Distributed clonal deletion prevents autoimmune disease progression Authors: Anna M. Newen, Uzair A. Ansari, Mikala J. Simpson, Fiona Flynn, Sukriti Sharma, Michael Ivanov, Ishaan Antani, Dylan Pfannenstiel, Baktiar Karim, Laura Bassel, Brian Capaldo, Qingrong Chen, Daoud Meerzaman, Indu Raman, Chengsong Zhu, Cornelius Y. Taabazuing, Hamid Kashkar, Christian T. Mayer Published: 2026 ABSTRACT SUMMARYSelf-reactive B cells arise during development and can increase pathogenicity through activation-induced cytidine deaminase (AID)-mediated diversification. Clonal deletion is thought to eliminate these cells, yet how deletion is distributed across developmental and activation stages to prevent autoimmune disease remains unclear. Here, we show that self-tolerance is enforced through temporally distinct mitochondrial outer membrane permeabilization (MOMP) checkpoints. Using conditional Bcl-2 expression to inhibit MOMP either from B cell development or activation, we find that early inhibition permits survival of autoreactive B cells after peripheral egress, expanding the pool available for activation and AID-dependent diversification. This results in broadened class-switched IgG autoreactivity, complement activation, kidney pathology, and drives lethal autoimmune disease. In contrast, post-activation MOMP inhibition promotes autoreactive cell accumulation and autoantibody production but causes limited tissue damage and normal survival. Together, these findings support a Distributed Clonal Deletion Model in which temporally distinct checkpoints cooperate to constrain autoimmune disease progression. INTRODUCTION Systemic autoimmune diseases such as systemic lupus erythematosus (SLE) arise when immune tolerance fails and B cells generate antibodies that target self-tissues. These conditions disproportionately affect women and can lead to chronic inflammation, renal failure, and organ damage. Autoreactive B cells are abundant in healthy individuals,1,2 yet most never produce autoreactive antibodies. Additionally, some individuals produce autoantibodies without developing disease.3–5 These three levels of tolerance - autoreactive B cells, autoantibodies, and autoimmune disease - remain incompletely understood. Defining how tolerance mechanisms operate at each level and cooperate to prevent disease progression is essential for understanding why tolerance breaks down in autoimmunity. B cell development generates a diverse preimmune repertoire through recombination-activating gene (RAG)–mediated V(D)J recombination in the bone marrow. Pro-B cells (Fraction B/C) rearrange immunoglobulin (Ig) heavy chains and proliferate as large pre-B cells (Fraction C′), followed by light-chain rearrangement in small pre-B cells (Fraction D), and progression to immature B cells (Fraction E).6–8 After leaving the bone marrow, immature B cells mature through transitional stages and enter the follicular (FO) or marginal zone compartments in the spleen.8–10 A long-standing but unresolved question is whether immature B cells destined for deletion die within the bone marrow or after peripheral egress,11–13 a distinction that has been difficult to resolve because dying cells are rapidly cleared in vivo. Upon antigen encounter in T cell-dependent immune responses, FO B cells express activation-induced cytidine deaminase (AID). AID mediates somatic hypermutation (SHM), primarily within germinal centers (GCs), which can generate or enhance autoreactivity, and it mediates class-switch recombination (CSR), which frequently occurs early during immune responses and can take place outside GCs.14–17 SHM alters antigen specificity, whereas CSR changes antibody isotype and thereby modifies effector properties. Because both RAG and AID can generate or modify autoreactivity, multiple apoptosis checkpoints censor autoreactive B cells,12,13,18–25 and genetic models that attenuate apoptosis are associated with autoimmune disease.26–32 Many of these checkpoints depend on mitochondrial outer membrane permeabilization (MOMP), a central event in the intrinsic apoptotic pathway, although MOMP can also influence cellular outcomes beyond apoptosis. Yet, autoreactivity and autoantibody production do not always progress to disease.1–5,33 Whether checkpoints before and after activation operate independently or cooperatively remains unclear, in part because prior studies of individual tolerance checkpoints have not directly examined how the timing of checkpoint failure across B cell compartments influences progression from autoreactivity to autoimmune disease. We hypothesized that clonal deletion is distributed across developmental and activation stages and that checkpoint timing determines disease outcomes by controlling which autoreactive B cells undergo AID-dependent diversification. Early checkpoints acting before B cell activation would restrict low-avidity autoreactive clones from undergoing CSR and SHM, whereas late checkpoints acting after activation would limit the persistence of already-diversified clones but act on an already-restricted substrate. Under this model, failure of early checkpoints would broaden the class-switched autoreactive repertoire and drive autoimmune disease progression, whereas failure of late checkpoints would permit autoreactive cell accumulation and autoantibody production without causing organ damage. To test this hypothesis, we used identical Rosa26LSL–Bcl2 alleles driven by distinct Cre recombinases to express Bcl-2 either beginning early in B cell development and persisting throughout subsequent stages (Bcl2Early) or only after activation (Bcl2Late). Bcl-2 inhibits MOMP, enabling direct comparison of the consequences of disrupting early versus late MOMP-sensitive tolerance checkpoints. We combined these models with in vivo EdU pulse-chase labeling, comprehensive disease assessment, and autoantibody profiling to define how checkpoint timing regulates B cell progression, AID-dependent diversification, and autoimmune disease outcomes. RESULTS Early but not late MOMP inhibition causes autoimmune disease We generated mice expressing human Bcl-2 (hBcl-2) and GFP from the Rosa26LSL–Bcl2 allele using Mb1Cre or AicdaCre to inhibit MOMP either beginning early in B cell development and persisting thereafter (Bcl2Early) or only after activation (Bcl2Late). Flow cytometry confirmed stage-specific Cre recombination (Figure S1 and S2A), and GC, switched memory B (swMem), and plasma cells (PCs) expressed comparable hBcl-2 levels (Figures S2B and S2C), enabling direct comparison. To determine how the timing of MOMP inhibition influences autoimmune disease progression, we monitored disease development in Bcl2Early, Bcl2Late, and control cohorts. Only Bcl2Early mice developed clinical illness, with reduced survival and a higher incidence of disease in females (Figures 1A and 1B). Renal histopathology revealed widespread glomerulonephritis and tubular injury in Bcl2Early mice, characterized by glomerular hypercellularity, capillary wall thickening, and crescent formation, together with inflammatory infiltrates (Figure 1C). Quantitative scoring confirmed significantly greater pathology in this group than in Bcl2Late and control mice, with more severe changes in females (Figures 1D–1F). Ig deposition within glomeruli was comparable between Bcl2Early and Bcl2Late mice, whereas cortical Ig deposition was increased in Bcl2Early kidneys (Figure 1G), indicating that differences in disease severity are not explained by glomerular Ig deposition alone. In contrast, complement C3d deposition was increased only in Bcl2Early mice (Figure 1H). Serum urea measurements indicated impaired renal function in a subset of Bcl2Early mice (25% of females; 0% of males) (Figures 1I and 1J), whereas Bcl2Late mice maintained normal urea levels despite the presence of autoreactive antibodies (Figure 1I). Together, these findings demonstrate that the disruption of early MOMP-sensitive checkpoints, together with subsequent failure of later checkpoints, is sufficient to drive autoimmune disease progression, whereas the disruption of post-activation checkpoints alone permits autoreactivity and autoantibody production without progression to organ damage. Bcl-2 expression from B cell development restricts IgG autoantibody breadth To determine whether checkpoint timing affects autoreactive antibody specificity, we performed autoantigen array analysis (Figures 2A–2C). Bcl2Early mice exhibited a marked expansion of IgG autoreactivity, with 23 significantly increased autoreactivities compared to 5 in Bcl2Late mice relative to controls (Figures 2A and 2C; Data S1). In contrast, IgM autoreactivity was increased but more comparable between Bcl2Early and Bcl2Late groups (Figures 2B and 2C; Data S2). Consistent with this, 13 IgG autoreactivities were significantly increased in Bcl2Early compared with Bcl2Late mice, whereas only a single IgM reactivity differed (Figure 2C), indicating that early checkpoint failure selectively expands class-switched autoreactive breadth rather than overall autoreactivity. Notably, among the most expanded IgG reactivities in Bcl2Early mice were antibodies targeting complement C3 and PCNA, which have been associated with severe nephritis in subsets of patients with SLE.34,35 Conversely, IgG reactivities against M2 and nucleolin were significantly reduced in Bcl2Early mice relative to controls, whereas apparent decreases in Sm and histone reactivities did not reach statistical significance. Total autoreactive IgG titers and dsDNA/ssDNA ELISAs were similar between Bcl-2 models (Figure S3A–E), with ELISA confirmation of anti-C3 and anti-PCNA IgG autoantibody differences (Figures S3F and S3G). To independently assess autoreactive repertoires, we performed phage immunoprecipitation sequencing (PhIP-seq). Although numerous peptide reactivities were detected across all groups, including controls (Figure S3H), gene set enrichment analysis (GSEA) of peptide-associated genes revealed qualitative differences between genotypes. Bcl2Early mice showed selective enrichment for gene sets related to lipid translocation, intracellular ligand-gated ion channel activity, transcriptional coactivator function, and immune response-associated transcriptional programs (Data S3), whereas Bcl2Late mice did not show enrichment of these categories. These findings indicate distinct functional profiles of autoreactive IgG repertoires despite broadly comparable numbers of individual peptide reactivities across groups. Focusing on peptides enriched in at least two of three Bcl2Early mice but absent from Bcl2Late and control mice identified 12 Bcl2Early-specific reactivities, including Prkcz, which is associated with autoimmune myopathies in humans36 (Figure 2D). Together, these data demonstrate that early checkpoint failure selectively widens class-switched autoreactive IgG breadth, whereas late checkpoint failure permits autoreactivity without broad class-switched diversification. AID is required for autoimmune pathology when MOMP is inhibited before activation Given the increased breadth of class-switched autoantibodies in Bcl2Early mice, we next tested whether AID-dependent processes are required for autoimmune disease in this setting. To do so, we analyzed Bcl2Early mice lacking AID (encoded by Aicda), thereby preventing SHM and CSR. Strikingly, Bcl2EarlyAicda−/− mice remained free of autoimmune disease and showed no renal injury (Figures 3A–3G), despite exhibiting the same early checkpoint disruption as Bcl2Early mice. Histopathologic evaluation revealed infection-associated mammary gland abscesses (Figures 3B and 3C), but no autoimmune pathology (Figures 3D–3G). One Bcl2Early mouse developed B cell lymphoma (Figures 3B and 3C). These findings demonstrate that AID-dependent diversification is required for autoimmune disease in the setting of early MOMP-regulated checkpoint failure, and that autoreactive B cell survival without CSR/SHM is insufficient to cause autoimmune organ damage. MOMP inhibition before activation expands peripheral B cell populations To determine how early MOMP checkpoint disruption alters the B cell pool available for activation, we analyzed B cell development in Bcl2Early and Bcl2Late mice (Figure 4A). Despite Bcl-2 expression beginning at the pro-B stage, bone marrow pro-B, pre-B, and immature B cells were not expanded in Bcl2Early mice, whereas a reduction in large pre-B cells (Fraction C′) was noted, consistent with decreased proliferation (Figure 4B). In contrast, Bcl2Early mice displayed significant expansion of splenic transitional (T1, T2), anergic (T3), and FO B cells, as well as increased recirculating mature B cells in the bone marrow (Figures 4B and 4C). Bcl2Late mice resembled controls across these compartments (Figures 4B and 4C). Both Bcl2Early and Bcl2Late mice accumulated post-activation GC B cells, switched memory B cells (swMem), and PCs. These findings indicate that early, but not post-activation, MOMP inhibition selectively expands peripheral immature and mature B cell populations. Immature B cell deletion occurs predominantly after bone marrow egress The absence of immature B cell expansion in Bcl2Early mice suggested either limited apoptosis at this stage or accelerated marrow egress. To test the latter, we performed vascular labeling of bone marrow sinusoids11 (Figure 4D). Immature B cells labeled by intravenously injected anti-CD45R/B220-PE-Cy7 did not increase in proportion or number in Bcl2Early mice relative to controls (Figures 4E and 4F), excluding enhanced marrow egress as a cause of the normal immature pool. Increased numbers of labeled mature B cells (Fraction F) were detected, reflecting expansion and increased bone marrow egress of this compartment (Figures 4B–4E and 4F). To determine whether non-apoptotic programmed cell death contributes to immature B cell loss, we analyzed Gsdmd−/− Gsdme−/− Mlkl−/− triple-deficient mice, which are unable to undergo pyroptosis and necroptosis (Figure 4G). Bone marrow and splenic B cell subsets were comparable to controls with only modest changes in Fraction C′, T3, and swMem cells (Figures 4H and 4I), while macrophages resisted pyroptotic and necroptotic stimuli (Figure S4A). These findings rule out pyroptosis and necroptosis, supporting a MOMP-sensitive peripheral deletion mechanism. Transitional B cells are the first peripheral stage regulated by MOMP To define how MOMP inhibition influences developmental progression, we performed EdU pulse-chase labeling (Figure 5A). In control mice, EdU labeled Fraction C′ within two hours and appeared in Fraction E, T1, T2, and T3 by day 3, with labeled transitional and anergic populations contracting by day 7 and labeled FO and Fraction F cells persisting (Figures 5B and 5C). In Bcl2Early mice, EdU labeling of Fraction C′ cells was reduced at two hours (Figure 5B), consistent with decreased proliferation. Fraction E labeling on day 3 was unchanged, whereas splenic EdU+ T1, T2, and T3 cells were significantly increased at day 3 and remained elevated at day 7 (Figure 5C). Early FO cells appeared similarly at day 3; however, by day 7, FO and Fraction F B cells were significantly expanded in Bcl2Early mice (Figure 5C). These findings identify transitional and anergic B cells as the first peripheral populations whose survival is limited by MOMP. FO-intrinsic MOMP inhibition is insufficient to account for peripheral expansion To assess the contribution of FO-intrinsic MOMP inhibition, we examined CD21CreRosa26LSL–Bcl2 (Bcl2Int) mice, in which recombination is efficient in FO B cells and partial in T2 and anergic T3 cells37 (Figure S4B). FO B cells were increased in Bcl2Int mice relative to controls but remained substantially fewer than in Bcl2Early mice (Figure 5D). Transitional and anergic T3 B cell numbers in Bcl2Int mice did not differ from controls (Figure 5E; Figures S4C and S4D). These findings indicate that FO-intrinsic MOMP inhibition contributes to but is insufficient to account for the Bcl2Early phenotype, highlighting the importance of early peripheral checkpoints acting at transitional stages. Together, these findings suggest that early MOMP checkpoint disruption permits the survival of autoreactive B cells after peripheral egress, leading to the expansion of transitional and FO B cell populations, increased class-switched IgG breadth, and AID-dependent auto immune disease, whereas post-activation checkpoint failure permits autoreactivity without progression to pathology. DISCUSSION This study demonstrates that the timing of MOMP-sensitive checkpoints determines whether autoreactive B cells progress to autoimmune disease. Inhibition of MOMP before B cell activation permits the survival and expansion of peripheral B cell populations, leading to increased class-switched autoreactive IgG breadth and severe, AID-dependent autoimmune pathology. In contrast, MOMP inhibition restricted to activated and post-GC stages allows autoreactive B cell accumulation and autoantibody production but does not result in organ damage. These findings provide a mechanistic explanation for how autoreactivity and autoantibodies can be present in the absence of clinical disease.1,3,4 Our results extend prior work on apoptosis and MOMP-sensitive checkpoints in B cells by directly comparing the consequences of disrupting early versus late MOMP-sensitive stages. Previous studies demonstrated apoptosis in developing and GC B cells11,38–40 and identified a post-GC checkpoint that limits autoreactive PC and memory B cell survival.33,41 However, how checkpoint timing influences progression from autoreactivity to autoimmune disease remained unresolved. Here, we show that early checkpoint disruption uniquely permits disease progression, whereas late checkpoint disruption alone is insufficient in our model, establishing that these checkpoints make non-equivalent contributions to immune tolerance. Because Bcl2Early mice continuously express Bcl-2 throughout B cell development and activation, the severe autoimmune phenotype likely reflects cumulative disruption of multiple sequential MOMP-sensitive checkpoints, including both pre-activation and post-activation stages. Mechanistically, early MOMP inhibition expands transitional and FO B cell populations without altering immature bone marrow subsets, indicating that a major clonal deletion checkpoint operates after peripheral egress. This extends prior apoptotic B cell quantification11 by defining functional consequences through genetic loss-of-function approaches. EdU tracing further identifies transitional B cells as the first stage at which survival is MOMP-limited, with downstream expansion of FO populations. These findings suggest that early peripheral checkpoints constrain the size and composition of the B cell pool entering immune responses, whereas later checkpoints primarily regulate the persistence of activated clones. The selective expansion of class-switched autoreactive IgG breadth in Bcl2Early mice provides a functional link between checkpoint timing and disease outcome. In contrast, Bcl2Late mice exhibited autoreactive antibodies without comparable IgG diversification or pathology. Despite comparable IgM autoreactivity and similar total autoreactive IgG titers in Bcl2Early and Bcl2Late mice, early checkpoint disruption broadened IgG specificity, including reactivities associated with severe SLE manifestations.34,35 Although autoreactive IgM can contribute to autoimmune pathology in certain settings, including IgM-mediated autoimmune disorders42 and AID-deficient models,43,44 the relatively limited pathology observed in Bcl2Late mice despite increased IgM autoreactivity suggests that expansion and diversification of class-switched autoreactive IgG is more closely associated with severe autoimmune disease progression in this model. Together with the complete protection observed in AID-deficient Bcl2Early mice, these findings demonstrate that AID-dependent diversification is required for disease and that survival of autoreactive B cells alone is insufficient to drive pathology in this model. The selective enrichment of particular autoreactivities in Bcl2Early mice likely reflects processes acting downstream of checkpoint failure, including stochastic repertoire generation, antigen accessibility, clonal competition, affinity-based selection, and ongoing diversification during immune responses. Because disease develops with age, these autoreactive profiles may evolve over time as autoreactive clones undergo further class switching and selection, rather than being fixed solely by the initial checkpoint defect. Thus, our data support a model in which early checkpoint disruption broadens the substrate available for pathogenic evolution, while the eventual emergence of specific autoreactivities reflects subsequent selection pressures. Because Bcl2Early mice continuously express Bcl-2 throughout subsequent developmental and activation stages, these autoreactive profiles likely reflect cumulative effects across multiple sequential MOMP-sensitive checkpoints rather than a single transitional-stage event. Because our experiments were performed in genetically controlled mice housed under specific pathogen-free conditions, environmental variation is likely less important in this setting than it may be in human autoimmune disease. Prior studies have demonstrated that SHM can generate pathogenic autoreactivity from autoreactive naive B cell precursors that escape early tolerance checkpoints.33,45,46 Although these observations do not establish SHM-dependence for individual autoreactivities in our model, they support the concept that broadened autoreactive precursor pools may undergo progressive diversification and selection during disease evolution. More broadly, defective early B cell tolerance checkpoints have been described across numerous human autoimmune diseases and are associated with the accumulation of autoreactive naive B cell populations47 that may subsequently undergo pathogenic diversification. Our findings are conceptually consistent with human inborn errors such as activated PI3Kδ syndrome (APDS), in which enhanced B cell survival together with dysregulated GC responses and altered immune selection are associated with broad autoreactivity and autoantibody production,48 although through mechanisms extending beyond altered B cell survival alone. These findings may help explain how similar early tolerance defects give rise to distinct autoimmune phenotypes depending on which autoreactive clones ultimately undergo expansion and diversification. AID has also been implicated in B cell tolerance,44,49–51 and AID deficiency has been reported to either attenuate52,53 or exacerbate54,55 autoimmune pathology depending on the model. In our system, AID deficiency completely prevented disease despite early checkpoint disruption, indicating that the pathogenic effects of AID-dependent diversification outweigh potential tolerance-promoting functions under these conditions. These findings suggest that the impact of AID on autoimmunity is context-dependent and shaped by the availability of autoreactive precursors. AID deficiency has also been associated with impaired control of commensal microbiota,56 which may contribute to secondary inflammatory or infectious phenotypes and could explain the infection-associated mammary abscesses observed in Bcl2EarlyAicda−/− mice in our study. Notably, Ig deposition within glomeruli was observed in both Bcl2Early and Bcl2Late mice, yet complement activation and renal pathology occurred only in Bcl2Early animals. This dissociation suggests qualitative differences in the autoreactive antibody response rather than differences in total antibody deposition. One possibility is that the antigenic targets differ, as Bcl2Early mice exhibit broader IgG reactivity, including antibodies such as anti-C3 that have been shown to enhance complement activation by interfering with regulatory mechanisms or stabilizing C3 convertases.57,58 Alternatively, differences in antibody isotype may influence complement fixation and pathogenic potential. These considerations indicate that the composition of the autoreactive repertoire, rather than the presence of immune complexes alone, is a key determinant of downstream tissue injury. These observations support a model in which clonal deletion is distributed across pre-activation and post-activation stages. Because receptor editing efficiently removes high-avidity autoreactive B cells in the bone marrow with minimal reliance on deletion,11,59,60 we propose that early peripheral checkpoints limit the entry of low-affinity or low-avidity autoreactive B cells into immune responses, thereby restricting the breadth of class-switched autoreactive repertoires, whereas later checkpoints constrain the persistence and expansion of autoreactive clones once activated (Figure 6). Failure of early checkpoints expands the pool of cells available for diversification and enables pathogenic autoantibody responses, while failure of late checkpoints permits autoreactivity without disease progression (Figure 6). This framework, which we term the Distributed Clonal Deletion Model, provides a unifying explanation for how substantial autoreactivity can be tolerated in healthy individuals yet lead to disease when early tolerance mechanisms fail. These findings have implications for human autoimmune disease, where defects in early tolerance checkpoints or altered survival of autoreactive precursors may expand the repertoire available for pathogenic diversification. Targeting pathways that promote autoreactive B cell deletion or limit class-switched autoantibody diversification and complement activation, rather than global B cell depletion, may therefore represent a strategy to reduce autoimmune pathology while preserving immune competence. Limitations of the study This study has several limitations. Bcl-2 inhibits MOMP but may also affect additional cellular processes, and thus our findings define MOMP-sensitive checkpoints rather than apoptosis-specific mechanisms. In addition, although AID dependence is demonstrated, the relative contributions of SHM and CSR were not resolved. RESOURCE AVAILABILITY Lead contact Requests for further information and resources should be directed to and will be fulfilled by the lead contact, Christian T. Mayer (christian.mayer@nih.gov). Materials availability This study did not generate new unique reagents. STAR★METHODS EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS Mice AicdaCre,61 C57BL/6J, Gsdmd−/−,62 Gsdme−/− 63 and Mb1Cre64 mice were from Jackson Laboratories. Aicda−/−65 mice were provided by Dr. Michel Nussenzweig (The Rockefeller University). CD21Cre37 mice by Dr. Jagan Muppidi (NCI). Mlkl−/−66 mice by Dr. James Murphy (WEHI) and Dr. Alan Sher (NIAID), and Rosa26LSL–Bcl2–IRES-GFP mice67 by Dr. James Murphy (WEHI) and Dr. Alan Sher (NIAID), and Rosa26LSL–Bcl2–IRES–GFP mice67 by Dr. Hamid Kashkar (University of Cologne). All mice were on C57BL/6 background. Mice were maintained under specific pathogen-free conditions at NCI (Frederick and Bethesda). Male and female mice between 7 and 52 weeks of age were used as indicated in the figure legends. Female Bcl2Early mice exhibited more severe autoimmune pathology and reduced survival than males. Therefore, female mice were preferentially used in experiments examining autoimmune disease pathogenesis and renal pathology, whereas both sexes were included in developmental and mechanistic analyses unless otherwise indicated. All procedures were approved by the NCI Animal Care and Use Committee (ACUC; #26–056 and EIB-114) and conformed with federal regulatory requirements and standards. The intramural NIH ACU program is accredited by AAALAC International. Primary bone marrow-derived macrophage cultures Bone marrow-derived macrophages (BMDMs) were generated from mouse bone marrow cells cultured in RPMI 1640 supplemented with 10% heat-inactivated fetal bovine serum, penicillin/streptomycin, and recombinant Fc-tagged human M-CSF (amino acids 33 to 190; produced and purified in-house) for 6 days prior to stimulation experiments. METHOD DETAILS Study design This study compared the effects of MOMP inhibition before or after B cell activation on B cell accumulation, autoantibody formation, and autoimmune pathology. Experiments were designed to analyze B cell subset composition, developmental kinetics, serological responses, and pathological outcomes in Bcl2Early, Bcl2Late, and control mice. Both male and female mice were analyzed unless otherwise indicated. Female mice were preferentially used in disease-focused experiments because female Bcl2Early mice exhibited more severe autoimmune pathology and reduced survival. Sample sizes were based on prior studies demonstrating sufficient statistical power to detect differences in B cell subset frequencies and autoantibody reactivity. No animals or data points were excluded. Histopathologic scoring was performed in a blinded fashion as described below. In vivo labeling experiments For EdU pulse–chase experiments, mice were injected intravenously with 1 mg 5-ethynyl-2′-deoxyuridine (EdU) (Thermo Fisher Scientific, A10044) and analyzed 2 h, 3 d, or 7 d later. EdU incorporation was detected using the Click-iT Plus EdU Pacific Blue Flow Cytometry Assay Kit (Thermo Fisher Scientific, C10636) according to the manufacturer’s instructions. For bone marrow sinus labeling experiments, mice were injected intravenously with 2 μg anti-CD45R/B220-PE-Cy7 antibody (Thermo Fisher Scientific, 25-0452-82) and analyzed 2 min later by flow cytometry to identify cells exposed to the bone marrow sinusoids. Pathology Tissues were fixed in buffered 10% formalin (Azer scientific, PFNBF-120) for 48 h at room temperature and stored in 70% ethanol prior to paraffin embedding. Sections were stained with hematoxylin & eosin and evaluated by a board-certified veterinary pathologist. All slides were evaluated in a blinded fashion using an Olympus BX46 microscope at 200× magnification. Images were captured using a Nikon DS-Ri2 camera. Two pathologists, blinded to genotype, graded glomerulopathy on a scale from 0 to 4 based on the extent and severity of glomerular changes. Grade 0 represented normal glomeruli, without abnormalities. Grade 1 indicated minimal hypercellularity involving less than 25% of glomeruli. Grade 2 was defined by mild hypercellularity affecting 25–50% of the glomerular tufts. Grade 3 denoted moderate hypercellularity in greater than 50 to 75% of glomeruli, with or without crescent formation and thickening of basement membrane. Grade 4 represented marked hypercellularity in greater than 75% of glomeruli, accompanied by crescent formation and thickening of basement membrane. Cortical tubular lesions were graded on a scale from 0 to 4 based on the extent of cortical involvement. Grade 0 (normal) indicated normal tubules without any pathological changes. Grade 1 (minimal) represented tubular lesions, including degeneration and/or regeneration, affecting less than 10% of the cortical area. Grade 2 (mild) involved tubular lesions affecting more than 10% but less than 25% of the cortical tubules. Grade 3 (moderate) indicated lesions involving more than 25% but less than 50% of the cortical tubules. Grade 4 (severe) denoted tubular lesions affecting more than 50% of the cortex. Additionally, lymphoplasmacytic infiltrates in the kidney were semi-quantitatively graded on a scale from 0 to 4, where 0 was normal, 1 was minimal, 2 was mild, 3 was moderate, and 4 was severe. Enzyme-linked immunosorbent assay (ELISA) Serum IgG autoantibodies were measured by ELISA as described41 using peroxidase-conjugated goat anti-mouse IgG Fc (Jackson ImmunoResearch, 115-035-164). Self-antigens included dsDNA (Sigma, D4522), single-stranded DNA (ssDNA, prepared from dsDNA), mouse C3 (Complement Technology, M113), and human PCNA (Novus Biologicals, NBC1-18428). Serum was tested at a 1:640 dilution and at three 4-fold dilutions (dsDNA, ssDNA), or at a 1:160 dilution (C3, PCNA). Monoclonal control antibodies (mGO53, non-reactive; ED38, highly polyreactive) with mouse IgG1 constant regions were tested at 4μg/ml and at three 4-fold dilutions. Absorbance at 405nm was measured with a SpectraMax iD3 Multi-Mode Microplate Reader (Molecular Devices). A405 values were corrected for PBS-only signal and area under the curve (AUC) values were calculated using GraphPad Prism. Kidney function Serum urea concentrations were measured with the Urea Nitrogen (BUN) Colorimetric Detection Kit (Thermo Fisher Scientific, EIABUN) according to the manufacturer’s instructions. Measurements were performed with a SpectraMax iD3 Multi-Mode Microplate Reader (Molecular Devices). Flow cytometry Flow cytometry was performed as described.11 EdU was detected using the Click-iT Plus EdU Pacific Blue Flow Cytometry Assay Kit (Cat. C10636, Thermo Fisher). FLOWMIST assay for autoreactivity The flow cytometry assay was performed as described.11 Serum was diluted 1:160 and autoreactive antibodies detected with 1μg/ml AlexaFluor647-conjugated F(ab’)2 goat anti-mouse IgG Fc (Jackson ImmunoResearch, 115-606-071). End titers were the last dilution producing a mean fluorescence intensity (MFI) ratio >3 over negative control serum. Autoantigen array Autoantigen arrays were performed as described33 with the following differences: Serum was tested at 1:640 dilution in PBS. Autoantibodies were detected with Cy3-conjugated goat anti-mouse IgG(H + L) (Invitrogen, Cat. A10521) and with AlexaFluor633-conjugated goat anti-mouse IgM (Invitrogen, Cat. No. A21046) secondary antibodies. Phage immunoprecipitation sequencing (PhIP-Seq) PhIP-Seq was performed by CDI Labs as described.70,71 Counts provided by the vendor were analyzed using edgeR.68 Modifications to the edgeR workflow were followed as described.72 Gene set enrichment analyses and over representation analyses were performed using the clusterProfiler package with the msigdbr package (https://CRAN.R-project.org/package=msigdbr) used for the gene set database.69 Complement staining Staining was performed on 5μm FFPE sections using a manual benchtop method. Antigen retrieval was performed with EDTA buffer for 10mins at 100°C. Nonspecific binding was blocked with an incubation of 2% normal donkey serum for 20 min. This was followed by an overnight incubation at 4°C with the C3d primary antibody (R&D Systems, AF2655) at a 1:250 dilution. Antibody detection was accomplished by incubation with Donkey anti-Goat Alexa Fluor 594 (Invitrogen, A11058), followed by incubation in DAPI. Slides were digitally imaged at 20x using a Leica Aperio FL fluorescence digital scanner. Immunoglobulin (Ig) staining FFPE blocks were sectioned at 5μm in preparation for IgM, IgG, IgA IHC. Staining was performed using a Leica Bond RX autostainer (Leica Biosystems). Antigen retrieval was performed using EDTA for 20 min at 100°C on the Bond autostainer. Endogenous biotin was blocked using the Avidin-Biotin blocking kit (Vector Labs, SP-2001) per the manufacturer’s instructions. The primary antibody for IgM, IgG, and IgA (Southern Biotechnology, 1012-08) was diluted 1:100, with a 30-min incubation time. Antibody detection was accomplished using Alexa Fluor 660 conjugated Streptavidin (Invitrogen, S21377) followed by DAPI counterstain. Slides were digitally imaged at 20x using a Leica Aperio FL fluorescence digital scanner. Image analysis Image analysis was conducted using the Area Quantification FL v2.3.3 algorithm in HALO (Indica Labs, Albuquerque, NM). Analyses were conducted for immunoglobulin/complement staining within the cortex and glomeruli to quantify the percentage of positive area. Bone marrow-derived macrophage stimulation Day 6 bone marrow-derived macrophages (BMDM) were primed with 1μg/ml Pam3CSK4 (InvivoGen, tlrl-pms) for 5h. To induce pyroptosis, BMDM were stimulated with 10μM nigericin (Invivogen, tlrl-nig) for 1h. To induce necroptosis, BMDM were stimulated with 100ng/ml TNFα (Sigma, T7539-10UG), 500nM birinapant (Apexbio, A4219) and 20μM Z-VAD-FMK (Tocris, 2163) for 16h. QUANTIFICATION AND STATISTICAL ANALYSIS Statistical significance was determined using GraphPad Prism software. Data were first evaluated for normal distribution using the Anderson-Darling test, D’Agostino & Pearson test, Shapiro-Wilk test and Kolmogorov-Smirnov test. If any test reported that N is too small for evaluation or if data were normally distributed, a two-tailed unpaired t test was used. If data did not pass normality tests, Mann-Whitney test was used. Survival curves were evaluated using the Log rank (Mantel-Cox) test. Test results are indicated in the Figures and figure legends. Supplementary Material Supplemental information can be found online at https://doi.org/10.1016/j.celrep.2026.117723.