠ The T-bet-represses-Blimp-1 claim carries no numbered reference at the point of assertion — unusual for an Annual Reviews article and notable given how much load it bears here. It is flagged as an unreferenced assertion on BLIMP-1 and T-bet, and independent verification is a Watch Item. Until verified, the Sutton/Jenks reconciliation above is a hypothesis with a plausible mechanism, not a resolution.
• Species. §3 and §4 — the mechanistic core — are essentially all mouse. The one human signalling observation (IFN-γ-induced T-bet without TLR ligands in CD27⁻ blood B cells) is an exception to the model, and Cancro’s explanation for it is speculative.
• Zero dengue content. ABCs in acute self-limiting viral infection are represented by influenza, yellow fever and vaccinia — transient increases only. Dengue fits Cancro’s “intracellular infection” criterion (cytoplasmic ssRNA → TLR7 ligand; strong IFN-γ), which makes the comparative extension reasonable, but it remains an extension.

Direct consequence for the pilot. The blood/tissue disequilibrium warning (§2.1) and the HIV/ART observation (§5.2) together mean a peripheral blood DN/ABC frequency should not be read as a proxy for total ABC burden. This cuts both ways and the honest framing is the second one: if circulating ABCs are an activated or mobilised state, then blood frequency during acute d5–8 dengue is arguably the more appropriate readout for an acute-response question than a pool-size measure would be — but the claim must be written as “circulating DN frequency,” never as “DN expansion,” and cross-sectional acute-versus-convalescent comparisons are measuring mobilisation, not accumulation. Both the caveat and the reframe are carried to Thesis Objectives and Grant Pitch.

Relation to existing wiki positions.

• Confirms: the TLR7 + IFN-γ/IL-21 differentiation programme underpinning DN2 B Cell and Extrafollicular Response; ABC association with autoantibody specificities (the cellular premise of the bridge-wiki/ thesis); CD11c as a primary marker.
• Qualifies: T-bet’s definitional status (CD11c may be cytokine-direct, not T-bet-downstream — converging with the Glaros2025 - Multilayered Identity of B Cell Memory demotion from a different direction, five years earlier); the ABC ≈ DN2 equation (Cancro maps ABC to human DN broadly, without the asymmetric-overlap precision of Lamprinou2026 - ABCs and DN B Cells).
• Apparent tension: Cancro states human ABC isotype skewing is toward IgG1; Sutton2021 - Alternative Lineage B Cells in Vaccination and Infection reports IgG3 enrichment in the alternative lineage. Different measurement bases (general isotype skew from an effector-function argument versus cluster-level transcriptomic enrichment) — logged on IgG Contradictions & Debates rather than smoothed over.

External references worth ingesting later: Di Niro et al. 2015 Immunity 43:120–31 (EF affinity maturation after Salmonella — a second EF-SHM primary alongside William2002 - Extrafollicular Somatic Hypermutation in Autoimmune Mice); Naradikian et al. 2016 J Immunol 197:1023–28 (the two-signal source); Sindhava et al. 2017 J Clin Invest 127:1651–63 (TLR9 tolerance checkpoint); Zumaquero et al. 2019 eLife 8:e41641 (human T-bet^hi B cells, CD40-independent — Jenks and Scharer co-authors); Russell Knode et al. 2017 J Immunol 198:1921–27 (ABC repertoire). Catalogued in External Citation Audit.

Questions Raised

1. Is the T-bet-represses-Blimp-1 claim correct, and what is the primary source? It is the mechanistic keystone for reconciling Sutton2021 with Jenks2018, and it is asserted here without a citation. Verification required before the wiki treats the contradiction as resolved.
2. Does the murine blood/spleen ABC disequilibrium hold in humans, and specifically in acute dengue? If circulating DN frequency is a mobilisation readout rather than a pool measure, what is the correct denominator and the correct verbal claim for a blood-only cross-sectional study? Untestable in the current pilot (no tissue), but it constrains how results may be phrased.
3. Are ABCs the cellular substrate of original antigenic sin in dengue? Cancro proposes ABCs as a mechanism for antigenic imprinting in sequential viral infection. Dengue is the paradigm case, and the wiki holds both halves separately — OAS at the antibody level (Priyamvada2016 - Cross-Reactive Memory Plasmablasts in Secondary Dengue) and DN/CD21⁻CD11c⁺ expansion at the cell level (Ansari2025 - Peripheral T Helper Subset Drives B Cell Response in Dengue) — with nothing joining them. Testable in principle: are DENV-specific cells within the DN2-phenotype gate biased toward the prior serotype?
4. Which of Cancro’s formative routes generates the dengue DN2-phenotype population? Antigen-driven memory versus homeostatic expansion predict different repertoire signatures (mutated/switched versus germline/unswitched). The wiki already holds an unresolved version of this in the GodoyLozano2016 - Lower IgG SHM Rates in Acute Dengue (low SHM) versus Priyamvada2016 - Cross-Reactive Memory Plasmablasts in Secondary Dengue (high SHM) split.
5. Does the CD11c⁻ → CD11c⁺ progenitor–successor relationship exist? If CD11c⁺ ABCs are a differentiation state rather than a stable subset, then CD11c-gated frequencies measure activation state, not compartment size — which changes what a between-arm comparison of DN2-phenotype percentage means.
6. Is the ~2/3 T-bet⁺ / ~half CD11c⁺ heterogeneity ratio preserved in human blood DN cells? The figures are murine and splenic. If they transfer, they set an expected ceiling for T-bet⁺ fraction within a CD11c⁺-gated dengue population; if they do not, the panel’s T-bet channel is measuring something else.
7. Does the TLR9-rescue tolerance model operate in dengue? Secondary dengue supplies immune complexes, apoptotic debris and viral ssRNA — the exact ingredients for BCR-delivered TLR ligand — plus abundant IFN-γ and IL-21. Cancro’s model predicts this should produce ABCs from cells that would otherwise be deleted, which is a candidate mechanism for the bridge-wiki/ cells→autoantibody arm that does not require a soluble self-antigen route.