Neutrophils Contribute to Excess Serum BAFF Levels and Promote CD4+ T Cell and B Cell Responses in Lupus-Prone Mice

Despite increased frequencies of neutrophils found in autoimmune diseases such as systemic lupus erythematosus (SLE), how they contribute to disease pathogenesis and the mechanisms that affect the accumulation of neutrophils are poorly understood. The aim of this study was to identify factors in autoantibody-mediated autoimmunity that controls the accumulation of spleen resident neutrophils and to determine whether neutrophils contribute to abnormal B cell responses. Increased levels of the cytokine BAFF have been linked to loss of B cell tolerance in autoimmunity, but the cellular source responsible for excess BAFF is unknown. B cell maturation antigen (BCMA) is a receptor for BAFF and is critical for the survival of bone marrow plasma cells. Paradoxically, BCMA deficiency exacerbates the formation of autoantibody-secreting plasma cells in spleens of lupus-prone mice and the reasons for this effect are not understood. Here we analyzed the phenotype, localization and function of neutrophils in spleens of healthy mice and congenic lupus-prone mice, and compared mice sufficient or deficient in BCMA expression. Neutrophils were found to be significantly increased in frequency and activation status in spleens of lupus-prone mice when BCMA was absent. Furthermore, neutrophils localized within T cell zones and enhanced CD4+ T cell proliferation and IFNγ production through the production of BAFF. Reduced BAFF and IFNγ serum levels, decreased frequencies of IFNγ-producing T cells, germinal center B cells, and autoantibody production after neutrophil depletion indicated the involvement of neutrophils in these autoimmune traits. Thus, we have identified a novel role for BCMA to control excess BAFF production in murine lupus through restraining the accumulation of BAFF-producing neutrophils. Our data suggests that devising therapeutic strategies to reduce neutrophils in autoimmunity may decrease BAFF levels and ameliorate disease.


Introduction
SLE is an autoimmune disorder characterized by a breakdown in B cell tolerance, leading to the generation of autoreactive plasma cells (PCs) that produce pathogenic autoantibodies. The factors that control the abnormal generation and maintenance of autoreactive PCs are poorly understood. Family members belonging to the B cell activating factor of the TNF family (BAFF) cytokine-receptor network have been closely linked to B cell homeostasis and tolerance [1,2,3]. B cell maturation antigen (BCMA) is a receptor expressed on PCs (but not on mature B cells) and is critical for survival of long-lived PCs in the bone marrow [4]. Signaling through BCMA on bone marrow PCs induces the expression of the anti-apoptotic molecule Mcl-1 that is necessary for survival [5]. In contrast, the BAFF receptor BR3 is expressed on mature B cells (but not on PCs) and is critical for their survival in peripheral tissues [6]. BR3 is also expressed on some CD4 + T cells and stimulates proliferation in response to BAFF [7,8,9,10].
Excess circulating BAFF levels in both lupus-prone mice and SLE patients are associated with a loss of B cell tolerance and autoantibody production [11,12,13]. In lupus-prone mice, neutralizing BAFF activity reduces both the frequency of peripheral B cells and activation of T cells, which is sufficient to prevent and treat the disease [14,15]. Yet, the mechanisms that control excess BAFF production in autoimmunity and which BAFF-producing cells contribute to disease pathogenesis are unknown.
The innate and adaptive arms of the immune system are thought to play essential roles in the development of SLE [16]. Neutrophils are a critical component of the innate immune system and the first line of defense against invading pathogens through uptake and destruction of microorganisms. The contribution of neutrophils to SLE pathology has been largely attributed to their ability to produce type I IFNs [16]. In addition, neutrophils undergo cell death by releasing neutrophil extracellular traps (NETs) that provide a source of autoantigens [17,18,19,20]. Neutrophils produce BAFF that is stored intracellular as preformed molecules, which are released when cells are stimulated with IFNc [21]. Neutrophils also express a membrane-anchored form of BAFF that is cleaved to a biologically active soluble form after stimulation [22]. Recently, a subset of human neutrophils has been shown to provide help to splenic B cells through the production of BAFF that enhances antibody production [23]. Thus, neutrophils may be a key cellular source of BAFF in SLE that contribute to abnormal B cell responses.
Given the important role of BCMA in maintaining long-lived PCs, we hypothesized that lupus-prone mice deficient in BCMA would have reduced survival of autoreactive PCs and therefore diminished pathogenic autoantibodies. Paradoxically, loss of BCMA in two different lupus-prone mouse models exacerbated disease through a CD4 + T cell-dependent mechanism that resulted in increased serum BAFF levels and autoantibody production despite reduced survival of bone marrow PCs [14]. Where and how BAFF production is controlled in these murine lupus models is unknown. We report that BCMA plays an important role in controlling the production of BAFF in autoimmunity. We found that BCMA deficiency resulted in an increased frequency of activated BAFF-producing neutrophils in spleens of lupus-prone mice. Interestingly, splenic neutrophils co-localized with CD4 + T cells and the accumulation of splenic neutrophils correlated with increased serum IFNc titers and the frequency of IFNc-producing T cells. Neutrophils from BCMA-deficient lupus-prone mice cultured with wild-type CD4 + T cells significantly promoted T cell proliferation and the production of IFNc compared to neutrophils from control mice. These cellular responses were dependent on BAFF signaling through the BAFF receptor BR3 expressed on CD4 + T cells. Long-term depletion of neutrophils using the mAb, 1A8, significantly reduced the frequency of IFNc-producing CD4 + T cells and the autoimmune phenotype in BCMA-deficient lupusprone animals. These findings extend the role of neutrophils in the pathogenesis of lupus and suggest neutrophils help shape CD4 + T cell responses via BAFF that contribute to the production of pathogenic autoantibodies.

Mice
Tnfrsf17 2/2 , B6.Fas lpr /J, B6.Fas lpr /JTnfrsf17 2/2 , Nba2, and Nba2Tnfrsf17 2/2 mice, fully backcrossed onto the C57BL/6 (B6) strain, were previously described [14,24]. Age-matched WT B6 mice were purchased from NCI. Mice were screened by PCR to determine inheritance of the entire Nba2 locus, the loss of Tnfrsf17, and lpr mutation as previously described [14]. All experiments were performed on individual mice using four-to-six month-old female mice unless indicated. Mice were randomized to experimental conditions and no mice were excluded from data analysis. For in vivo experiments, mice were treated in their home cage between the hours of 2-5 pm. Intra-peritoneal (I.P.) administration of 1A8 mAb was selected to achieve systemic responses.

Ethics statement
Mice were housed in a specific pathogen-free animal facility at the University of Virginia. All animal procedures were conducted in compliance with the National Institutes of Health guidelines and were approved by the Institutional Animal Care and Use Committee of the University of Virginia (Protocol #3506).
Animals received food and water ad libitum and were housed in groups of 5, whenever possible. All efforts were made to minimize suffering and mice were euthanized by carbon dioxide inhalation.

In vivo pristane treatment and neutrophil depletion
Three-month-old female mice were given a single I.P. injection of 0.5 ml PBS (Gibco) or TMPD (2,6,10,14-tetramethylpentadecane), commonly known as pristane (Sigma-Aldrich), as previously described [25]. After one month, animals were euthanized and spleen and serum were prepared for flow cytometric analysis of neutrophils, T-and B-cell populations, and ELISA to measure serum BAFF, IFNc, and autoantibody titers. To deplete neutrophils, B6.Fas lpr /JTnfrsf17 2/2 mice were given I.P. injections of 400 mg rat IgG (Sigma-Aldrich) or 1A8 (BioXCell) every other day for 4 weeks. Following treatment, mice were euthanized and spleen and serum were harvested for analysis.

In vitro assays
CD4 + T cells from spleens of WT mice were isolated to greater than 90% purity (Miltenyi MACS kit). Neutrophils from spleens of lupus-prone mice were isolated to greater than 90% purity, as previously described [26]. All cells were cultured in complete RPMI (Gibco) supplemented with 10% FBS (Gibco) unless otherwise indicated. Cell viability was determined using live/dead AQUA (Invitrogen) according to manufacturer's instructions. For T cell and neutrophil co-cultures, CD4 + T cells were labeled with CellTrace Violet (Life Technologies) and cultured (5610 4 ) alone or with equivalent numbers of splenic neutrophils in the presence of 1 mg/ml anti-CD3 (145-2C11, CedarLane Labs) for 3 days. In some cases, 10 mg/ml of anti-BR3 blocking antibody was added (R&D Systems). T cells were analyzed by flow cytometry for proliferation by CellTrace Violet dilution and culture supernatants were analyzed for IFNc, IL-4, and IL-17 production by ELISA. To measure the frequency of IFNc-producing CD4 + T cells, spleen cells were cultured in the presence of 50 ng/ml PMA (Sigma-Aldrich), 500 ng/ml ionomycin (Sigma-Aldrich) in the presence of GolgiStop (BD Biosciences). After 5 hours of stimulation, CD4 + T cells that produced IFNc were measured by intracellular flow staining.

Flow cytometry
Mouse spleen single-cell suspensions were prepared as previously described [26]. Cells were spun at 800 rpm (130 G) for 8 minutes and maintained on ice for all manipulations. Cells were stained at a concentration of 122610 6 cells/well in a 100 ml volume. Following staining, cells were fixed in 1% formaldehyde

Imagestream
To confirm the morphology of neutrophils, cells were stained with anti-CD11b-FITC and anti-Ly6G-PE-Cy7 as described above. Neutrophils were fixed and permeabilized using the BD Cytofix/Cytoperm Kit according to manufacturer's instructions (Becton Dickinson). DAPI (Life Technologies) was used for nuclear staining. Samples were acquired on an ImageStreamX (Amnis) at a 60X magnification and analyzed using IDEAS software.

Real-time PCR
Cells were suspended in RLT/2-ME buffer and total RNA was isolated using the RNAqueous Micro Kit (Ambion). cDNA was generated using the iScript cDNA synthesis kit (Bio-Rad), according to manufacturer's instructions. PCR reactions were performed using iQSYBR Green Supermix (Bio-Rad) and the following primer sequences (

Statistical analysis
Statistics were determined using Prism software v5.0 (GraphPad Software). All error bars represent the mean 6 standard error of the mean. All p-values were derived from the two-tailed Student's t test, one-way ANOVA, or two-way ANOVA, as appropriate, and stated in the figure legend (*p,0.05, **p,0.01, ***p,0.001).

BAFF levels are influenced by BCMA in murine lupus
We previously demonstrated that lupus-prone mice with a deficiency in BCMA leads to increased serum BAFF levels [14]. To evaluate further the role of BCMA on the expression of BAFF in autoimmunity, we studied the BAFF-producing innate immune cells in spleens of female naïve C57BL/6 (WT) mice and congenic lupus-prone B6.Fas lpr /J mice, and compared mice sufficient or deficient in the gene Tnfrsf17 that encodes BCMA. Twenty-five-week old mice were analyzed due to robust differences in the amount of circulating BAFF among the genotypes. B6.Fas lpr /J and B6.Fas lpr /JTnfrsf17 2/2 mice showed increased serum BAFF levels, with significantly higher BAFF levels in the absence of BCMA compared to WT and Tnfrsf17 2/2 controls, indicating that heightened BAFF levels were dependent on lpr predisposition (Fig. 1A). These results confirm our previous findings, which further demonstrated no difference in APRIL expression among the mouse strains [14]. We previously showed increased frequencies of splenic CD11c + dendritic cells (DC) in B6.Fas lpr / JTnfrsf17 2/2 mice compared to B6.Fas lpr /J mice [14], suggesting that the heightened level of circulating BAFF in these mice may be derived from elevated numbers of BAFF-producing DCs. Interestingly, hematoxylin and eosin staining of spleen sections revealed significant accumulation of polymorphonuclear (PMN) leukocytes with a neutrophil phenotype in B6.Fas lpr /JTnfrsf17 2/2 mice compared to the other strains (Fig. 1B). PMNs in the spleens of B6.Fas lpr /JTnfrsf17 2/2 mice were found in abundance surrounding enlarged follicles with a poorly defined mantle zone, whereas fewer PMNs surrounding follicles were observed in spleens of B6.Fas lpr /J mice. Analysis of spleens from WT and Tnfrsf17 2/2 mice indicated that PMNs were found predominantly within the mantle zone. These data suggest that neutrophils, together with DCs, may contribute to excess BAFF production in autoimmunity and are modulated in cell frequency or BAFF production by BCMA.
To determine whether BCMA deficiency affects the accumulation and BAFF expression of DCs and neutrophils, we examined these cells by immunofluorescent staining for Ly6G and CD11c and quantified BAFF co-localization in spleens of mice. Confocal imaging confirmed increased frequencies of CD11c + DCs in spleens of B6.Fas lpr /JTnfrsf17 2/2 mice compared to the other strains (Fig. 1C, left panel). Despite increased splenic CD11c + DCs in B6.Fas lpr /JTnfrsf17 2/2 mice, we observed no differences in CD11c + cells that co-localized with BAFF expression compared to the other strains. To verify this observation, we quantified the intensity of fluorescently stained BAFF on each CD11c + cell as a measurement of the amount of BAFF protein using the image analysis software, CellProfiler [27]. No significant differences in BAFF intensities co-localizing to CD11c + DCs were measured among the mouse strains (Fig. 1C, graph). We also did not observe any differences in BAFF intensities co-localizing to F4/80 + cells, suggesting that BCMA deficiency does not affect BAFF protein expression in splenic macrophages (data not shown).
Neutrophils are phenotypically defined as Ly6G + CD11b + and have been shown to be potent BAFF producers [23,28,29]. To investigate further whether BCMA deficiency controls BAFF production via neutrophils, we examined Ly6G and BAFF expression in spleen. WT and Tnfrsf17 2/2 mice had few Ly6G + cells present in spleen (Fig. 1C, right panel). In contrast, B6.Fas lpr / J and B6.Fas lpr /JTnfrsf17 2/2 mice consistently had higher frequencies of splenic Ly6G + cells. Quantification of BAFF intensity per Ly6G + cell demonstrated increased BAFF intensities of Ly6G + cells in spleens of Tnfrsf17 2/2 and B6.Fas lpr /J mice, with higher BAFF intensities in B6.Fas lpr /JTnfrsf17 2/2 mice compared to WT controls (Fig. 1C). Taken together, these data indicate that BCMA deficiency in lupus-prone mice controls the abnormal accumulation of splenic CD11c + DCs and neutrophils, and suggests that neutrophils contribute to excess circulating BAFF levels in B6.Fas lpr /JTnfrsf17 2/2 mice. We next evaluated the steady state frequencies of splenic neutrophils in 6-week-old and 25-week-old mice. No differences in the percentages and numbers of Ly6G + CD11b + neutrophils were observed among 6-week-old mice ( Fig. 2A and data not shown). In contrast, B6.Fas lpr /JTnfrsf17 2/2 mice showed significantly higher numbers of splenic neutrophils at 25 weeks of age compared to the other strains. ImageStream analysis confirmed that the Ly6G + CD11b + cells isolated from spleens of B6.Faslpr/J and B6.Fas lpr /JTnfrsf17 2/2 mice were neutrophils, as defined by the classic polymorphonuclear granulocyte morphology (DAPI staining; Fig. 2B). We did not observe any differences in the frequencies of Ly6G + CD11b + neutrophils in the bone marrow of lupus-prone mice compared to control animals, suggesting that the increased frequency of splenic neutrophils in lupus-prone mice resulted from accumulation and not increased neutrophil development (data not shown). Increased frequencies of splenic neutrophils were also found in BCMA deficient congenic Nba2 lupus-prone mice [14,24], which provided independent confirmation that BCMA expression is critical for restraining accumulation of neutrophils in autoimmunity (Fig. 2C). Finally, to determine whether these findings were limited to murine autoimmune susceptibility genes, we examined splenic neutrophils in WT and Tnfrsf17 2/2 mice under environmentally induced lupus conditions after treatment with the naturally occurring hydrocarbon oil pristane. Administration of pristane to naïve wild-type mice results in chronic inflammation and increased production of IFNc that induces a mild lupus-like disease, including loss of B cell tolerance [25]. Tnfrsf17 2/2 mice had significantly higher numbers of splenic neutrophils compared to WT mice following pristane treatment and mock-treated controls (Fig. 2D). These data indicate that BCMA is a key regulator of splenic neutrophil accumulation in autoimmunity.
A previous report has demonstrated that increased BAFF production by neutrophils is linked with an elevated activation state [21]. Following activation, neutrophils upregulate the low affinity IgG Fc receptors (CD16/CD32), the high affinity receptor for IgG (CD64), and LFA-1 (comprised of CD11a and CD18), as well as downregulate CD62L expression [30,31,32]. Therefore, we tested whether the increased frequency of splenic BAFF-producing neutrophils from B6.Fas lpr /JTnfrsf17 2/2 mice was associated with cell activation using these markers. We observed minimal differences in membrane expression levels of CD11a and CD16/ CD32 on neutrophils among the strains (Fig. 2E). In contrast, neutrophils from B6.Fas lpr /JTnfrsf17 2/2 mice had markedly higher expression levels of CD64 and CD18, and lower CD62L compared to neutrophils from control mice. Splenic neutrophils from Nba2;Tnfrsf17 2/2 mice also expressed increased levels of CD64 and CD18, and decreased CD62L expression (data not shown). These data indicate that the absence of BCMA in lupusprone mice contributes to an activated phenotype of spleenresident neutrophils.
We have recently shown in lupus-prone mice that, in addition to plasma cells, BCMA is expressed in a subset of CD4 + T cells and serves as a negative regulator of the germinal center response (Coquery et al, manuscript submitted). To determine if BCMA is expressed in neutrophils and therefore directly involved in controlling the accumulation of splenic neutrophils in lupus-prone mice, we analyzed BCMA expression by measuring mRNA transcript levels. We chose this method since detection of BCMA protein on the cell surface of murine cells is unreliable due to poor reagents [33]. Using a protocol developed by our laboratory to isolate spleen-resident murine neutrophils [26], we found that BCMA was not expressed in neutrophils from B6.Fas lpr /J mice compared to WT, Tnfrsf17 2/2 , and B6.Fas lpr /JTnfrsf17 2/2 mice (Fig. 2F). Plasma cells from WT and B6.Fas lpr /J mice served as a positive control, whereas naïve B cells that do not express BCMA served as negative controls. Additionally, no differences in the viability of purified neutrophils among the mouse strains were observed (Fig. 2G). These data indicate that the accumulation and activation of neutrophils in B6.Fas lpr /JTnfrsf17 2/2 spleens is not due to BCMA deficiency in neutrophils, but results from the loss of BCMA in other cell types.
Neutrophils co-localize with CD4 + T cells in spleens of lupus-prone mice and contribute to T cell response through a BAFF-dependent mechanism Low numbers of BAFF-producing neutrophils have been detected in the perifollicular regions of spleens from healthy humans [23]. To determine further where neutrophils accumulate in spleens of lupus-prone mice under steady state conditions, we stained tissue sections for Ly6G and CD11b, as well as for PNA and IgD to identify B cell follicles (PNA 2 IgD + ) and germinal centers (PNA + IgD 2 ). We detected Ly6G + CD11b + neutrophils in the perifollicular area of spleens from WT mice (Fig. 3A). Neutrophils were also detected outside B cell follicles and germinal centers in spleens of Tnfrsf17 2/2 and B6.Fas lpr /J mice, with even more perifollicular neutrophils that extended outward, beyond the follicular mantle in spleens of B6.Fas lpr /JTnfrsf17 2/2 mice. Interestingly, neutrophils were positioned in the T cell zones of Tnfrsf17 2/2 , B6.Fas lpr /J and B6.Fas lpr /JTnfrsf17 2/2 mice compared to WT mice, as measured by co-localization of Ly6 + and CD4 + cells (Fig. 3B). These data suggest that splenic neutrophils accumulating in the T cell zones of B6.Fas lpr /J mice, and even more in B6.Fas lpr /JTnfrsf17 2/2 mice, may interact with CD4 + T cells.
We hypothesized that the accumulation of splenic neutrophils in T cell zones of lupus-prone mice may influence CD4 + T cell responses through a BAFF-dependent mechanism. A portion of murine splenic mature CD4 + T cells express low levels of the BAFF receptor BR3 [8,34,35]. Consistent with these studies, we found that BR3 is expressed at low levels on splenic CD4 + T cells of all mouse strains compared to the high expression levels of BR3 on B cells (Fig. 3C). BR3 expression levels on CD4 + T cells from lupus-prone mice were slightly higher, as measured by mean fluorescence intensity (MFI) values but were not significant. To directly test whether splenic neutrophils from lupus-prone mice affected CD4 + T cell function, we developed a co-culture system to measure neutrophil-mediated CD4 + T cell responses in vitro. Purified neutrophils from B6.Fas lpr /J and B6.Fas lpr /JTnfrsf17 2/2 mice were co-cultured with mature resting CD4 + T cells from WT mice and T cell proliferation and cytokine production in the culture supernatants were measured. Using this assay, we measured greater CD4 + T cell proliferation from co-cultures with neutrophils from B6.Fas lpr /JTnfrsf17 2/2 mice compared to B6.Fas lpr /J mice, and T cells cultured alone (Fig. 3D). Confirmation that neutrophils from B6.Fas lpr /JTnfrsf17 2/2 mice induced T cell proliferation by producing BAFF that signals through BR3 on CD4 + T cells was determined by adding a BR3 neutralizing mAb to the cultures, which reduced T cell proliferation (Fig. 3D). Neutrophils do not express BR3 and therefore cannot respond to BAFF (Fig. 3E). These data are consistent with our earlier findings that neutrophils from B6.Fas lpr /JTnfrsf17 2/2 mice have a more robust activated phenotype compared to neutrophils from B6.Fas lpr /J mice (Fig. 2D).
Previous work using an OT-II system suggests that neutrophils may induce both Th1 and Th17 differentiation [36]. To determine whether autoimmune-prone neutrophils are driving different effector cell subsets, we co-cultured resting WT CD4 + T cells with neutrophils from B6.Fas lpr /J or B6.Fas lpr /JTnfrsf17 2/2 mice and measured IFNc, IL-4 and IL-17 in culture supernatant to determine the relative differentiation of Th1, Th2, and Th17 responses, respectively. The results demonstrated that T cells cultured alone produced low amounts of IFNc, IL-4 and IL-17 (Fig. 3D). Addition of neutrophils from both B6.Fas lpr /J and B6.Fas lpr /JTnfrsf17 2/2 mice to the cultures increased IFNc production by T cells, with the IFNc levels induced by neutrophils from B6.Fas lpr /JTnfrsf17 2/2 mice substantially higher. Neutrophils from both B6.Fas lpr /J and B6.Fas lpr /JTnfrsf17 2/2 mice promoted minimal IL-4 responses and induced equivalent levels of IL-17 (Fig. 3D). To evaluate the relative contribution of BAFF generated by neutrophils from lupus-prone mice to mediate cytokine production in CD4 + T cells of WT mice, we utilized a BR3 blocking mAb. Blocking BR3 reduced IFNc production by T cells to basal levels, indicating that BAFF produced by splenic neutrophils is responsible for IFNc production in CD4 + T cells in this system. However, blocking BR3 had no impact on IL-4 production. Interestingly, blocking BR3 increased the IL-17 production by T cells when cultured with neutrophils from B6.Fas lpr /JTnfrsf17 2/2 mice but not B6.Fas lpr /J mice; however this increase was not statistically significant (Fig. 3D).
Excess production of IFNc by CD4 + T cells is associated with the pathogenesis of SLE [37]. Analysis of serum IFNc titers in lupus-prone mice demonstrated significantly higher levels from B6.Fas lpr /JTnfrsf17 2/2 mice with age compared to B6.Fas lpr /J mice and control animals, which had little IFNc (Fig. 3F). Moreover, increased numbers of IFNc-producing CD4 + T cells in spleens of B6.Fas lpr /JTnfrsf17 2/2 mice were observed compared to B6.Fas lpr /J mice (Fig. 3G). These data suggest that the elevated amount of BAFF generated by neutrophils in B6.Fas lpr / JTnfrsf17 2/2 mice compared to B6.Fas lpr /J mice signals through BR3 on CD4 + T cells to preferentially drive the differentiation of Th1 cells.
Depletion of neutrophils reduces excess BAFF and IFNc levels and autoimmunity in B6.Fas lpr /JTnfrsf17 2/2 mice The above in vitro findings support the possibility that, like DCs present in spleens of B6.Fas lpr /JTnfrsf17 2/2 mice [14,38], neutrophils may also serve as a key source of BAFF for promoting IFNc production and humoral autoimmunity in vivo. To address this possibility, we examined the impact of 1A8 antibody mediated (anti-Ly6G [39]) in vivo depletion of neutrophils in B6.Fas lpr / JTnfrsf17 2/2 mice on BAFF and IFNc production, germinal center B cell responses, and autoantibody titers after 4 weeks of treatment. Administration of the Ly6G depleting antibody reduced splenic neutrophil frequencies by 80% (Fig. 4A). CD11b + monocytes and dendritic cells expressing intermediate levels of GR1 were unaffected by 1A8 depletion. Neutrophil depletion resulted in a substantial decrease in both serum BAFF and IFNc levels ( Fig. 4B and 4C). We found reduced numbers of CD4 + T cells, including IFNc-producing T cells, in spleens of B6.Fas lpr /JTnfrsf17 2/2 mice compared to isotype antibody-treated mice (Fig. 4D). Since reduced levels of BAFF and CD4 + T cell help would be expected to affect peripheral B cell responses, we evaluated spleen cellular responses. Mice that were neutrophil depleted had significantly decreased numbers of B cells, germinal center B cells, and plasma cells (Fig. 4E). Moreover, we observed reduced serum anti-dsDNA IgG titers and immune complex deposition in kidneys of mice after neutrophil depletion ( Fig. 4F  and 4G). These data indicate that neutrophils drive autoimmunity by promoting abnormal BAFF and IFNc production.

Discussion
Neutrophils play a key role in the innate host defense to infection [40,41]. They eliminate pathogens through phagocytosis, the release of antimicrobial products, and the release of neutrophil extracellular traps (NETs) that trap and destroy invading pathogens. In addition, neutrophils have emerged as important Purified neutrophils from spleens of lupus-prone mice were co-cultured with CD4 + T cells from WT mice at a 1:1 ratio in the presence or absence of BR3 blocking antibody. After 3 days of culture, CD4 + T cell proliferation was measured by CellTrace Violet dilution and cytokine concentrations in culture supernatant were measured by ELISA. Combined data from three mice/genotype. One of three independent experiments with similar results is shown. (E) Gene expression levels of Tnfrsf13C (BR3) in purified neutrophils from mice of the indicated genotype. Samples were normalized to HPRT and the expression in WT animals was set to one. Total B cells are shown as a positive control. Combined data from 3 mice/genotype. (F) Serum IFNc levels from 6-and 25-week-old mice of the indicated genotype were measured by ELISA. Combined data from 5-9 mice/genotype. (G) Total splenocytes from mice of the indicated genotype were cultured in the presence or absence of PMA/ionomycin with GolgiStop. The frequency of IFNc-producing CD4 + T cells was assessed following 5 hour in vitro stimulation. Combined data from 4 mice/genotype at 6 and 25weeks of age. Error bars indicate mean 6 SEM. Statistics determined with a two-way ANOVA using a Bonferroni post test (D), or a one-way ANOVA with a Tukey post test (F, G), and denoted as follows: *p,0.05, **p,0.01, ***p,0.001, ns -not significant. doi:10.1371/journal.pone.0102284.g003 regulators of adaptive immune responses to infection by modulating antibody production, dendritic cell activation, and antimicrobial T cell responses. The abnormal accumulation and function of neutrophils are linked with autoimmune diseases such as SLE, yet their role in disease pathogenesis is unclear. Neutrophils derived from SLE patients have impaired phagocytic capacity, increased activation state and, in some patients, a propensity to release NETs [16,18,42]. Thus, neutrophils in autoimmunity may contribute to the presentation of autoantigens, leading to autoantibody production. Neutrophils are also a source of cytokines, including BAFF, that are important for B cell development, differentiation, and survival [21,22,23,28]. In lupus-prone mice and SLE patients, increased circulating BAFF levels correlate with increased numbers of peripheral B cells, B cell hyperactivity, and autoantibody production [43,44]. Thus, neutrophils may also play immunoregulatory roles in autoimmunity in secondary lymphoid organs by producing excess BAFF that promotes survival of autoreactive B cells and the production of autoantibodies. Here we found that the accumulation, activation, and BAFF expression of neutrophils in lupus-prone mice are increased further in the absence of BCMA. We also show that neutrophils from BCMA deficient lupus-prone mice induce CD4 + T cell proliferation and IFNc production in a BAFF-dependent manner more potently compared to neutrophils from BCMA sufficient lupus-prone mice. These findings correlated with higher BAFF and IFNc serum levels, as well as increased frequencies of IFNc-producing CD4 + T cells, in BCMA deficient lupus-prone mice compared to control animals. Reduced BAFF and IFNc serum levels, decreased frequencies of IFNc-producing T cells, GC B cells, and autoantibody production after neutrophil depletion indicated the involvement of neutrophils in these autoimmune traits.
The relevant BAFF receptor(s) through which increased BAFF levels promote lupus is unknown. Given that signals transduced through BR3 on mature B cells and through BCMA on plasma cells support cell survival, it has been thought that the expression of these receptors on B lineage cells is largely responsible for BAFF-mediated B cell hyperactivity and autoantibody production in SLE. However, both lpr and Nba2 mice deficient in BCMA unexpectedly develop an accelerated lupus-like disease that is associated with aberrant BAFF levels, germinal center formation, and CD4 + T cell-dependent autoantibody production [14], suggesting an important role for BCMA in the context of murine lupus. Our findings indicate that BCMA critically influences BAFF production by controlling the accumulation and activity of spleenresident neutrophils. BAFF signaling through BR3 expressed on T cells has been shown to induce proliferation and cytokine production [7,10,34,35]. Thus, apart from excess BAFF directly affecting the activation and survival of autoreactive B cells, BAFFdependent regulation of CD4 + T cell responses may also contribute to B cell hyperactivity and loss of tolerance in autoimmunity.
Although B6.Fas lpr /JTnfrsf17 2/2 neutrophils induce CD4 + T cell proliferation and IFNc secretion through BAFF, IL-4 and IL-17 production were driven via an alternative mechanism. These findings are consistent with previous reports demonstrating that neutrophils induce IL-17 via an MHC-dependent mechanism [36,45]. Importantly, in autoimmune-prone mice, neutrophils localize within the T cell zone in situ, suggesting our in vitro findings are biologically relevant. Although the role of type I IFNs in SLE has been well established, the significance and impact of the type II IFNs, has only more recently been appreciated. IFNc is elevated in the sera of SLE patients [46,47,48]. Furthermore, multiple groups have shown the benefits of disrupting the IFNc signaling pathway in autoimmune disease. Both MRL/lpr and NZBWF1 mice deficient in IFNc or the IFNc receptor showed significant improvement in disease [49,50,51]. Our findings suggest that neutrophils may contribute to increased IFNc production. Consistent with this notion, depletion of neutrophils in vivo led to a substantial reduction in serum IFNc and IFNc-producing CD4 + T cells, as well as an overall improvement in disease severity.
BCMA expression was not detected in neutrophils. This finding indicates that BCMA deficiency in neutrophils themselves is not involved in the differences of neutrophil accumulation and activation between B6.Fas lpr /J and B6.Fas lpr /JTnfrsf17 2/2 animals. We propose that the overall inflammatory environment in B6.Fas lpr /JTnfrsf17 2/2 mice affects neutrophil accumulation compared to B6.Fas lpr /J mice [14]. How neutrophils are recruited to secondary lymphoid organs is unknown; however, IFNc has been shown to mediate neutrophil recruitment to the lung following tissue injury or infection [52,53]. Thus, further work will be necessary to determine whether a feedback loop exists between BAFF-producing neutrophils and IFNc-producing CD4 + T cells in the recruitment of neutrophils in autoimmune disease. Taken together, our findings demonstrate a role for neutrophils in the pathogenesis of murine lupus.