Council Final Report: Sanz2025

Paper: Sanz I (2025). “Human Atypical B Cells. An Overview.” Immunological Reviews 334:e70058. Review date: 2026-05-15 Council composition: Default 4-member panel (Methodology Critic, Claims Validator, Contextual Critic, Strengths Advocate)

---

The Verdict

Sanz2025 is a valuable reference-class review from the defining authority in DN2/EF B cell biology, providing the field’s most comprehensive nomenclature reconciliation and a practical gating framework. However, it functions as a Sanz lab position paper as much as a neutral synthesis — its central thesis (abandon “AtB” for DN1/DN2/DN3) rests on conceptual argument and lab-of-origin authority rather than experimental evidence showing the DN classification is predictively superior. The review’s most impactful new concept for this wiki is the memory DN2 cell, which partially revises the lab’s own prior framing of DN2 as exclusively acute EF effectors. Dengue literature is entirely absent from the otherwise comprehensive cross-disease synthesis — the wiki’s connection between dengue data and the Sanz framework remains an original analytical contribution, not a field-established linkage.

---

Claim-by-Claim Assessment

#	Claim	Evidence strength	Overclaim risk	Consensus notes
1	AtB cells are a normal component of immune responses	MODERATE	Moderate	Presence in health is established; “normal” conflates prevalence with non-pathological function. Malaria chronic accumulation data underweighted.
2	AtB label should be abandoned for DN1/DN2/DN3	WEAK	High	Conceptual argument, not experimental comparison. DN2 itself is heterogeneous (effector vs. memory). Relocates rather than resolves classification ambiguity.
3	In primary responses, ABC-phenotype cells are predominantly naïve-derived	WEAK	High	Based on SLE flare + severe COVID-19 — extreme inflammatory contexts. Not validated in mild infections or controlled vaccine settings.
4	DN2 cells are not exhausted or anergic	MODERATE	Moderate	Supported in specific contexts (TLR7 stimulation). Does not refute exhaustion in HIV/chronic malaria. Context-dependent, not universal.
5	Self-limited EF autoreactivity is normal	WEAK	High	Single disease context (severe COVID-19). Extraordinary inflammatory conditions. Cannot generalise to routine infections.
6	ZEB2 is the primary driver of ABC formation	MODERATE	Moderate	Necessity well-established (Dai 2024). “Primary” not operationally defined; T-bet, IRF5, SWEF also important.
7	EF endotypes predict clinical outcomes	MODERATE	Moderate	Strong in SLE (severity + vaccine response). Preliminary in cancer. Most evidence cross-sectional, not prospective.
8	Selective therapeutic targeting of ABC/DN2 is feasible	WEAK	High	Review’s own heterogeneity argument undermines marker-based targeting precision. TLR7 inhibition contraindicated in acute viral infections (including dengue).

---

Top Strengths

1. Nomenclature reconciliation (STRONG). Table 1, Figure 1 (Venn diagram), and Figure 2 (gating strategy) are the most comprehensive reference artifacts available for resolving the AtB/ABC/DN2/CD21lo terminological confusion. These will be widely cited as a practical standardisation resource regardless of whether the field adopts “DN” as the sole label.

2. Memory DN2 concept — honest self-revision (MODERATE). The acknowledgment that antigen-specific DN2 cells persist >1 year post-vaccination (>50% of spike/RBD⁺ cells) fundamentally revises the lab’s own 2018 characterisation of DN2 as transient EF effectors. This is intellectually honest and opens a new research direction: distinguishing effector DN2 from memory DN2.

3. ZEB2→Mef2b GC-exclusion mechanism (MODERATE). The synthesis placing ZEB2 as a primary ABC driver that mechanistically represses GC entry (via Mef2b) provides the clearest molecular basis for EF/GC pathway antagonism published to date. This is assembled from prior work but framed more cleanly than any single paper.

---

Top Concerns

1. Self-referential bias — MAJOR CONCERN. The review’s historical narrative is structured to present the DN classification as the inevitable resolution of prior nomenclature confusion. Counter-proposals from other groups (Meffre lab, scRNA-seq atlas-based classifications) receive insufficient comparative analysis. The review should be read as a position paper from the DN2 defining lab, with appropriate epistemic discounting.

2. “Normal” framing overclaims — MAJOR CONCERN. The claim that AtB cells are “normal” conflates their presence in healthy individuals with non-pathological function. The review itself documents extensive pathological roles (SLE nephritis, COVID-19 severity, cancer resistance to immunotherapy). A cell that is present in health and amplified in disease is a context-dependent responder, not simply “normal.” Chronic malaria settings — where AtB accumulation correlates with impaired humoral protection — are underweighted.

3. DN classification relocates ambiguity — MAJOR CONCERN. The review’s own acknowledgment of effector vs. memory DN2 cells, combined with the lack of standardised CD11c gating thresholds and the requirement for intracellular T-bet staining, means the DN system introduces its own resolution problems. No experimental evidence is presented showing that DN1/DN2/DN3 is predictively superior to existing classifications in any clinical or mechanistic context.

---

What the Paper Proves vs. What It Implies

Proves:

• The AtB/ABC/CD21lo nomenclature captures heterogeneous populations that differ in origin, function, and fate
• DN2 cells exist in healthy individuals, not only in disease
• ZEB2 is necessary for ABC transcriptional signature (in TLR7-driven murine lupus)
• Antigen-specific DN2 cells can persist >1 year post-vaccination
• ABC/DN2 cells function as potent APCs across mouse and human studies

Implies (but does not prove):

• DN1/DN2/DN3 is a superior classification to AtB/ABC
• DN2 cells are the same functional entity across diseases (SLE, COVID-19, malaria, infections)
• Naïve-derived DN2 is the dominant ABC population in all primary responses
• Self-limited EF autoreactivity is a generalizable feature of healthy immune responses
• EF/GC endotypes are stable, predictive patient stratifiers
• Selective ABC/DN2 therapeutic targeting is clinically feasible

---

Remaining Gaps

• Effector vs. memory DN2 distinction: The extended phenotype, transcriptional features, and functional properties that distinguish effector DN2 from memory DN2 “remain to be fully understood” (paper’s own words). This is the most critical unresolved question for the field.
• Cross-lab gating standardisation: CD11c threshold for DN2 identification is not specified with antibody clone, fluorochrome, or instrument parameters. Until this is standardised, cross-study DN2 frequency comparisons remain unreliable.
• Dengue entirely absent: No dengue immunology paper is cited despite dengue featuring massive plasmablast waves (Wrammert2012), low SHM consistent with EF origin (GodoyLozano2016), and VH4-34 enrichment paralleling the self-limited autoreactivity model (Appanna2016). The wiki’s dengue-Sanz framework synthesis is original.
• Prospective endotype validation: EF/GC endotype clinical utility requires prospective studies showing endotype assignment at diagnosis predicts outcomes. Current evidence is retrospective/cross-sectional.
• DN2 identity in dengue unconfirmed: Ansari2025’s CD21⁻CD11c⁺ cells in acute dengue lack T-bet/CXCR5/FCRL5 staining. Whether these are canonical DN2, memory DN2, or a dengue-specific variant is unresolved.

---

Council Recommendation

How to treat this paper’s claims in the wiki:

1. Continue using DN nomenclature as the wiki’s primary classification system — it is the most precise currently available, and Sanz2025 confirms this choice. But add a caveat that the DN system is the Sanz lab’s framework and has not been independently validated as predictively superior to alternatives.

2. Integrate the memory DN2 concept. The DN2 B Cell entity page and Extrafollicular Response concept page should acknowledge that DN2 cells can be either effector (acute EF) or memory (durable, antigen-specific). Singh2026’s 18-month DENV-specific CD27⁻CD21⁻ data should be flagged as potentially representing memory DN2 rather than effector DN2.

3. Soften “concurrent EF+GC” framing. The wiki’s Extrafollicular Response page should reframe the CXCL13-based “concurrent EF+GC” claim to “possible mixed endotype” rather than demonstrated co-existence — consistent with Sanz2025’s competing endotype model and the existing CXCL13 watch item.

4. Do not modify claims about dengue DN2 identity based on this review alone. Sanz2025 provides no dengue-specific data. The wiki’s existing watch items (T-bet/CXCR5 staining needed, DN1 vs DN2 resolution needed) remain the correct framework.

5. Add a new watch item for the memory DN2 question in dengue: “Do DENV-specific CD27⁻CD21⁻ cells at late convalescence (Singh2026, 18 months) represent memory DN2 (per Sanz2025/Faliti2024) rather than effector DN2?”

6. The review’s therapeutic targeting section has limited dengue relevance. TLR7 inhibition is contraindicated during active DENV viraemia. IL-21 (Ansari2025) remains the more appropriate dengue therapeutic target. No wiki changes needed on this point.

Note: This council report is advisory only. No wiki pages have been modified. The curator must explicitly direct any changes based on council findings.