EFB Dengue Literature Review

Switched Memory B Cell

12 min read

Switched Memory B Cell

Overview

Switched memory B cells (sM, sometimes SWM) are isotype-class-switched, antigen-experienced memory B cells defined by flow cytometry as IgD⁻CD27⁺. Having undergone Class Switch Recombination, they express surface IgG or IgA (a minority are IgM-only), carry substantial Somatic Hypermutation, and are the canonical product of the Germinal Center reaction — the “textbook” long-lived, high-affinity memory pool that mediates fast recall.

In the IgD-vs-CD27 gating quadrant, sM (IgD⁻CD27⁺, lower-right) sits directly beside the DN population (IgD⁻CD27⁻, lower-left): both are class-switched, differing only by CD27. This adjacency makes sM the indispensable germinal-center comparator for the wiki’s extrafollicular/atypical spine — it is the GC-derived reference against which DN/DN2 EF effectors are contrasted (see DN2 Gating Strategy). sM can be further split by CD21 into resting switched memory (CD21⁺CD38⁻) and activated switched memory (CD21⁻), the latter forming a phenotypic continuum toward the atypical/DN2 state.

Key Points from Literature

  • CD27 was historically the universal memory marker, and CD27⁺ switched cells consistently carry significant somatic hypermutation, originating primarily from GC reactions (see Wei2007 - DN Memory B Cells in SLE, n=29 HC + 36 SLE cross-sectional, citing Klein et al. 1998).

  • sM and DN2 are the two B cell subsets epigenetically closest to ASCs, yet driven by different transcription-factor programmes: sM by NF-κB/EBF/OCT2 (GC-associated), DN2 by T-BET/AP-1/EGR (EF-associated). This establishes sM as a genuinely distinct differentiation endpoint — GC-derived memory — not an alternate maturation stage of the EF lineage (see Scharer2019 - Epigenetic Programming in SLE B Cells, RRBS + ATAC-seq, n=9 SLE + 12 HC).

  • DN1 is transcriptionally near-identical to sM (only 22 DEGs by RNA-seq), expresses TCF7 and CXCR5, and is proposed as an early sM precursor that has not yet acquired CD27 — placing DN1 and sM as immediate neighbours while DN2 diverges (see Jenks2018 - DN2 B Cells and EF Pathway in SLE, RNA-seq + ATAC-seq). UMAP projection of the composite COVID-19 dataset confirmed DN1 and sM occupy overlapping space, clearly separable from the DN2 cluster (see Woodruff2020 - EF B Cell Responses in COVID-19, Fig. 2c–d).

  • Autoreactive clones concentrate in the switched memory compartment: VH4.34-encoded autoreactive B cells within switched memory (CD27⁺IgD⁻) were 16.2 ± 11.9% in SLE vs. 1.3 ± 0.3% in healthy controls (P=0.03, n=6 SLE), normalising to 1.92 ± 0.7% after rituximab reconstitution. Residual B cells surviving 95–99% rituximab depletion are predominantly switched-memory phenotype (CD20⁺CD38^low CD27⁺IgD⁻) (see Anolik2004 - Rituximab and B Cell Abnormalities in SLE, n=6 SLE).

  • An IgM-only memory layer exists within the IgD⁻CD27⁺ gate: elevated IgM sequence frequency in the IgD⁻CD27⁺ compartment in SLE (20.9–68.1% vs 1.5–37.5% in vaccinated controls) is attributed to GC-independent EF generation of IgM-only memory — the first memory layer laid down before class switch (see Tipton2015 - ASC Diversity and Origin in SLE, citing Dogan et al. 2009). Note: a switched-memory gate defined on IgD⁻CD27⁺ alone (without IgG/IgA confirmation) will include this IgM-only fraction.

  • CD27 is not a stable lineage marker — this blurs the sM↔DN boundary: CD27 can be downregulated or shed (CD70/TLR/cytokine-induced; ADAM17 cleavage in high-inflammatory settings), so CD27⁻ status does not prove a cell never expressed CD27. Within the CD27⁺ memory compartment, T-bet⁺/FcRL5⁺ “memory ABC” cells are poised for ASC differentiation, while T-bet⁻/FcRL5⁻ canonical memory cells are stem-like (see Sanz2025 - Human Atypical B Cells Overview, invited review).

  • Dengue — switched memory is reprogrammed and recalled: DENV-specific class-switched IgD⁻ MBCs are significantly higher in secondary than primary immunity (p<0.001) and durable to 18 months (see Singh2026 - DENV-Specific Memory B Cell Subsets, n=58 samples, longitudinal). The acute IgG-dominant plasmablast burst reflects recall of class-switched memory (see Wrammert2012 - Plasmablast Responses in Acute Dengue, n=46 cohort), is functionally memory-derived via original-antigenic-sin serotype bias (see Priyamvada2016 - Cross-Reactive Memory Plasmablasts in Secondary Dengue, n=4 secondary DHF), and is driven by a Tph→IL-21→class-switched memory→plasmablast axis (see Ansari2025 - Peripheral T Helper Subset Drives B Cell Response in Dengue, n=170 cohort, T-B coculture). sM is the resting reservoir this recall draws from.

  • ★ Switched memory is NOT a reliable proxy for germinal-center origin. The use of class-switch status as indirect evidence of GC participation assumed CSR occurs predominantly within GCs; newer work indicates CSR takes place primarily prior to GC entry, with switched cells then gaining a competitive advantage inside GCs and becoming enriched over the reaction. GC-specific fate mapping confirms the proxy fails in both directions: GC-independent early MBCs include a significant fraction of class-switched cells, while GC-derived MBCs can remain IgM⁺ or IgM⁺IgD⁺ (see Glaros2025 - Multilayered Identity of B Cell Memory, review, no original data, mouse). Wiki consequence: the sM (IgD⁻CD27⁺) gate remains the best available GC-comparator population for the DN/EF story, but it should be described as the canonical/classical memory comparator rather than as a GC-origin-certified one.

  • Long-lived switched memory can be marginal-zone-phenotype. Long-lived anti-smallpox memory B cells with a marginal-zone phenotype and transcriptional signature dominated the switched memory compartment in the spleens of individuals immunized decades earlier — though whether this reflects preferential generation by that vaccine or preferential long-term survival is unclear (see Glaros2025 - Multilayered Identity of B Cell Memory, review, human). See Tissue-Resident Memory B Cell.

  • Where a booster is given can change what switched memory does. Memory B cells in draining lymph nodes reenter secondary GCs more efficiently than those in non-draining nodes, with transcriptional differences underpinning distinct intranodal localization — implying that administering a booster in the same versus the opposite arm can influence the quality of the resulting response (see Glaros2025 - Multilayered Identity of B Cell Memory, review, mouse).

  • ★ The sM↔DN arrow runs in both directions — RNA velocity does not settle precedence. Single-cell RNA velocity analyses have been reported showing a high flow from DN1 → sM (DN1 as sM precursor) and a high flow from sM → DN1 (DN1 as sM progeny), plus a flow from unswitched memory → DN2/DN3. A 2023 review that surveys both concludes the question is open: “Some DN1 cells could be precursors of SM B cells, while others could be progeny of SM B cells” (see Beckers2023 - Origins and Functions of DN B Cells, review, citing Stewart 2021 for both velocity analyses; see Single-Cell RNA Sequencing).

  • The clonal relationship between sM and DN is real but small: 0.2–2.2% overlap. AIRR sequencing of paired sM and DN compartments found only a 0.2–2.2% clonal overlap, with a similar mutation load in the clonally related cells — so a shared-origin route exists but accounts for a minority of either compartment. The majority of DN and sM cells are clonally distinct, with different IgV(D)J family and gene usage (see Beckers2023 - Origins and Functions of DN B Cells, review, own AIRR data; see BCR Sequencing).

  • A repertoire study reading the arrow the other way. One AIRR analysis found genealogical trees pointing to DN cells as the progenitor of sM cells, with no significant differences in VH gene family usage or in the types and locations of Ig mutations between class-switched DN and sM cells; those authors concluded most class-switched DN cells are memory cells that acquired their mutations by the same GC-derived mechanism as sM cells (see Beckers2023 - Origins and Functions of DN B Cells, review, citing Wu 2011).

  • The proposed mechanism for sM → DN conversion is immunosenescence. CD27 downregulation under chronic antigen stimulation — during normal aging, chronic viral infection, and some autoimmune diseases — is the route by which sM cells would become DN, previously described as “exhausted memory B cells.” The review offers no transcriptional or epigenetic mechanism for the downregulation itself (see Beckers2023 - Origins and Functions of DN B Cells, review; see CD27).

  • Corroboration of the 22-DEG figure from an independent review. The near-identity of DN1 and sM transcriptomes (22 differentially expressed genes) is restated as one of the two strongest arguments that DN1 and sM are developmentally adjacent (see Beckers2023 - Origins and Functions of DN B Cells, review, citing Jenks 2018).

  • [2026-08-27] Classic switched memory falls in IgG4-related disease while a CXCR5⁻ fraction of the same compartment expands. Splitting IgD⁻CD27⁺ cells by CXCR5 (n=38, 13-colour FCM): CXCR5⁺ classic/resting switched memory decreased and the CXCR5⁻ “ABC-like” fraction expanded, both p<0.01, while total switched memory was unchanged — the shift is internal to the compartment and would be invisible to a panel without CXCR5. Unswitched memory / marginal-zone cells fell as a share of total B cells (p<0.01) (see Allard-Chamard2023 - DN3 B Cells Infiltrate Inflamed Tissues, n=38).

  • [2026-08-27] Switched memory B cells are largely absent from tissue T-B conjugates. In IgG4-RD salivary gland (n=4), cell-cell contacts with T cells occurred primarily between DN B cells and T cells and between plasmablasts and T cells, and rarely between switched CD27⁺CD20⁺ memory B cells and T cells — despite switched memory being present in the tissue. In blood, switched memory did not discriminate IgG4-RD from health (logistic regression OR 1.001, p = 0.9824) (see Allard-Chamard2023 - DN3 B Cells Infiltrate Inflamed Tissues, n=4 tissue IF + n=38 blood).

  • Two things this review adds about the switched memory gate. First, the Bm1–Bm5 scheme “coalesces different types of memory B cells” and does not separate resting naive from IgD⁺ unswitched memory — part of the case for IgD/CD27 as the primary scheme (see Bm Classification). Second, and quantitatively: CXCR5⁻ switched memory cells do express T-bet, at an MFI significantly lower than DN2 and activated naive but not zero — so a loosely drawn T-bet⁺ gate will include SWM (see Sanz2019 - Consistent Classification of Human B Cell Populations, review — no original data).

  • [2026-08-29] On the Emory lupus panel, switched memory is a derived quantity, not a directly gated one. Switched memory frequency was obtained by subtracting plasmablasts (IgD⁻CD27⁺⁺CD38⁺⁺) from the IgD⁻CD27⁺ compartment rather than by positive gating — so PB gate placement propagates directly into the SM number (see Jenks2021 - B Cell Subset Composition in Cutaneous Lupus, n=207 + 46 HCD, cross-sectional). Practical consequence for panel design: SM and PB are not independent measurements on this scheme.

  • [2026-08-29] Switched memory is the subset where lupus patients are most heterogeneous. Rather than shifting as a group, 17% of SLE+/CCLE−, 19% of SLE+/CCLE+ and 16% of primary CCLE patients had switched-memory frequencies more than 2 SD above the healthy-donor mean — a minority-stratum pattern rather than a group difference. DN1, by contrast, was described as transcriptionally linked to resting switched memory (see Jenks2021 - B Cell Subset Composition in Cutaneous Lupus).

  • Switched memory is the benchmark DN2 is measured against — and DN2 comes close. In the 2.5-day sorted-subset assay, conventional switched memory (IgD^neg^CD27⁺) and DN1 memory formed IgG ASCs efficiently while naive B cells failed entirely; T-bet^hi^ DN2 cells produced only 2–3-fold fewer ASCs than switched memory (and ≥50-fold more than naive) (see Zumaquero2019 - IFN-gamma Programs T-bet-hi B Cells for ASC Differentiation, 3 independent experiments, 3 SLE donors). Separately, RNA-seq of Be1-derived B_DN_ cells was enriched for ASC-associated genes relative to switched memory (p<0.001) as well as to naive and mature B cells (see Zumaquero2019 - IFN-gamma Programs T-bet-hi B Cells for ASC Differentiation, RNA-seq n=3/subset). Switched memory therefore remains the most differentiation-competent conventional subset, but the CD27⁻ DN2 population is functionally within reach of it.

  • A numerically small but functionally competitive slice of the switched memory pool, held in a different place. BrdU given from day 4 for 10 days and read at day 30 put T-bet⁺CD11c⁺ cells at ~5% of BrdU⁺ IgD^lo^CD38⁺GL-7⁻ memory B cells. On sorted transfer into MD4 recipients they nonetheless produced a recall anti-LCMV IgG response comparable to memory B cells lacking the phenotype and higher than naive. i.v. anti-CD45 labelled the majority of T-bet⁺CD11c⁺ memory cells against only ~30% of total memory B cells, i.e. the subset keeps its marginal-zone position into memory while the bulk memory pool does not (see Song2022 - Tfh Outside Germinal Centers Drive T-bet CD11c B Cells, mouse, LCMV-Armstrong, n=3–5 mice/group).

Contradictions & Debates

Memory B Cell, Double-Negative B Cell, DN2 B Cell, Activated Naive B Cell, CD27, IgD, CD21, Class Switch Recombination, Somatic Hypermutation, Germinal Center, Plasmablast, Extrafollicular Response, Original Antigenic Sin, DN2 Gating Strategy, Early Memory B Cell, Tissue-Resident Memory B Cell, Class Switch Recombination, DN3 B Cell

Sources

Graph View

Backlinks