key: cord-0278588-xoqmbr8r
authors: Ricker, Edd; Manni, Michela; Flores-Castro, Danny; Jenkins, Daniel; Gupta, Sanjay; Rivera-Correa, Juan; Meng, Wenzhao; Rosenfield, Aaron M.; Pannellini, Tania; Chinenov, Yurii; Sculco, Peter K.; Jessberger, Rolf; Luning Prak, Eline T.; Pernis, Alessandra B.
title: Sex-specific differences in the function and differentiation of ABCs mark TLR7-driven immunopathogenesis
date: 2021-01-21
journal: bioRxiv
DOI: 10.1101/2021.01.20.427400
sha: 92f1e5c247bb56d430de2bf4174f810108a77334
doc_id: 278588
cord_uid: xoqmbr8r

Sex differences characterize immune responses to viruses like SARS-CoV2 and autoimmune diseases like SLE. ABCs are an emerging population of CD11c+ T-bet+ B cells critical for antiviral responses and autoimmune disorders. DEF6 and SWAP70, are two homologous molecules whose combined absence in double-knock-out mice (DKOs) leads to a lupus syndrome in females marked by an accumulation of ABCs. Here we demonstrate that DKO ABCs exhibit sex-specific differences in their expansion, upregulation of an ISG signature, and further differentiation. BCR sequencing and fate mapping experiments reveal that DKO ABCs undergo oligoclonal expansion and differentiate into both CD11c+ and CD11c− effector populations with pathogenic and proinflammatory potential. Tlr7 duplication in DKO males overrides the sex-bias and further augments the dissemination and pathogenicity of ABCs resulting in severe pulmonary inflammation and early mortality. Thus, sexual dimorphism shapes the expansion, function, and differentiation of ABCs contributing to the sex-bias that accompanies TLR7-driven immunopathogenesis.

Sex-dependent differences in immune responses have been well documented in viral infections, as recently highlighted by the COVID-19 pandemic, vaccination outcomes, and autoimmune diseases like Systemic Lupus Erythematosus (SLE), a heterogeneous disorder that often includes upregulation of interferon stimulated genes (ISGs) in addition to autoantibody production and multi-organ involvement 1, 2 . Both sex hormones and genes on the Xchromosome have been implicated in this sexual dimorphism 3 . Notably, TLR7, an endosomal TLR critical for responses to viruses like SARS-CoV-2 and lupus pathogenesis, is encoded on the X-chromosome and has recently been shown to escape X-chromosome inactivation leading to greater TLR7 expression in female than male B cells, monocytes, and pDCs 4 .

TLR7 engagement promotes the formation of Age/autoimmune-associated B cells (ABCs), also known as DN2 in humans, a B cell population that preferentially expands with age in female mice 5, 6, 7 . ABCs exhibit a distinctive phenotype and, in addition to classical B cell markers, express the transcription factor T-bet and myeloid markers like CD11c. T-bet and CD11c are often, but not always, co-expressed 8, 9, 10 . ABCs are an important component of antiviral responses and are inappropriately controlled in several viral infections including HIV and SARS-Co-V2 11, 12, 13 . Aberrant expansion and activation of ABCs is also associated with autoimmune pathogenesis, especially SLE 14 . In this disease, ABCs accumulate to a greater extent in African-American patients, rapidly differentiate into plasmablasts/plasma cells (PB/PC), are major producers of autoantibodies, and correlate with disease activity and clinical manifestations 7, 15, 16 . Despite the emerging biological and clinical importance of ABCs, the full spectrum of their function and differentiation capabilities are incompletely understood.

Although T-bet is a well-known marker for ABCs, reliance of ABCs on this transcription factor differs depending on the setting. B-cell T-bet is important for protective flu-specific IgG2c antibodies, but its absence has variably impacted the generation of ABCs and disease parameters in lupus murine models 9, 17, 18, 19 . This is likely due to the presence of additional regulators of autoimmune ABCs such as IRF5 (Interferon Regulatory Factor 5), whose dysregulation promotes ABC accumulation and lupus development 20 . The ability of IRF5 to drive ABC expansion can be restrained by the SWEF proteins, Def6 and SWAP-70, two homologous proteins that also control cytoskeletal reorganization by regulating Rho GTPase signaling and whose combined absence in mice results in a lupus syndrome that primarily affects females 21, 22, 23, 24 . An important role for these molecules in immune responses, inflammation, and autoimmunity is supported not only by murine genetic models but also by human studies. The CORO1A-DEF6 blood transcription module correlates with responses to flu vaccination and malaria 25, 26 . Furthermore, SWAP70 is a susceptibility locus for RA 27 and CVD 28 while DEF6 is a risk factor for human SLE 29, 30 . Mutations in DEF6 moreover result in early-onset autoimmune manifestations, often associated with viral infections, which include autoantibody production and upregulation of an ISG signature 31, 32 .

In this study we have exploited the sex-bias exhibited by mice lacking both SWEF proteins (Double-Knock-out or DKOs) to investigate the impact of sexual dimorphism on the ABC compartment. We demonstrate that ABCs from DKO females and males differ in their ability to expand, upregulate an ISG signature, and further differentiate. BCR sequencing and fate mapping experiments reveal marked oligoclonal expansion and interrelatedness of ABCs with both CD11c + and CD11ceffector populations, which include CD11c + pre-GC B cells and CD11c + PBs. In addition to IRF5, DKO ABCs also require IRF8 but are less dependent on T-bet.

Notably, Tlr7 duplication in DKO males overrides the sex-bias and augments the pathogenicity of ABCs resulting in severe pathology and early mortality. Thus, in autoimmune settings, ABCs can give rise to a heterogenous population of effector cells with distinct pathogenic potentials that are controlled in a sexually dimorphic manner.

Similar to human SLE, the lupus syndrome that develops in DKOs preferentially affects females providing a powerful model to delineate the cellular and molecular mechanisms that underlie sexual dimorphism in autoimmunity. Given the key role of ABCs in lupus, we first assessed whether the sex-bias that accompanies lupus development in DKOs was associated with differences in ABC expansion. Significantly more ABCs accumulated in DKO females than agematched DKO males although DKO males still contained greater numbers of ABCs than WT controls (Fig. 1A) . Furthermore, ABCs sorted from DKO males secreted significantly lower levels of anti-dsDNA IgG2c upon stimulation with a TLR7 agonist, imiquimod, than ABCs from DKO females (Fig. 1B) . Thus, both the accumulation and the function of ABCs in DKOs are controlled in a sex-specific manner.

Tlr7 can be expressed biallelically in female B cells due to incomplete X chromosome inactivation 4 . In line with these findings, ABCs from DKO females expressed higher levels of Tlr7 than ABCs from DKO males (Fig. S1A ). ABC accumulation in DKO females was furthermore dependent on TLR7, as DKO females crossed to Tlr7 -/mice exhibited a profound reduction in ABC accumulation (Fig. 1C) . Tlr7-deficient DKO females also displayed significant decreases in GC B cells, PB/PCs, and TFH cells, and lacked anti-dsDNA IgG2c antibodies (Fig. S1B-E) . Thus, ABC expansion and lupus pathogenesis in DKO females are dependent on Tlr7.

To further assess the importance of Tlr7 in the sex-bias of DKOs, we crossed DKO males to C57BL/6 mice carrying the Y-linked genomic modifier Yaa (termed Yaa-DKOs), in which a portion of the X-chromosome has translocated onto the Y-chromosome resulting in a 2-fold increase in Tlr7 expression in males 33 . Tlr7 duplication in DKO males markedly increased the frequencies and numbers of splenic ABCs reaching levels that were even greater than those observed in DKO females ( Fig. 1D ; S1F). Tlr7 duplication in DKO males also rescued the ability of sorted male ABCs to secrete anti-dsDNA IgG2c antibodies upon stimulation (Fig. 1E ).

Increased ABC accumulation and function in Yaa-DKO males were accompanied by autoantibody production, the classical clinical feature of SLE ( Fig. 1F-G) . Yaa-DKO males also exhibited significantly decreased survival as compared to both DKO males and females (Fig.   1H ). Thus, duplication of Tlr7 in Yaa-DKO males overrides the sex-bias and promotes the development of a severe lupus syndrome in DKO males marked by greatly enhanced accumulation of ABCs and autoantibody responses.

In addition to ABC accumulation, DKO females also exhibit robust GC and PB/PC responses 22 prompting us to examine whether sex-specific differences could also be observed in these compartments. DKO females contained more GL7 + Fas + GC B cells than DKO males, a difference that was again reversed by Tlr7 duplication in Yaa-DKO males ( Fig. 2A) .

Immunofluorescence staining confirmed these findings and revealed that GCs in Yaa-DKO males were smaller and less well-organized than those in DKO females ( Fig. 2B; S2A ). DKO females also demonstrated a greater expansion of PB/PCs than DKO males (Fig. 2C) . Tlr7 duplication in Yaa-DKO males reversed this effect and strongly promoted the accumulation of PB/PCs (Fig. 2C) . No sex-based differences were instead detected in other B cell compartments ( Fig. S2B-D) . Other parameters known to promote spontaneous GC responses such as the ratio between TFH and TFR or the dual production of IFNg and IL-21 were comparable between DKO females and males and only minimally affected by Tlr7 duplication (Fig. 2D -E; S2E-F). Thus, DKOs exhibit a sex-specific accumulation of GC B cells and PB/PCs, which can be regulated in a Tlr7-dependent manner.

Given the sex-bias in the accumulation of ABCs as well as of GC B cells and PB/PCs, we next investigated whether these populations might be related. An analysis of the GC B cell population in DKO females revealed that a subset of these cells expressed CD11c and that the numbers of CD11c + GC B cells were significantly greater in DKO females than DKO males or WT controls ( Fig. 2F; S2G ). CD11c + GC B cells were greatly increased in Yaa-DKO males (Fig. 2F) . We also identified a population of CD11c + PB/PCs that accumulated in DKO females and, to an even greater extent, in Yaa-DKO males ( Fig. 2G; S2H) . Thus, sex differences in lupus development in DKOs are accompanied by the aberrant accumulation of CD11c-expressing B cell effector subsets.

To gain further insights into the effector B cell subsets that differentially expand in the spleens of DKO mice, we next compared their BCR repertoires. To evaluate the clonal landscape, we began by determining the contribution of the top 20 ranked clones to the overall repertoire by computing the D20 index 34 . We observed that the D20 index, or fraction of sequence copies contributing to the sum of the top 20 ranked clones, was lowest in follicular B cells and increased in ABCs, followed by GCB and finally being highest (most expanded) in the PB/PC pool (Fig.   3A ). This order of large clone contribution by B cell subset was preserved in both DKO females and Yaa-DKO males (Fig. S3A) . Furthermore, when one studies the level of resampling of clones between replicate sequencing libraries as an independent measure of clone size, the same trend is preserved, with FoBs having the lowest degree of overlap and PB/PCs having the highest Fig. 3E-F) . Taken together, this initial global repertoire analysis revealed that general repertoire features tended to map by B cell subset rather than by mouse strain, with FoBs having the greatest diversity, smallest clone size, and lowest level of SHM and GCB/PBs having the highest. ABCs instead tended to be intermediate with most of these measures.

To further assess the relationships of ABCs with the other B cell effector populations that aberrantly expand in DKO females and Yaa-DKO males, we sorted and analyzed the various populations from individual mice, stratifying the GCB and PB/PC subsets by CD11c expression.

CD11c is differentially expressed on DN B cells in human lupus and we wondered if similar interconnections existed in the context of the DKO females and Yaa-DKO male models. In particular, CD11c + DN2 cells are transcriptionally and epigenetically poised to become PBs in human SLE 35 . If this also occurs in these models of murine SLE, then CD11c + ABCs, which are phenotypically and functionally overlapping with DN2, may exhibit a higher degree of clonal overlap with PB/PCs than with GCBs or with CD11csubsets. We therefore visualized overlapping clones from DKO female and Yaa-DKO male mice as strings and in Venn Diagrams subsets, including hundreds of clones that were present in all the subsets. To quantify and compare the level of overlap between the different subsets, we compared the Jaccard index, in which each clone is only counted once in each subset, and the Cosine similarity, in which clone size is also taken into account ( Fig. 3I ; S3K-J). Neither measure revealed a consistent pattern of similarity with respect to CD11c status. However, ABCs were consistently most highly associated with PB/PCs.

To further verify the relationships between ABCs and the other effector populations in DKOs, we crossed them with Tbet-zsGreen-T2A-CreER T2 -Rosa26-loxP-STOP-loxP-tdTomato mice (termed ZTCE-DKO), where cells expressing T-bet co-express zsGreen and can be traced with   Tamoxifen-inducible tdTomato expression 36 . These mice allow for the detection of stable T-betexpressing cells (zsGreen + tdTomato + ) and cells that previously but no longer express T-bet (zsGreen -tdTomato + ) 36 . ABCs expressed high levels of both ZsGreen and tdTomato, indicating stable expression of T-bet ( Fig. 3J-K) . CD11c + and CD11c -GC B cells contained both zsGreen + tdTomato + and zsGreen -tdTomato + although the CD11csubset was preferentially zsGreen -tdTomato + suggesting that both populations can originate from T-bet expressing cells but that CD11c + GC B cells include a greater fraction of stable T-bet expressors ( Fig. 3J-K) .

Although no longer expressing ZsGreen, a substantial fraction of CD11c + PB/PCs and CD11c -PB/PCs were tdTomato + , suggesting that both these populations can derive from T-betexpressing B cells (Fig. 3J-K) . In combination with our clonal overlap analyses, these findings suggest that, in this autoimmune setting, ABCs share lineage relationships with both CD11c + and CD11c -GC B cell and PB/PC populations.

While ABCs from female and male DKOs exhibited similar BCR repertoire features, their marked differences in function and differentiation suggested that they might employ distinct molecular programs. To gain insights into these mechanisms, we compared their transcriptome by RNAseq (Fig. 4A ). Gene set enrichment (GSEA) and CPDB pathway analyses revealed that ABCs from DKO females were enriched for pathways related to SLE pathogenesis, interferon (IFN) responses, and TLR and complement cascades ( 37 . Male ABCs were instead enriched for pathways related to RhoGTPase signaling and platelet activation ( Fig. 4D-E; S4C ). Given the profound effects of Tlr7 duplication on the ABCs of DKO males, we next sorted ABCs from Yaa-DKO males and compared their transcriptome to that of ABCs from DKO males (Fig. 4F) . Similar to what was observed in female ABCs, the top pathways upregulated in ABCs from Yaa-DKO males were those related to IFN responses ( Fig. 4G-H; S4D ). ABCs from DKO males were instead enriched for genesets related to hemostasis and platelet activation (S4D-F). Thus, enrichment for an ISG signature differentiates ABCs from DKO females and males and Tlr7 duplication promotes the upregulation of ISGs in male ABCs.

We next employed ATAC-seq to investigate the chromatin landscape of ABCs derived from the different DKOs. We identified at least 85,000 peaks in ABCs from female, male, and Yaa-DKO male mice ( Fig. S4G -H, Table S1 -2). Tlr7 overexpression induced sufficient changes in the chromatin landscape to enable a motif analysis of the differentially accessible regions (DAR) of Yaa-DKO male ABCs versus male ABCs. Peaks upregulated in ABCs from Yaa-DKO males were enriched for motifs known to be bound by IRFs and NFkB family members (Fig. 4I ). Peaks upregulated in ABCs from DKO males were instead enriched for ETS binding sites (Fig. 4J) .

Consistent with the transcriptional profiles and the enrichment in IRF binding motifs, loci that were differentially accessible in ABCs from Yaa-DKO males as compared to DKO males included a number of ISGs like Cxcl11, Ifi44, and Ifitm3 (Fig. 4K) . Thus, differences in ISG expression by ABCs are accompanied by a differential enrichment for IRF binding motifs.

In addition to ABCs, populations of CD11c-expressing GC B cells and PB/PCs also accumulate to a greater extent in DKO females than DKO males suggesting that they also contribute to the sex-bias in disease development. Since the phenotypic and molecular characteristics of these populations are largely unknown we investigated them in greater detail. CD11c + GC B cells shared several phenotypic features with ABCs including high expression of T-bet, Fcrl5, and Cxcr3 (Fig. S5A) . As compared to CD11c -GC B cells, moreover, the transcriptome of sorted CD11c + GC B cells was enriched for ABC genesets and for IFN responses (Fig. 5A-B ; S5B-C).

Despite expressing comparable transcript levels of several classical GC target genes, including Bcl6, Irf8, and Spib, the levels of BCL6 protein were lower in the CD11c + than in the CD11cpopulations ( Fig. 5C; S5D ). In line with the notion that intermediate expression of BCL6 protein in B cells has been associated with a pre-GC B cell state 38, 39 , the CD11c + population contained increased frequencies of BCL6 mid IRF4 + B cells, a profile associated with pre-GC B cells, and was enriched in a pre-GC B cell signature by GSEA (Fig. 5D-E) . BCR signaling, as monitored by the phosphorylation of SYK and LYN, was furthermore significantly higher in CD11c + than CD11c -GL7 + Fas + B cells (Fig. 5F ). The CD11c + subset furthermore were less proliferative than CD11c -GL7 + Fas + B cells ( Fig. 5G-H) . In contrast, the CD11c + population upregulated pathways related to migration and apoptotic cell clearance and expressed high levels of MerTK, a critical efferocytic receptor (Fig. 5I-J) . A greater percentage of CD11c + GL7 + Fas + B cells than CD11c -GL7 + Fas + B cells furthermore could engulf apoptotic thymocytes and this was coupled with upregulation of surface MHC-II expression (Fig. S5E-F) . Thus, the CD11c + GL7 + Fas + B cells that aberrantly expand in DKO females likely represent ABCs that have acquired a pre-GC B cell phenotype and can both engulf and present apoptotic debris, thus potentially augmenting autoreactive responses.

To gain insights into the molecular profiles of the CD11c + PB/PCs that also aberrantly expand in DKO females, we sorted CD11c + and CD11c -PB/PCs from DKO females and compared their transcriptome by RNA-seq (Fig. 6A ). Consistent with their expression of CD11c, CD11c + PB/PCs were enriched for ABC signatures and IFN responses ( Fig. 6B ; S6A-C). CD11c + PB/PCs furthermore expressed higher surface levels of ABC markers like Cxcr3 and Fcrl5 than CD11c -PB/PCs, although, consistent with the fate mapping studies, they expressed only low levels of T-bet (Fig. 6C) . Thus, CD11c + PB/PCs share several transcriptional and phenotypic similarities with ABCs despite downregulating T-bet expression. CD11c + and CD11c -PB/PCs expressed similar transcript levels of Prmd1 and Irf4 although the transcriptional programs normally regulated by Blimp1 and Irf4 in PB/PCs were enriched to a greater extent in CD11c -PB/PCs than in CD11c + PB/PCs ( Fig. 6D; S6D ). CD11c + PB/PCs were more proliferative than CD11c -PB/PCs and expressed higher levels of B220, MHC-II, and Ciita ( Fig. 6E-F) . As compared to CD11c -PB/PCs, CD11c + PB/PCs upregulated pathways related to migration including chemokine receptors like Ccr3 as well as the expression of pro-inflammatory cytokines such as Tnf and Il1b and of inflammasome components like Nlrp3 (Fig. 6H-J) . Taken together these data thus suggest that CD11c + PB/PCs represent a population of PBs with distinctive migratory and pro-inflammatory characteristics.

The finding that a substantial proportion of effector B cell populations were related to ABCs and previously expressed T-bet prompted us to examine the role of T-bet in the accumulation of ABCs and their progeny in DKO females. We thus generated CD23-Cre Tbx21 flox/flox .DKO mice to specifically delete T-bet in B cells from DKO mice. Despite successful T-bet deletion, lack of B-cell T-bet did not significantly decrease the formation of ABCs as assessed by staining with CD11c and CD11b and other ABC markers such as Fcrl5 and Cxcr3 ( Fig. S7A-C) . Lack of B cell T-bet also did not affect the accumulation of total or CD11c + GL7 + Fas + B cells and PB/PCs or the TFH/TFR ratio (Fig. S7D-H) . Lack of B cell T-bet did, however, result in a profound decrease in anti-dsDNA IgG2c without a corresponding increase in anti-dsDNA IgG1 (Fig. S7I) . Thus, in autoimmune-prone DKO females, B cell T-bet is not necessary for ABC generation or differentiation but is specifically required for the production of IgG2c autoantibodies.

Given that global IRF5 deletion profoundly decreased ABC formation in DKO females 20 , we next employed a similar strategy to investigate whether B-cell expression of IRF5 was specifically required for ABC generation and differentiation. Since IRF8 was also identified as a potential upstream regulator of ABCs in DKO females versus males (Fig. S7J) , we also extended this analysis to B cell IRF8. ABC accumulation was significantly decreased in CD23-Cre. Deleting IRF8 in DKO B cells impaired the ability of IRF5 to bind to these regions and lack of IRF5 decreased binding of IRF8 to these sites (Fig. 7K) . Taken together these data support the notion that cooperation between IRF5 and IRF8 promotes the aberrant accumulation, function, and differentiation of ABCs in DKO females.

While SLE preferentially affects females, males with lupus often exhibit a more rapid and severe course. Similarly, the shorter survival of Yaa-DKO males than DKO females suggested a more severe immunopathogenesis. Interestingly, the early mortality and renal damage of Yaa-DKO males was accompanied by greater frequencies of ABCs in the blood and kidneys ( CD11c + and CD11c -PB/PCs as well as activated T cells ( Fig. 8E; S8D ). DKO females also exhibited lung inflammation but, as compared to age-matched Yaa-DKO males, the findings were less severe (Fig. 8D ). Lung infiltrates in DKO females, however, worsened with age and were markedly ameliorated in aged CD23-Cre.Irf8 flox/flox DKO females suggesting a crucial contribution of ABCs to the pulmonary inflammation (Fig. 8F) . Thus, Tlr7-driven expansion of ABCs and their progeny can promote their accumulation in the lungs and the development of severe pulmonary inflammation.

Given that the marked lung inflammation in the setting of TLR7 dysregulation was reminiscent of the pathophysiology not only of SLE but also of severe viral infections such as COVID-19, we conducted additional hematologic and serologic analyses to assess whether other parameters known to be altered in this infection were similarly affected. A peripheral blood count demonstrated lower lymphocyte and platelet counts but increased monocytes in Yaa-DKO males than DKO males (Fig 8G; S8E ). Yaa-DKO males also exhibited elevated levels of serum TNFa and IL4 (Fig. 8H ). We also assessed the production of antiphospholipid antibodies. While all DKOs produced higher levels of anti-phosphatidylserine (pS) IgM antibodies than WT controls, only Yaa-DKO males produced anti-pS IgG, anti-cardiolipin, and anti-MDA-LDL IgG (Fig. 8I ). In line with the known ability of ABCs to produce antiphospholipid antibodies in response to pathogens and mediate hematologic abnormalities 40 , anti-pS IgG and anti-cardiolipin antibodies correlated with ABC and PB/PC frequencies and there was a significant inverse association between ABC frequencies and platelet counts ( Fig. S8F-G) . Thus, Tlr7 duplication in DKO males results in hematologic and serologic abnormalities that can be associated not only with SLE but also with severe viral infections like COVID-19.

To gain insights into the molecular features that might result in a more rapid and severe TLR7induced immunopathogenesis in Yaa-DKO males than in DKO females, we compared the transcriptomes of ABCs and their progeny sorted from Yaa-DKO males with those of the corresponding populations sorted from DKO females. Only minimal differences were observed by GSEA between the ABCs of DKO females and those of Yaa-DKO males (Fig. S8H ). However, a comparison of CD11c + preGC B cells and CD11c + PBs demonstrated that, as compared to DKO females, the populations derived from Yaa-DKO males were enriched for pathways involved in the production of antimicrobial peptides, cytokine interactions and signaling, and transcriptional regulation of granulopoiesis ( Fig. 8J -K; S8I-J). Thus, CD11c + effectors in Yaa-DKO males upregulate pathways that can enhance their pathogenicity and potentially enable their trans-differentiation, as has been recently proposed for plasmablasts in COVID-19 patients 41 .

The mechanisms that underlie the sexual dimorphism observed in responses to infections and vaccinations, and in autoimmune diseases remain incompletely understood. Here, we delineate sex-specific differences in the function and differentiation of ABCs, a subset of B cells that are emerging as critical mediators of antiviral responses and pathogenic players in autoimmunity.

Using a spontaneous model of lupus where disease preferentially develops in females, we demonstrate that female ABCs exhibit a greater ability than male ABCs to accumulate, acquire an ISG signature, and further differentiate into effector populations, which include CD11c + pre-GC B cells and CD11c + PBs. BCR sequencing and fate mapping reveal oligoclonal expansion and relatedness amongst ABCs, GC B cells, and PB/PC populations irrespective of the expression of CD11c. Genetic studies demonstrate a critical role for TLR7, IRF5, and IRF8 in promoting these abnormalities in females. Duplication of Tlr7 in males overrides the sex-bias and triggers severe immunopathology marked by intense lung inflammation and early mortality.

Thus, sex-specific differences permeate several aspects of ABC biology in autoimmune settings suggesting that this compartment may be uniquely endowed to function in a sex-specific manner.

Our studies demonstrate marked differences in the ability of female and male DKO ABCs to express an ISG signature, which has recently been shown to be upregulated in ABCs and PCs from SLE patients, and is one of the best-known features not only of this disease but also of antiviral responses 7, 42, 43, 44 . The ISG signature observed in SLE patients indeed overlaps with that detected upon viral infections and immunizations and an inability to upregulate ISGs has recently been shown to distinguish severe from mild-to-moderate COVID-19 patients 45, 46, 47 . Our epigenetic and genetic analyses furthermore indicate that this system is critically reliant on the dysregulation of IRF5 and IRF8 activity, well-known controllers of ISGs whose variants have long been associated with lupus pathogenesis and whose activity can be a target of viral evasion strategies 48, 49 . Given that, in addition to ISGs, the IRFs can also induce the production of IFNs, sex-specific differences in the control of IRF activity could result in self-reinforcing loops that help confer sexual dimorphism to IFN responses and, while predisposing to autoimmune pathology, could enable females to clear viral pathogens, like SARS-CoV-2, and respond to vaccinations more effectively.

Sex-based differences were also observed in the ability of ABCs to further differentiate into effector subsets. Despite similar TFH responses to DKO males, DKO females exhibited a more robust expansion of GC B cells and PB/PCs, which in addition to classical CD11csubsets, also included CD11c + expressing subsets. BCR sequencing and fate mapping uncovered surprising relationships of ABCs not only with CD11c + pre-GC B cells and CD11c + PBs, but also with CD11c -GC B and CD11c -PC, which exhibited remarkably different transcriptional profiles from the CD11c + B cell populations. The IRFs may again be a crucial component of this dysregulation.

Indeed, the known ability of IRFs to homo/heterodimerize and target ISREs as well as interact with other transactivators like the Ets proteins PU.1 48 may be well-suited to help confer a high degree of heterogeneity to ABCs and their progeny and fine-tune their transcriptional profiles.

The pairing of IRF5 with IRF8 in this compartment may provide ABCs and their CD11c + progeny with a more "innate" quality than traditional CD11c -B cell subsets and facilitate their distinctive combination of innate and adaptive functions. Changes in IRF8 activity, like those mediated by ROCK2 phosphorylation 50 , could instead result in the downregulation of typical ABC transcriptional targets like CD11c and help promote its interaction with other transactivators enabling ABCs to give rise to both CD11c + and CD11ceffector progeny. In support of this notion preliminary studies indicate that TLR7 engagement can inhibit ROCK2 activation thus directly affecting this balance. Thus, in autoimmune settings, ABCs can differentiate into a heterogenous pool of progeny with a broad range of pathogenic effector functions, whose relationships may not be easily surmised from their transcriptional profiles.

VH sequencing demonstrated profound oligoclonal expansion and further confirmed common clonal relationships between ABCs, GC B cells, and PB/PCs, which again could be observed within both CD11c + and CD11ccompartments. Interestingly, different degrees of clonal overlap could be observed between ABCs and different progenies within distinct mice suggesting that while both extrafollicular and GC-like differentiation pathways may be available to these cells, the precise routes employed by ABCs to undergo terminal differentiation can vary depending on the specific inflammatory milieus that they are exposed to. Although the spontaneous GCs observed in this autoimmune setting displayed some atypical features, as evidenced by the finding that these GCs were exquisitely sensitive to the absence of IRF8 unlike those observed upon T-dependent immunization 51 , the ABC-derived GC populations did exhibit higher levels of SHM than ABCs suggesting that these GCs were functional. Notably, several of the VH regions overexpressed in DKO ABCs and their progeny, including VH1 (J558) and VH14 (SM7), have previously been associated with the production of lupus autoAbs and with the expanded ABCs of SLC -/mice, another spontaneous autoimmune model 52 .

The plasticity of ABCs and a limited capability to acquire GC-like as well as PB phenotypes have been previously observed in Ehrlichia and influenza infection models 9, 10, 53 suggesting that, upon encountering a pathogen, these differentiation pathways are available to ABCs, albeit in a restricted manner. One of the consequences of TLR7 overexpression in our autoimmune model furthermore was the dissemination of ABCs in the blood and their accumulation in the lungs, a pattern that was also transiently exhibited by T-bet + B cells shortly after influenza infection 9 .

Strikingly, the lung infiltrates in aged DKO females were markedly ameliorated in mice lacking IRF8 in B cells supporting a key role for ABCs and their progeny in promoting the pulmonary inflammation observed in DKOs. Thus, some of the pathogenesis of SLE may reflect a breakdown in the regulatory mechanisms aimed at restricting the differentiative routes and dissemination of ABCs in time and space. In this regard, the ability of the SWEF proteins to coordinate cytoskeletal organization and IRF function 21 may be particularly important. Indeed, by ensuring the proper positioning, cell-cell interactions, and transient migration of ABCs and by restricting the accessibility to IRF controlled transcriptional programs, these molecules could help ensure that a rapid initial response is coupled with the transient generation of a pool of progeny, which is diverse but limited in size.

Although lower levels of TLR7 responsiveness by male ABCs may protect DKO males from the development of lupus, the transcriptional profile of male ABCs showed enrichment for Rho GTPase signaling pathways. Deregulation of these pathways has been linked to well-known disorders such as hypertension and cardiovascular disease, which also show a sexually dimorphic pattern but preferentially affect males 54, 55 . This raises the intriguing possibility that lower levels of TLR7 engagement by male ABCs could not only result in suboptimal antiviral defenses but also promote a distinct set of pathogenic responses marked by vascular inflammation and thrombosis. Surprisingly, TLR7 duplication in Yaa-DKO males resulted in more profound immunopathogenesis than that observed in DKO females suggesting that females may have evolved mechanisms to contain the overwhelming inflammatory effects that may accompany the greater levels of TLR7 stimulation that they are predisposed to. In particular, the smaller and more disorganized GCs and much greater PB/PC responses observed in Yaa-DKO males than DKO females suggest a lesser ability of Yaa-DKO males to restrain extrafollicular B cell responses.

In this regard, several of our findings are reminiscent of the pathophysiology of COVID-19 where disease outcomes have shown striking age-and sex-dependent differences 1 . Indeed, expansion of T-bet + B cells (including activated naïve and DN) and PBs, which in some cases can exhibit reprogramming potential and lower levels of SWAP-70, have all been observed in severe COVID-19 patients 12, 13, 41, 56 . Interestingly, some of the features known to accompany disease severity in COVID-19 such as lymphocytopenia, thrombocytopenia, production of antiphospholipid antibodies, and a broad array of cytokine responses 57, 58 could be observed in Yaa-DKO males in the absence of any viral infections suggesting that these effects could directly result from unbridled TLR7 stimulation. Whether the reported ability of ABC-like cells to accumulate with age in adipose tissue in an IL-1-dependent manner 59 could provide a ready depot of these cells and lower the threshold of TLR7 stimulation needed by these cells to acquire full pathogenic functions will be an important question to be addressed given the link between obesity and COVID-19 outcomes. Thus, the sex-dependent and heterogeneous properties of the ABC compartment and the need to tightly control their TLR7 responsiveness to ensure effective responses while avoiding pathophysiology could be a critical component of the diverse pathogenic responses to this virus and may need to be taken into account in the design of effective and safe vaccines for SARS-CoV2.

The sex-specific regulation and heterogenous effector potential of the ABC compartment embody essential elements of SLE pathogenesis and could have important therapeutic implications. The remarkable plasticity of ABCs could indeed underlie some of the challenges that have long accompanied the development of therapeutics for SLE and may also need to be taken into account in the design of successful antiviral approaches. Furthermore, the marked sex-differences that we have uncovered in the pathways utilized by female and male ABC subsets suggest that the efficacy of approaches targeting these cells may demonstrate sexdependent differences. Such a scenario may be encountered, for instance, in the case of statins or geranylgeranylation inhibitors that can inhibit the activation of Rho-GTPases and have been shown to have vaccine adjuvant properties 60 . Sex-specific differences in ABC function and differentiation could thus contribute to the well-known sex-bias that underlies several autoimmune diseases and broadly impact responses to pathogens and vaccination.

DEF6-deficient (Def6 tr/tr ) mice were generated by Lexicon Pharmaceuticals, Inc. using a gene trapping strategy as previously described 22 . Swap70-deficient (Swap70 -/-) were generated as previously described 22 . Def6 tr/tr Swap70 -/-(DKO) mice were generated by crossing Def6 tr/tr mice with Swap70 -/mice that had been backcrossed onto C57BL/6 background for >10 generations 22 . and Irf5 f/f DKO mice were previously described 20, 63 . CD23-cre mice were provided by Jayanta Chaudhuri and were previously described 64 . Tbet-zsGreen-T2A-CreER T2 -Rosa26-loxP-STOP-loxP-tdTomato DKO (ZTCE-DKO) mice were generated by crossing DKO mice with Tbet-zsGreen-T2A-CreER T2 provided from Jinfang Zhu 36 and with B6.Cg-Gt(ROSA)26Sor tm14(CAG-tdTomato)Hze /J (#007914, Jackson). ZTCE-DKO mice were treated with Tamoxifen by oral gavage 3d before experiments were conducted. All mice used in the experiments were kept under specific pathogen-free conditions. All the experiments were carried out following institutional guidelines with protocols approved by the Institutional Animal Care and Use Committee of the Hospital for Special Surgery and WCMC/MSKCC.

The following monoclonal antibodies to mouse proteins were used for multi-parameter flow cytometry: B220 (RA3-6B2; 400x), CD4 (RM4-5; 400x), CD45 (HI30; 400x), CD11b (M1/70; 

CD23 + B cells were purified from single cell suspensions of splenocytes with biotinylated anti-CD23 (BD Bioscience) and streptavidin microbeads (Miltenyi) according to manufacturer's instructions. Cells were cultured for 2-3d in RPMI 1640 medium (Corning) supplemented with non-essential amino acids (Corning), 2mM L-Glutamine (Corning), 25mM HEPES (pH 7.2-7.6), and 50µM b-Mercaptoethanol and stimulated with 5µg/mL F(ab')2 anti-mouse IgM (Jackson ImmunoResearch), 5µg/mL purified anti-mouse CD40 (BioXcell), and 50ng/mL IL21 (Peprotech). Thymocyte engulfment assays were performed as previously described 65 . In brief, thymocytes were harvested and isolated from 4-6 week-old WT mice and treated with 50µM

Dexamethasone for 4hr to induce apoptosis. Apoptotic thymocytes were then stained with 1µM CypHer5E (GE Healthcare) for 45min at 37 o C in serum-free Hank's Balanced Sodium Solution (HBSS). Stained thymocytes were co-cultured with splenocytes at a 1:10 splenocyte to thymocyte ratio. Apoptotic thymocytes were removed by washing with cold PBS and splenocytes were assessed for efferocytosis by flow cytometry. MHC-II expression was compared on efferocytic (CypHer5E + ) and non-efferocytic (CypHer5E -) B cells.

Total RNA was isolated using the RNeasy Plus Mini kit (Qiagen). cDNAs were prepared using the iScript cDNA synthesis kit (Biorad). Real-time PCR was performed using the iTaq Universal SYBR Green Supermix (Biorad). Gene expression was calculated using the DDCt method and normalized to Ppia. Primers for Ciita and Tlr7 were from Qiagen. Custom primers were used for 

For anti-dsDNA ELISA, plates were coated with 100µg/mL salmon sperm DNA (Invitrogen; #AM9680) at 37 o C overnight and blocked in 2% BSA in PBS at room temperature for 2hr. For anti-cardiolipin and anti-phosphatidylserine ELISA, Immulon 2HB plates (Thermo) were coated with 75µg/mL of cardiolipin or with 30µg/mL phosphatidylserine dissolved in 100% ethanol overnight. Sera were diluted 1:200 and incubated on coated plates at 25 o C for 2hr. Supernatants from sorted ABCs were used neat after 7d culture with 1µM Imiquimod. Plates were then incubated with HRP-labeled goat anti-mouse IgG or IgG2c Fc antibody for 1hr (eBioscience). Frozen blocks were cut into 9µm section with cryotome and stored at -80 o C. Upon thawing, sections were let dry at room temperature and stained. Spleen sections were stained with B220 (BD; RA3-6B2) and GL7 (BioLegend). Kidney sections were stained with FITC-labeled goat antimouse IgG (Jackson ImmunoResearch). Specimens were captured by Q capture software on a Nikon Eclipse microscope and quantifications were calculated using ImageJ software.

Quality of all RNA and library preparations were evaluated with BioAnalyzer 2100 (Agilent).

Sequencing libraries were sequenced by the Epigenomics Core Facility at Weill Cornell Medicine using a HiSeq2500, 50-bp single-end reads at a depth of ~15-50 million reads per sample. Read quality was assessed and adaptors trimmed using FASTP 68 . Reads were then mapped to the mouse genome (mm10) and reads in exons were counted against Gencode v27 with STAR2.6 Aligner 69 . Differential gene expression analysis was performed in R using edgeR3.24.3. Genes with low expression levels (<2 counts per million in at least one group) were filtered from all downstream analyses. Differential expression was estimated using quasi-likelihood framework.

Benhamini-Hochberg false discovery rate (FDR) procedure was used to correct for multiple testing. Genes with an unadjusted p-value less than 0.01 were considered differentially expressed. Downstream analyses were performed in R using a visualization platform built with Shiny developed by bioinformaticians at the David Z. Rosensweig Genomics Research Center at HSS.

Gene set enrichment analysis was performed using GSEA software (Broad Institute) 70 . Genes were ranked by the difference of log-transformed count per million (cpm) for contrasted conditions. Molecular Signatures DataBase v7.0 (Broad Institute) was used as a source of gene sets with defined functional relevance. Gene sets were also curated from RNA-seq datasets of ABCs from female DKO mice and from the blood of SLE patients 15, 20 and from previously published PB/PC datasets 71, 72 . Gene sets ranging between 15 and 2500 genes were included into the analysis. Nominal p-values were FDR corrected and gene sets with FDR q <0.25 were used to crease GSEA enrichment plots. Analyses of differentially expressed genes were also performed using the online webtool CPDB and upstream regulator analyses were conducted using the Enrichr databases 73, 74, 75 .

The nuclei of sorted ABCs from female DKO, male DKO, or Yaa-DKO mice were prepared by incubation of cells with nuclear preparation buffer (0.30M sucrose, 10mM Tris pH 7.5, 60mM

KCl, 15mM NaCl, 5mM MgCl2, 0.1mM EGTA, 0.1% NP40, 0.15mM spermine, 0.5mM spermidine, and 2mM 6AA) 71 . Libraries were prepared as previously described 76 . Paired-end 50bp sequences were generated from samples with an Illumina HiSeq2500 and, following adapter trimming with FastP, were aligned against mouse genome (mm10) using bowtie2 with -local -q -p options. Peaks were called with MACS2 with --macs2 callpeak -f BAMPE --nomodel --shit -100 --extsize 200 --B --SPMR -g $GENOMESIZE -q 0.01 options. Peak-associated genes were defined based on the closest genes to these genomic regions using RefSeq coordinates of genes. We used the annotatePeaks command from HOMER to calculate ATAC-seq tag densities from different experiments and to create heatmaps of tag densities. Sequencing data were visualized by preparing custom tracks for the UCSC Genome browser. De novo transcription factor motif analysis was performed with motif finder program findMotifsGenome from HOMER package on ATAC-seq peaks. Peak sequences were compared to random genomic fragments of the same size and normalized G+C content to identify motifs enriched in the targeted sequences.

Genomic DNA was extracted from sorted cells using the QIAGEN Gentra DNA purification kit (Qiagen, No. 158689). Primer sequencies and library preparation were described previously 9 .

Samples were amplified in duplicate (2 biological replicates per sample) using 100ng of input DNA per replicate (Table S3) . Sequencing was performed on an Illumina MiSeq instrument in the Human Immunology Core facility at the University of Pennsylvania using a 2x300bp pairedend kit. Sequencing data analysis was also previously described 9 . Sequences were quality controlled with pRESTO 77 . Briefly, paired reads were aligned, sequences with low quality scores were discarded, and base calls with low confidence were masked with "N"s. IgBLAST was then used to align sequences to V-and J-genes in the IMGT database 78 . To group related sequences together into clones, ImmuneDB hierarchically clusters sequences with the same VH gene, same JH gene, same CDR3 length, and 85% identity at the amino acid level within the CDR3 sequence 79 . Clones with consensus CDR3 sequences within 2 nucleotides of each other were further collapsed to account for incorrect gene calls. Sequencing data were submitted to SRA under project number PRJNA663307 in accordance with the MiAIRR standard 80 . Prism v8.4.3 was used for D20, Jaccard index, SHM, CDR3 length, and VH gene usage histogram plots.

Morpheus (Broad Institute) was used for VH heatmap. Venn Diagrams were generated in (http://www.interactivenn.net/). Other calculations were performed as described previously 79 .

p-values were calculated with two-tailed t-tests or ANOVA followed by multi-group comparisons, as indicated in the figure legends. Survival data was tested by Kaplan Meyer analysis with significance determined by the log-rank (Mantel-Cox) test. Correlation data was tested by paired Pearson correlations. p-values of <0.05 were considered significant. Ns: not significant, * p <0.05, ** p <0.01, *** p <0.001, **** p <0.0001. Statistical analysis was performed with Graphpad Prism 8.

The data that support the findings of this study are available from the corresponding author upon request. The RNA-seq and ATAC-seq data will be deposited.

We thank members of the HSS Research Institute for thoughtful discussions and reagents. This 

Ifi204

Interferon Stimulated Genes Cx3cr1  Tnfsf13b  Ccl5  Il7r  Ccl3  Il12a  Tnfrsf19  Il2ra  Il18rap Kdr Amh Cd70 Ccl4 Platelet Activation showing the production of IL21 and IFNg in T cells from the indicated mice. Data pooled from at least 3 mice per genotype and show mean +/-SEM; p-value by 1-way ANOVA followed by Tukey's test for multiple comparisons. * p <0.05, ** p <0.01, *** p <0.001, **** p <0.0001. Figure S8 , related to Figure 8 . 

Representative images of PAS staining in kidneys from the indicated mice. Data shows mean +/-SEM; n>=4 per genotype; p-value by 1-way ANOVA followed by Tukey's test for multiple comparisons. (B) Representative images of IgG deposition in kidneys from the indicated mice. Data shows mean +/-SEM; n>=10 fields across 2-3 mice per genotype

CD4+ PD1 hi CXCR5 + ) in the lungs of the indicated mice. Data show mean +/-SEM; n>=2 per genotype; p-value by 1-way ANOVA followed by Tukey's test for multiple comparisons. (E) Plots showing the numbers of neutrophils and monocytes and the levels of hemoglobin in the blood from the indicated mice. Data show mean +/-SEM; n>=6 per genotype; p-value by 1-way ANOVA followed by Tukey's test for multiple comparisons. (F) Plots showing the correlations between the frequencies of ABCs or PB/PCs in the spleen and serum levels of anti-pS IgG and anti-Cardiolipin IgG antibodies. Data from DKO(F) (red circles), DKO(M) (purple squares), and Yaa-DKO(M) (maroon squares) are shown (n>=5 mice per genotype; p-value by Pearson correlation). (G) Plots showing the correlations between platelet counts and the frequencies of CD11c + CD11b + ABCs in the spleen or blood. (H) Plots showing the top pathways enriched in ABCs from Yaa-DKO(M) mice as compared to those from DKO(F) mice as determined by GSEA

GC) B cells from Yaa-DKO(M) mice. (I) Plots showing the top pathways enriched in CD11c + pre

Dotted line indicates significance threshold at FDR q <0.25. (J) Heatmaps showing the expression of genes enriching the antimicrobial peptide, cytokine signaling, and IL7 signaling genesets in Yaa-DKO(M) CD11c + pre-GC B cells

Considering how biological sex impacts immune responses and COVID-19 outcomes

Taming lupus-a new understanding of pathogenesis is leading to clinical advances

Sexual dimorphism in autoimmunity

TLR7 escapes X chromosome inactivation in immune cells

A B-cell subset uniquely responsive to innate stimuli accumulates in aged mice

Toll-like receptor 7 (TLR7)-driven accumulation of a novel CD11c(+) B-cell population is important for the development of autoimmunity

Distinct Effector B Cells Induced by Unregulated Toll-like Receptor 7 Contribute to Pathogenic Responses in Systemic Lupus Erythematosus

Age-Associated B Cells

The Transcription Factor T-bet Resolves Memory B Cell Subsets with Distinct Tissue Distributions and Antibody Specificities in Mice and Humans

T-Bet(+) IgM Memory Cells Generate Multi-lineage Effector B Cells

Overexpression of T-bet in HIV infection is associated with accumulation of B cells outside germinal centers and poor affinity maturation

Critically ill SARS-CoV-2 patients display lupus-like hallmarks of extrafollicular B cell activation. medRxiv

The Loss of Bcl-6 Expressing T Follicular Helper Cells and the Absence of Germinal Centers in COVID-19

Extrafollicular responses in humans and SLE

IL-21 drives expansion and plasma cell differentiation of autoreactive CD11c(hi)Tbet(+) B cells in SLE

The immune cell landscape in kidneys of patients with lupus nephritis

B cells expressing the transcription factor T-bet drive lupus-like autoimmunity

Generation of functional murine CD11c(+) age-associated B cells in the absence of B cell T-bet expression

CD11c(+) T-bet(+) B Cells Require IL-21 and IFN-gamma from Type 1 T Follicular Helper Cells and Intrinsic Bcl-6 Expression but Develop Normally in the Absence of T-bet

Regulation of age-associated B cells by IRF5 in systemic autoimmunity

Regulation of systemic autoimmunity and CD11c(+) Tbet(+) B cells by SWEF proteins

Dual regulation of IRF4 function in T and B cells is required for the coordination of T-B cell interactions and the prevention of autoimmunity

SWAP-70-like adapter of T cells: a novel Lck-regulated guanine nucleotide exchange factor coordinating actin cytoskeleton reorganization and Ca2+ signaling in T cells

SWAP-70 is a guanine-nucleotide-exchange factor that mediates signalling of membrane ruffling

Induction of Human T-cell and Cytokine Responses Following Vaccination with a Novel Influenza Vaccine

A Molecular Signature in Blood Reveals a Role for p53 in Regulating Malaria-Induced Inflammation

Genome-wide association study in a Korean population identifies six novel susceptibility loci for rheumatoid arthritis

The Integrated Landscape of Biological Candidate Causal Genes in Coronary Artery Disease

High-density genotyping of immune-related loci identifies new SLE risk variants in individuals with Asian ancestry

Transancestral mapping and genetic load in systemic lupus erythematosus

Human DEF6 deficiency underlies an immunodeficiency syndrome with systemic autoimmunity and aberrant CTLA-4 homeostasis

DEF6 deficiency, a mendelian susceptibility to EBV infection, lymphoma, and autoimmunity

Control of toll-like receptor 7 expression is essential to restrict autoimmunity and dendritic cell proliferation

Computational Evaluation of B-Cell Clone Sizes in Bulk Populations

Epigenetic programming underpins B cell dysfunction in human SLE

Dynamic expression of transcription factors T-bet and GATA-3 by regulatory T cells maintains immunotolerance

Modular transcriptional repertoire analyses of adults with systemic lupus erythematosus reveal distinct type I and type II interferon signatures

The Amount of BCL6 in B Cells Shortly after Antigen Engagement Determines Their Representation in Subsequent Germinal Centers

Germinal center B cell development has distinctly regulated stages completed by disengagement from T cell help

Atypical memory B-cells are associated with Plasmodium falciparum anemia through anti-phosphatidylserine antibodies

A single-cell atlas of the peripheral immune response in patients with severe COVID-19

Interferon target-gene expression and epigenomic signatures in health and disease

Interferon-stimulated genes: a complex web of host defenses

Mapping systemic lupus erythematosus heterogeneity at the single-cell level

Cell-specific type I IFN signatures in autoimmunity and viral infection: what makes the difference?

Broad immune activation underlies shared set point signatures for vaccine responsiveness in healthy individuals and disease activity in patients with lupus

Impaired type I interferon activity and inflammatory responses in severe COVID-19 patients

Direct Inhibition of IRF-Dependent Transcriptional Regulatory Mechanisms Associated With Disease

Interferon regulatory factors: critical mediators of human lupus

Serine-threonine kinase ROCK2 regulates germinal center B cell positioning and cholesterol biosynthesis

Transcription factors IRF8 and PU.1 are required for follicular B cell development and BCL6-driven germinal center responses

Age-associated B cells expanded in autoimmune mice are memory cells sharing H-CDR3-selected repertoires

Liver Is a Generative Site for the B Cell Response to Ehrlichia muris

Sex Hormones and Sex Chromosomes Cause Sex Differences in the Development of Cardiovascular Diseases

RhoGTPase in Vascular Disease

Deep immune profiling of COVID-19 patients reveals distinct immunotypes with therapeutic implications

Longitudinal analyses reveal immunological misfiring in severe COVID-19

Autoinflammatory and autoimmune conditions at the crossroad of COVID-19

Aging Induces an Nlrp3 Inflammasome-Dependent Expansion of Adipose B Cells That Impairs Metabolic Homeostasis

The Mevalonate Pathway Is a Druggable Target for Vaccine Adjuvant Discovery

Recognition of single-stranded RNA viruses by Toll-like receptor 7

The Yaa gene model of systemic lupus erythematosus

Regulation of Effector Treg Cells in Murine Lupus

Instructive role of the transcription factor E2A in early B lymphopoiesis and germinal center B cell development

Interpreting an apoptotic corpse as anti-inflammatory involves a chloride sensing pathway

Identification of autoantibody clusters that best predict lupus disease activity using glomerular proteome arrays

Protein array autoantibody profiles for insights into systemic lupus erythematosus and incomplete lupus syndromes

fastp: an ultra-fast all-in-one FASTQ preprocessor

STAR: ultrafast universal RNA-seq aligner

Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles

Multifunctional role of the transcription factor Blimp-1 in coordinating plasma cell differentiation

IRF4 Activity Is Required in Established Plasma Cells to Regulate Gene Transcription and Mitochondrial Homeostasis

Enrichr: interactive and collaborative HTML5 gene list enrichment analysis tool

Enrichr: a comprehensive gene set enrichment analysis web server 2016 update

The ConsensusPathDB interaction database: 2013 update

Transposition of native chromatin for fast and sensitive epigenomic profiling of open chromatin, DNA-binding proteins and nucleosome position

pRESTO: a toolkit for processing high-throughput sequencing raw reads of lymphocyte receptor repertoires

a comprehensive database for human and mouse immunoglobulin and T cell receptor genes

An atlas of B-cell clonal distribution in the human body

Adaptive Immune Receptor Repertoire Community recommendations for sharing immune-repertoire sequencing data