Systemic Autoimmunity and Lymphoproliferation Are Associated with Excess IL-7 and Inhibited by IL-7Rα Blockade

Lupus is characterized by disturbances in lymphocyte homeostasis, as demonstrated by the marked accumulation of activated/memory T cells. Here, we provide evidence that proliferation of the CD8+ precursors for the accumulating CD4–CD8– T cells in MRL-Faslpr lupus-predisposed mice is, in part, driven by commensal antigens. The ensuing lymphadenopathy is associated with increased production of IL-7 due to expansion of fibroblastic reticular cells, the primary source of this cytokine. The excess IL-7 is not, however, consumed by CD4–CD8– T cells due to permanent down-regulation of IL-7Rα (CD127), but instead supports proliferation of autoreactive T cells and progression of autoimmunity. Accordingly, IL-7R blockade reduced T cell activation and autoimmune manifestations even when applied at advanced disease stage. These findings indicate that an imbalance favoring production over consumption of IL-7 may contribute to systemic autoimmunity, and correction of this imbalance may be a novel therapeutic approach in lymphoproliferative and autoimmune syndromes.


Introduction
Lupus, the archetypal systemic autoimmune disease, is characterized by a broad array of T and B cell abnormalities and a plethora of autoantibodies, among which those directed against nucleosomal (DNA, histones) and spliceosomal (small nuclear ribonucleoproteins, snRNP) antigens predominate. Consequent to chronic immune system activation and the effects of various autoantibodies, lupus is characterized by an expedited accumulation of activated T cells as well as lymphopenia, both implying severe disturbances in lymphocyte homeostasis. These disturbances may be attributed to a constellation of factors, including the continuous T cell activation by the ever-present self-antigens, defective activation-induced cell death, and excess of cytokines that promote T cell activation and/or survival. The contribution of these factors may depend on the genetic defect(s) that underlies the pathogenic predisposition. These issues can best be addressed in murine strains that spontaneously develop lupus-like systemic autoimmunity. Among these, the MRL-Fas lpr model exhibits the most evident disturbance in lymphocyte homeostasis due to an early retroviral transposon insertion in the gene encoding the apoptosis-mediating Fas protein, resulting in defective activationinduced cell death and accumulation of activated/memory T cells.
Recent studies have considerably advanced our understanding of the mechanims by which normal T cell homeostasis is controlled, with two cytokines, IL-7 and IL-15, playing primary roles [1,2]. IL-7 is mostly produced by fibroblastic reticular cells (FRCs), a mesenchymal cell population found in the stromal environment of lymphoid organs [3,4]. Binding of IL-7 to the IL-7 receptor (IL-7R), composed of the IL-7Ra chain (CD127) and the common cytokine c chain (cc, CD132), activates several signaling pathways that enhance cellular metabolic functions and survival of naïve, early effector and memory CD4 + and CD8 + T cells, primarily by inducing anti-apoptotic Bcl-2 family members [5]. Similarly, IL-15, primarily expressed by activated monocytes and dendritic cells, binds to IL-15Ra (CD359) on accessory cells and is trans-presented to T cells expressing a functional IL-15R, composed of IL-2/15Rb (CD122) and cc chains [6]. IL-15 promotes the long-term survival of memory CD8 + T cells and, in part, naïve CD8 + and memory CD4 + T cells, but cannot fully compensate for the requirement of IL-7 [1].
In contrast to physiologic conditions in which the availability of IL-7 and IL-15 is rather limited, surplus of these cytokines caused by either reduced consumption (as in lymphopenic mice) or increased production (as in transgenic mice) induces a self-MHC/ peptide-dependent T cell expansion, referred to as ''homeostatic proliferation'', that ceases only when the equilibrium between cytokine levels and T cell numbers is reestablished [1,2]. Although largely polyclonal, homeostatic proliferation appears to favor expansion of T cell clones with higher affinity for self-peptides [1,7,8], as well as acquisition of several surface markers associated with conventional antigen-induced activation [9,10] and even effector functions [11][12][13]. We and others, therefore, proposed that continuous or recurrent lymphopenia and the associated cytokine excess may promote the preferential activation and expansion of self-reactive T cells and autoimmunity in predisposed individuals [14][15][16][17].
In addition to slow-paced homeostatic proliferation occurring under conditions of acute lymphopenia, a second fast-paced form of proliferation termed ''spontaneous proliferation'' has been observed upon T cell transfer in mice that are chronically lymphopenic due to the absence of recombination activating genes (RAG1 or RAG2) or T cell receptor-encoding genes (TCRa or TCRb) [18]. In contrast to homeostatic proliferation, spontaneous proliferation does not require IL-7 and is likely driven by commensal rather than self-antigens since it is significantly reduced in germ-free recipients [19][20][21].
Here, we report that both spontaneous and homeostatic proliferation coexist in the MRL-Fas lpr lupus model. The origin of this homeostasis disturbance could be attributed to self and commensal antigen-induced lymphadenopathy, resulting in expansion of IL-7-producing FRCs and down-regulation of IL-7R in chronically activated T cells. These changes culminate in an excess of IL-7 that sustains the autoimmune process and, thus, blockade of IL-7R signaling significantly decreased disease manifestations in this model.

A Subset of MRL-Fas lpr T Cells Exhibit Phenotypic Markers of Homeostatic Proliferation
In autoimmunity, T cell activation might be induced by either conventional antigen-mediated engagement or excess of T cell trophic cytokines and homeostatic proliferation. These two types of activation can be distinguished by differences in phenotypic cell surface markers, as cognate antigen-activated T cells are CD44 hi CD69 + CD25 + CD62L low , whereas homeostatically-expanded T cells are CD44 hi CD69 -CD25 -CD62L hi [9,22]. As previously described [23], with age and disease progression, MRL-Fas lpr mice exhibited a moderate expansion of lymph node (LN) CD4 + and CD8 + single-positive (SP) T cells, and a much larger expansion of CD4 -CD8double-negative (DN) T cells ( Figure 1A). Interestingly, substantial fractions of SP and DN T cells expressed high levels of CD44 and CD62L, but lacked expression of CD25 and CD69, suggesting that their expansion was caused by excess of T cell-trophic cytokines ( Figure 1B). This interpretation is further supported by the finding that most CD25 + T cells (99.160.1% in young and 96.560.9% in old MRL-Fas lpr mice) were CD4 + Foxp3 + regulatory T cells. However, ,50% of CD4 + and DN T cells, and ,25% of CD8 + T cells in LNs of older mice were CD69 hi , indicating a concurrent conventional self or foreign antigen-driven activation. These findings suggest that both conventional antigenmediated engagement and homeostatic proliferation contribute to T cell activation in MRL-Fas lpr mice.

DN T Cells Lack Receptors for IL-7 and IL-15
Because T cell survival and homeostatic proliferation are dependent on signaling by IL-7 and IL-15, we examined the expression profiles of the receptors for these cytokines. Depending on age, 45-61% of CD4 + T cells and 85-87% of CD8 + T cells in spleen and LNs expressed high levels of IL-7Ra, while 43-75% of CD8 + T cells, but only 2-4% CD4 + T cells, expressed IL-2/15Rb (Figure 2A), frequencies similar to those in normal mice. Strikingly, however, virtually all DN T cells lacked both these receptors, particularly at advanced age ( Figure 2A). Accordingly, in vitro survival of CD4 + and CD8 + T cells was markedly enhanced by IL-7, whereas DN T cells were unaffected by this cytokine ( Figure 2B). Moreover, as previously reported for normal T cells [24], IL-7Ra was down-regulated on both CD4 + and CD8 + T cells cultured in the presence of IL-7 and up-regulated in the absence of this cytokine, but receptor expression remained undetectable on DN T cells in either condition ( Figure 2C), suggesting that IL-7Ra is irreversibly down-regulated on DN T cells. Since most DN T cells are thought to derive from CD8 + precursors [25][26][27][28][29][30][31], we next examined the effects of IL-15 and IL-21, two cytokines known to support survival and proliferation of naïve and memory CD8 + T cells [6,32]. Consistent with IL-2/ 15Rb expression, IL-15 enhanced the in vitro survival of CD8 + , but not DN, T cells. In contrast, IL-21 significantly enhanced survival of DN T cells ( Figure 2D, upper and middle panels) without inducing proliferation ( Figure 2D, lower panel), an effect that correlated with expression of both IL-21R chains (a and cc) by these cells (Figure 2A). These results indicate that DN T cells lose IL-7Ra and IL-2/15Rb, cease using IL-7 and IL-15, and switch to alternative survival resources, such as IL-21.

Commensal Antigens May Contribute to Conversion of CD8 + Precursors to DN T Cells
Persistent down-regulation of IL-7Ra and IL-2/15Rb has been observed on T cells undergoing extensive activation and reaching a functionally defined ''exhausted'' state [33,34]. Although accumulation of DN T cells in MRL-Fas lpr mice is due to defective Fas-mediated apoptosis, the nature of the antigens (self or foreign) that drive activation of the CD8 + precursors for DN cells has not been defined. To differentiate between self and foreign stimuli, experiments were performed with purified, LN-derived T cell subsets transferred into TCRb -/-MRL-Fas lpr recipients. In these chronically immunodeficient hosts, T cells with slow division, defined as ''homeostatic proliferation'', are thought to respond to self-peptide/MHC ligands, whereas those with rapid division, defined as ''spontaneous proliferation'', are thought to proliferate in an IL-7-independent manner [18], likely in response to microbiota-derived ligands [19][20][21]. As observed in chronically T cell-deficient normal background mice, at 1 week post-transfer of syngeneic T cells into TCRb -/-MRL-Fas lpr recipients, ,14% of CD4 + T cells recovered from peripheral LNs (a pool of inguinal, brachial and axillary) had undergone slow-paced homeostatic proliferation (1 to 7 divisions) and ,68% fast-paced spontaneous proliferation (.7 divisions) ( Figure 3A). Remarkably, however, the vast majority of CD8 + T cells (83%) exhibited homeostatic proliferation, whereas DN T cells showed the opposite, i.e. almost exclusively spontaneous proliferation ( Figure 3A). At 2 weeks posttransfer, spontaneous proliferation of CD8 + T cells was more evident, particularly in mesenteric LNs, the primary site of commensal antigen recognition [19,20], and in spleen, but remained significantly lower than spontaneous proliferation of control CD8 + T cells from C57BL/6 mice transferred into chronically immunodeficient (RAG2 -/-) syngeneic recipients ( Figure 3B).
The reduced spontaneous proliferation of MRL-Fas lpr CD8 + T cells was not due to lack of CD4 + T cells or competition with other cell types such as NK and cd T cells, since slow division of CD8 + T cells was maintained when total LN cells were transferred into sublethally-irradiated TCRb -/recipients. Instead, reduced spontaneous proliferation could be due to conversion of CD8 + T cells to DN cells. Indeed, a significant proportion (,51-82%) of the CD8 + T cells that had undergone .7 divisions in LNs of TCRb -/recipients converted to DN T cells ( Figure 3C) and down-regulated both IL-7Ra and IL-2/15Rb ( Figure 3D). In contrast, only 3-9% of CD4 + T cells that had undergone .7 divisions converted to DN T cells. Interestingly, this conversion was not observed in T cells undergoing slow-paced homeostatic proliferation ( Figure 3C and 3D). Hence, the evidence provided by these experiments is compatible with the possibility that recognition of commensal antigens leads to strong proliferation of CD8 + T cells, which is associated with down-regulation of both the CD8 coreceptor and the receptors for major T cell trophic cytokines.

Accumulation of DN T Cells Leads to Excess IL-7
Down-regulation of IL-7Ra should make DN T cells inefficient consumers of trophic cytokines and therefore unable to inhibit homeostatic proliferation of SP T cells. Moreover, reduced consumption by these cells might create an excess of IL-7 sufficient to provoke proliferation of T cells under non-lymphopenic conditions. Indeed, homeostatic proliferation of CFSEstained MRL-Fas lpr SP T cells in LNs of TCRb -/recipients was not inhibited by co-transfer of large numbers (100 6 10 6 ) of purified DN T cells ( Figure 4A). In contrast, efficient inhibition was achieved by co-transfer of total LN cells (50 6 10 6 ) from young MRL-Fas lpr donors, which are mostly composed of naïve T cells with minimal DN T cell expansion ( Figure 4A). More importantly, a significant fraction of SP T cells proliferated when transferred into unmanipulated (non-lymphopenic) older MRL-Fas lpr mice with lymphadenopathy and DN T cell expansion ( Figure 4B), whereas no significant proliferation was detected in non-lymphopenic younger MRL-Fas lpr mice with no lymphadenopathy ( Figure 4C). Interestingly, this proliferation was IL-7dependent and could be inhibited with an anti-IL-7Ra antibody ( Figure 4D). These findings indicate that, in parallel with the accumulation of DN T cells lacking IL-7Ra, older MRL-Fas lpr mice develop excess of IL-7 of sufficient magnitude to induce activation and expansion of potentially autoreactive T cells.

Lymphadenopathy is Associated with Expansion of IL-7-Producing Stromal Cells
Because lymphadenopathy may also lead to the expansion of cells that produce IL-7, we examined IL-7 expression as well as frequency of fibroblastic reticular cells (FRCs), the major producers of this cytokine in secondary lymphoid organs [3]. IL-7 transcripts were significantly increased in the enlarged LNs of older MRL-Fas lpr mice ( Figure 5A), while IL-7Ra transcripts were decreased, resulting in ,7-fold higher ratio of IL-7 to IL-7Ra expression in older compared to younger mice ( Figure 5A). Similarly, higher IL-7 protein levels were detected in LNs from older MRL-Fas lpr mice compared to younger MRL-Fas lpr mice or C57BL/6 controls ( Figure 5B). These findings correlated with an ,7.6-fold increase in the ratio between the numbers of FRCs and IL-7Ra + T cells ( Figure 5C-5D). Furthermore, analysis of MRL-Fas lpr mice with various degrees of lymphadenopathy indicated that FRCs accumulated proportionally to the number of total LN cells, whereas IL-7Ra + T cells accumulated at much slower rate ( Figure 5E). Thus, excess of IL-7 in MRL-Fas lpr mice is caused by a combination of decreased consumption due to receptor downregulation by DN T cells, and increased production due to FRC expansion.

Anti-IL-7Ra Antibody Treatment Reduces Disease in MRL-Fas lpr Mice
Since excess of IL-7 signaling may decrease the activation threshold and provoke proliferation of autoreactive T cells, we examined whether blockade of IL-7R could exert therapeutic effects in the MRL-Fas lpr disease. Initial experiments showed that an anti-IL-7Ra monoclonal antibody (A7R34, rat IgG2a) [35] effectively blocked this receptor in vivo ( Figure 6A), and inhibited IL-7-mediated STAT5 phosphorylation in vitro (mean fluorescence intensity for p-STAT5 in CD4 + T cells: 29565 for non-stimulated cells, 587634 for cells treated with IL-7, and   278610 for cells treated with IL-7 + anti-IL-7Ra). In vivo, this antibody was effective for up to 4 weeks, with progressive declines thereafter due to immune response against the heterologous antibody. Nonetheless, treatment initiated at early disease stages (6 weeks of age) resulted in significant reductions in dermatitis, lymphadenopathy, splenomegaly and total serum IgG2a, and marginal reduction in anti-chromatin IgG2a autoantibodies, compared to PBS-injected controls ( Figure 6B-6E). Treated mice also displayed reduced numbers of CD4 + , CD8 + and DN T cells, as well as immature (T1) and follicular (T2-F0) B cells, whereas marginal zone B cells were unaffected ( Figure 6D). The frequency of peritoneal B-1 cells was also unaffected by this treatment (33.162 A separate group of mice was then similarly treated beginning at 14 weeks of age to assess whether IL-7R blockade could inhibit established disease. Indeed, treatment significantly reduced proteinuria, glomerulonephritis, and lymphocyte infiltrates in the kidneys ( Figure 6F). Impressively, at 24 weeks of age all antibodytreated mice were alive, compared to .50% mortality in the control group ( Figure 6G).

Discussion
We performed studies to define homeostatic characteristics of T cells in the MRL-Fas lpr lupus model, focusing on potential imbalances in the production and consumption of the major T cell prosurvival cytokines. We found that a substantial proportion of T cells displayed phenotypic markers resembling those of T cells undergoing homeostatic proliferation. In addition, the DN T cells that massively accumulate in this Fas-defective lupus model were found to permanently down-regulate the receptors for IL-7 and IL-15, probably due to chronic activation in part mediated by commensal antigens, leading to reduced consumption of these cytokines. Lymphoaccumulation in these mice was also associated with expansion of FRCs, increased IL-7 production, and enhanced IL-7-dependent T cell proliferation. Consequently, treatment with an IL-7Ra-blocking antibody significantly reduced autoimmune disease manifestations in this model (Figure 7).
We and others have hypothesized that, in predisposed genetic backgrounds, recurrent homeostatic T cell proliferation triggered by lymphopenia and excess of T cell-trophic cytokines may be a contributing factor in the pathogenesis of systemic autoimmunity [14][15][16][17]. This possibility stems from the observation that during homeostatic proliferation T cells acquire effector functions and are selected for high affinity to self-antigens [11][12][13]. Indeed, a considerable proportion of T cells in MRL-Fas lpr mice displayed surface markers comparable to those acquired by normal T cells during homeostatic proliferation, and similar phenotypic characteristics have been reported for C57BL/6-Fas lpr T cells [36]. Although this phenotype may also be acquired by conventional antigeninduced effectors during conversion to central memory T cells, this seems unlikely since classical T cell activation is associated with transient expression of CD25 [9,37], yet this marker was primarily detected on Foxp3 + regulatory T cells and, at low levels, in a very small population (,5%) of Foxp3 -CD4 + T cells. Nonetheless, a substantial fraction of CD4 + , CD8 + and DN T cells expressed the CD69 marker, typically upregulated during antigen-induced activation. Thus, both conventional (self) antigens and homeostatic proliferation seem to contribute to T cell activation in this model.
The present study indicates that homeostatic T cell proliferation in these mice is driven by excess of IL-7, since adoptivelytransferred SP T cells proliferated in unmanipulated older MRL-Fas lpr recipients, and this proliferation was significantly blocked by administration of anti-IL-7Ra antibody. Similar observations were reported for C57BL/6-FasL gld mice, which also exhibit expansion of DN T cells with reduced expression of IL-7Ra and IL-2/15Rb [38]. Two major mechanisms emerge from the present study that might explain how IL-7 excess is attained in these mice: increased production due to the expansion of FRCs in the enlarged LNs, and decreased consumption due to persistent down-regulation of IL-7Ra by DN T cells.
FRCs constitute a unique type of stromal cells that enwrap conduits, consisting of collagen fibers and other extracellular matrix proteins, that distribute lymph and small molecules throughout the T cell zone of secondary lymphoid organs [4]. FRCs have been identified as the major source of IL-7 and other factors essential for T cell homeostasis and trafficking [3]. Presumably, lymphoaccumulation of DN T cells induces reorganization of the microarchitecture in the enlarged LNs, leading to increased FRC numbers, as was reported to occur during LN organogenesis, or following certain infections that transiently deplete FRC networks [4].
Regarding the second mechanism, transient down-regulation of IL-7Ra has been previously shown in thymocytes during transition   from double-negative to double-positive stages [5], and in mature T cells following either TCR engagement or signaling by IL-7 and other pro-survival cytokines [24]. This transient IL-7Ra down-regulation is required for efficient IL-7 signaling [39], but has also been proposed as an ''altruistic'' mechanism to ensure availability of this survival-promoting cytokine to unstimulated T cells [5]. Contrastingly, for the MRL-Fas lpr DN T cells, downregulation of IL-7Ra (and IL-2/15Rb) appears irreversible. Irreversible down-regulation of IL-7Ra also occurs in functionally ''exhausted'' T cells that accumulate during persistent viral infections, such as HIV or LCMV [33,34] and, interestingly, high levels of IL-7 can be detected in the serum of AIDS patients [33]. Therefore, it is likely that the DN T cells of MRL-Fas lpr mice have also reached an exhaustion state, which is further suggested by an age-related progressive decline in proliferative capacity of these cells both in vitro and in vivo [31,40]. The molecular mechanism by which IL-7Ra expression is regulated has not been fully explained. Nonetheless, IL-7Ra endocytosis is mediated by clathrin-coated pits followed by Jak3-and proteasome-dependent degradation [39], while resynthesis is regulated by several factors, including the transcriptional repressor growth factor independence-1 (GFI1) and the transcription factor Forkhead box O1 (Foxo1) [24,41].
Absence of IL-7R and IL-15R in DN T cells raises questions on the nature of alternative resources utilized by these cells for survival. We obtained evidence that this function might, in part, be subserved by IL-21, another prosurvival cytokine for naïve and activated/memory T cells [32]. Interestingly, blockade of IL-21 has been reported to reduce the MRL-Fas lpr disease, including lymphadenopathy [42], and this cytokine has been shown to support the survival and function of chronically-activated CD8 + T cells during a persistent viral infection [43].
The nature of the activating stimulus that promotes conversion of CD8 + precursors to DN T cells has not been directly defined. However, based on the present adoptive transfer experiments in chronically immunodeficient recipients, we have hypothesized that commensal antigens may in part be responsible for driving this process. Specifically, we found that a) DN T cells exhibited spontaneous proliferation, which is attributed to microbiota-derived antigens and DC activation by innate stimuli [19][20][21], but not homeostatic proliferation, which is attributed to self-peptide/MHC recognition [1,2]; b) the CD8 + T cell repertoire of MRL-Fas lpr mice was largely depleted of cells able to undergo spontaneous proliferation, presumably because of their conversion to DN cells; and c) the small CD8 + fraction that displayed spontaneous proliferation converted to DN T cells and down-regulated IL-7Ra and IL-2/15Rb. Previous studies indicated that downregulation of CD8 may result from sustained antigen stimulation [44,45], or reduced IL-7 signaling [46], suggesting that downregulation of IL-7Ra may precede that of CD8. Moreover, the small population of DN T cells present in normal mice was reported to display anti-bacteria specificity [47], and MRL-Fas lpr mice bred in a germ-free environment and fed an antigen-free diet had significant reductions in lymphadenopathy and other disease characteristics [48]. The findings as a whole suggest that activation of CD8 + T cells and conversion to DN cells may, in part, be driven by commensal rather than self-ligands, likely due to defects in the anatomic barriers that limit bacterial translocation [49], and/or defects in the suppressive mechanisms that curtail responses to these ligands [1]. Adoptive transfer experiments, similar to those described in the present study, with germ-free MRL-Fas lpr recipients will allow more definitive conclusions with regard to the role of commensal antigens in driving CD8 + T cell activation and conversion to DN cells in this model. Studies with Fas lpr mice lacking MHC class I, MHC class II, CD4, or CD8 suggested that CD8 + T cells and the derivative DN cells are not essential for autoimmune disease development, whereas CD4 + T cells are required [26][27][28][29][30][31]. Nonetheless, in some of these studies, mice lacking CD8 + and DN T cells showed significant reductions in serologic and histologic disease parameters [28][29][30][31]. The present results suggest that DN T cells contribute to disease indirectly by reducing the cellular ''sinks'' for IL-7 and promoting expansion of cells that produce this cytokine. Similar changes may occur in SLE and in autoimmune lymphoproliferative syndrome (ALPS), where there is expansion of DN T cells [50][51][52].
The ultimate effect of IL-7 excess in autoimmunity is a reduction of the activation threshold of autoreactive T cells, which, when superimposed on a predisposed genetic background, leads to disease enhancement. Hence, treatment of MRL-Fas lpr mice with an IL-7Ra-blocking antibody exerted significant disease-decreasing effects even when applied at an advanced disease stage. These effects are compatible with the primary role of IL-7 in survival and proliferation of naïve, early effector, and memory CD4 + T cells. It is possible, however, that additional mechanisms may contribute to the therapeutic effects, including reduction of autoreactive B cells, the development of which is dependent on IL-7 in mice [5,6]. Moreover, IL-7Ra is also a component of the thymic stromal lymphopoietin (TSLP) [53], a cytokine previously shown to promote systemic autoimmunity when overexpressed in transgenic mice [54]. Thus, disease reduction in anti-IL-7Ra antibody-treated mice may be due to inhibition of both IL-7 and TSLP, and additional experiments with specific antibodies or targeted gene deletion may clarify the differential contribution of these cytokines in systemic autoimmunity. Considering that autoimmune disease pathogenesis invariably involves the participation of CD4 + T cells, it is likely that blockade of IL-7Ra signaling will be an efficient treatment in a wide spectrum of autoimmune and inflammatory disorders.

Mice
Animals were housed in a facility approved by the Association for Assessment and Accreditation of Laboratory Animal Care International. All animal experiments and protocols were performed according to the NIH Guide for the Care and Use of Laboratory Animals and approved by The Scripps Research Institute Animal Care Committee (permit numbers 09-0082 and 09-0101). MRL-Fas lpr , C57BL/6 (B6) and C57BL/6.Rag2 -/mice were obtained from The Jackson Laboratory (Bar Harbor, ME), and T cell-deficient TCRb -/-MRL-Fas lpr mice have been previously described [55]. These mice were maintained at the mouse facility of The Scripps Research Institute under specific pathogen-free conditions.

Cell Preparations
Single cell suspensions were prepared from bone marrow, thymus, spleen and LNs (inguinal, axillary, brachial, cervical and, when indicated, mesenteric). To analyze B cell precursors in bone marrow, cells were stained with antibodies to Igk and CD19, and pre-/pro-B cells (CD19 + Igk -), newly formed B cells (CD19 low Igk low ) and recirculating B cells (CD19 hi Igk hi ) were identified by flow cytometry. Similarly, flow cytometry was used to enumerate T cell precursors (DN-I to DN-IV, TCRb low CD69 low , TCRb hi CD69 hi ) upon thymocyte staining with antibodies to CD4, CD8, CD44, CD25, CD69 and TCRb. To analyze FRCs, LN tissues were digested with collagenase IV and DNase I, and the obtained single cell suspensions were stained with antibodies to CD45, gp38 (podoplanin) and CD31, and FRCs were identified by flow cytometry as CD45 -CD31 -gp38 + cells, as described [3]. For some in vivo and in vitro experiments, purified CD4 + , CD8 + and DN T cell subsets were isolated by cell sorting using monoclonal antibodies to mouse CD19 and I-A k (to exclude non-T cells), as well as CD4 and CD8. As assessed by flow cytometry using fluorescent antibodies to TCRb, CD4 and CD8, purity was typically 96-98%.

Real Time Quantitative RT-PCR
LN cells were obtained from young (6 wks) and older (20 wks) MRL-Fas lpr mice and control C57BL/6 mice (12 wks) (n = 3 mice/ group). RNA was extracted from total LN cells (RNeasy Plus Mini kit, Qiagen) and cDNA synthesized (RT First Strand Kit, SABiosciences). PCR was performed in triplicate using 400 ng cDNA, the RT SYBR Green qPCR Master Mix (SABiosciences), primer sets specific for mouse IL-7, IL-7Ra or HPRT1 gene sequences (SABiosciences), and an ABI PRISM 7900HT Sequence Detection System (PE Biosystems). IL-7 and IL-7Ra transcript levels in individual mice were normalized using HPRT1 values and expressed as fold change in comparison to levels in control C57BL/6 (B6) mice.

Immunoblot
LNs were obtained from young (6 wks) and older (18 wks) MRL-Fas lpr mice and control C57BL/6 mice (16 wks) (n = 3 mice/group). LNs were minced with a razor blade and lysed overnight at 4uC in M-Per mammalian extraction buffer containing 150 mM NaCl 2 (Thermo Scientific). Protein concentrations were determined (Bio-Rad Protein Assay) and, for each individual mouse, 75 mg separated on a precast 4-12% gradient polyacrylamide gel (Invitrogen) under reducing conditions. Proteins were transferred to a nitrocellulose membrane using Invitrogen iBlot technology and antibody incubation and washes were performed using an Invotrogen BenchPro 410 card processing station. Briefly, membranes were incubated with rat anti-mouse IL-7 monoclonal antibody (R&D Systems) followed by peroxidase-conjugated goat anti-rat IgG antibody (Jackson ImmunoResearch Labs) and SuperSignal chemiluminescent substrate (Thermo Scientific). After film (ThermoScientific) exposure, nitrocellulose membranes were stripped and re-probed using a rabbit anti-mouse b-actin antibody (Cell Signaling Technology) followed by peroxidase-conjugated donkey antirabbit IgG antibody (BioLegend). Films were examined by densitometry (ImageJ software) and, for each individual mouse, band intensities were normalized by subtracting the background and calculating the ratio IL-7/b-actin.
In Vivo Treatment with Anti-IL-7Ra Antibodies MRL-Fas lpr mice were treated i.p. with 200 mg anti-IL-7Ra rat monoclonal antibody (clone A7R34) three times per week for 1 to 10 weeks. Treatment was started either before (6 weeks of age) or after (14 weeks of age) appearance of disease manifestations (i.e. dermatitis, lymphadenopathy and autoantibody titers).

Serology
Total and anti-chromatin serum IgG subclasses were assessed by ELISA using 96-well plates coated with goat anti-mouse IgG (Jackson ImmunoResearch Laboratories) or mouse chromatin, respectively. Bound antibodies were detected using alkaline phosphatase-conjugated goat antibodies (Caltag Laboratories) to mouse IgG or IgG2a, the main autoantibody subclass in this model. Standard curves were generated using calibrated mouse serum (Accurate Chemical and Scientific Company).

Statistical Analysis
Group comparisons were analyzed by unpaired two-tailed Student's t test. Survival was analyzed by Kaplan-Meier plot and log rank test. p,0.05 was considered significant.