Quick viewing(Text Mode)

Development of Long-Lived Plasma Cells Autoimmunity by Suppressing

Development of Long-Lived Plasma Cells Autoimmunity by Suppressing

Foxp3+ Regulatory T Cells Control Humoral by Suppressing the Development of Long-Lived Plasma Cells

This information is current as Eunkyeong Jang, Wang Sik Cho, Mi-La Cho, Hyun-Joo of September 30, 2021. Park, Hye-Joa Oh, Sang Mee Kang, Doo-Jin Paik and Jeehee Youn J Immunol 2011; 186:1546-1553; Prepublished online 5 January 2011;

doi: 10.4049/jimmunol.1002942 Downloaded from http://www.jimmunol.org/content/186/3/1546

References This article cites 41 articles, 21 of which you can access for free at: http://www.jimmunol.org/content/186/3/1546.full#ref-list-1 http://www.jimmunol.org/

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists by guest on September 30, 2021 • Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2011 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Foxp3+ Regulatory T Cells Control Humoral Autoimmunity by Suppressing the Development of Long-Lived Plasma Cells

Eunkyeong Jang,* Wang Sik Cho,† Mi-La Cho,‡ Hyun-Joo Park,* Hye-Joa Oh,‡ Sang Mee Kang,* Doo-Jin Paik,† and Jeehee Youn*,†

Foxp3+ regulatory T cells (Tregs) are crucial for maintaining tolerance, but their role in humoral autoimmunity remains unclear. To address this, we combined a model of -dependent arthritis (K/BxN) with Foxp3 mutant scurfy mice to generate Treg-deficient K/BxN mice, referred to as K/BxNsf mice. The disease symptoms of K/BxNsf mice were exacerbated, and this coincided with increases in extrafollicular Th cells, follicular Th cells, and germinal centers. Surprisingly, the K/BxNsf mice exhibited an abnormal accumulation of mature plasma cells in their spleens and a corresponding loss of plasma cells. The plasma cells were unresponsive to the bone marrow homing chemokine CXCL12, despite normal expression of the chemokine receptor CXCR4. Importantly, they were long-lived and less susceptible to the cytotoxic action of cyclophosphamide. Downloaded from They also expressed less FcgRIIb and were less apoptotic in response to autoantigen–autoantibody immune complexes. This suggests that Tregs control susceptibility to cell death induced by engagement of FcgRIIb with immune complexes. Direct cytotoxic effects of Tregs also contribute to the death of plasma cells. Thus, our results reveal that Tregs suppress the emergence of long-lived splenic plasma cells by affecting plasma cell-autonomous mechanisms as well as T cell help, thereby avoiding the persistence of humoral autoimmunity. The Journal of Immunology, 2011, 186: 1546–1553. http://www.jimmunol.org/

ystemic autoimmune diseases, such as systemic lupus cursors that emerge from these reactions are dividing, rapidly erythematosus (SLE) and rheumatoid arthritis, are char- turning over, Ab-secreting plasmablasts, which in turn differenti- S acterized by activation of autoreactive T and B cells (1, 2). ate into nondividing PCs (6). Most PCs either die within 3–4 d in This activation represents a functional loss of self-tolerance and these organs, which apparently provide only few survival niches leads to autoantibody production. trigger com- for the cells, or leave the organs in search of survival niches plement- and FcR-mediated inflammatory responses in the target provided mainly by the bone marrow (BM). After arriving at the tissue, which in turn promote tissue destruction, more extensive BM, PCs terminally differentiate to end-stage Ab-secreting cells loss of self-tolerance, and ultimately symptoms of autoimmunity and attain a long life span of months or even years. Therefore, by guest on September 30, 2021 (3). Therefore, persistent production of pathogenic autoantibodies although a small proportion of long-lived PCs persist in the spleen is central to the chronic, destructive clinical manifestations of sys- and LNs, most long-lived PCs reside in the BM. These PCs are not temic autoimmune diseases. eliminated by treatments targeting B lineage cells, such as irra- The persistence of autoantibodies in autoimmune disorders can diation, prednisone, cyclophosphamide, and anti-CD20 Abs (7– be attributed either to a continuous supply of short-lived plasma- 10). Thus, along with their persistent Ab production, the resistance blasts or to the activity of long-lived plasma cells (PCs). Both of of these PCs to cytostatic treatment contributes to the difficulty of these cell populations develop as a result of Ag-specific, cognate resolving long-lived PC-mediated diseases. Despite the patho- interactions with extrafollicular T helper (TEFH) cells at extra- logical significance of long-lived PCs, little is known about how follicular foci or with follicular T helper (TFH) cells within they develop and persist in vivo in autoimmune states. germinal centers (GCs) formed in peripheral lymphoid organs, Several of the steps generating peripheral autoimmunity can be such as spleen and lymph nodes (LNs) (4, 5). The early PC pre- censored by CD4+Foxp3+ regulatory T cells (Tregs). These cells were originally identified by their capacity to suppress the pro- *Department of Biomedical Sciences, College of Medicine, Hanyang University, liferation and activation of other T cells (11). However, recent Seoul 133-791, Korea; †Department of Anatomy and Cell Biology, College of Med- studies have extended their targets to diverse immune cells in- icine, Hanyang University, Seoul 133-791, Korea; and ‡Research Institute of Medical Science, The Catholic University of Korea, Seoul 137-701, Korea cluding B cells. Tregs have been shown to suppress directly the Received for publication September 2, 2010. Accepted for publication November 30, activation of B cells, in addition to their indirect effect through 2010. suppressing the activity of TFH cells (12, 13). Their suppressive This work was supported by a National Research Foundation grant funded by the activities include perforin/granzyme-dependent cytotoxicity tar- Korean government (MEST) (2008-0058736 by Mid-career Researcher Program and geting B cells as well as T cells (14–16). In agreement with these 2009-0081790). in vitro studies, the titer of autoantibodies in autoimmune subjects Address correspondence and reprint requests to Dr. Jeehee Youn, Department of Anatomy and Cell Biology, College of Medicine, Hanyang University, 17 is inversely correlated with the activity of Tregs. For instance, Haengdang-dong, Sungdong-gu, Seoul 133-791, Korea. E-mail address: jhyoun@ depletion and transfer of Tregs in autoimmune animals lead to in- hanyang.ac.kr creased and decreased autoantibody production, respectively (17, Abbreviations used in this article: 7-AAD, 7-aminoactinomycin D; BM, bone mar- 18). Reversal of the numerical deficit of Tregs in SLE reduces titers row; dLN, joint-draining lymph node; GC, germinal center; GPI, glucose-6- phosphate isomerase; LN, lymph node; PC, plasma cell; PSGL-1, P-selectin glyco- of autoantibodies in vivo (19). However, how Tregs affect the ac- protein ligand 1; SLE, systemic lupus erythematosus; TEFH, extrafollicular helper T; tivity of autoantibody-producing PCs in vivo remains unknown. T , follicular helper T; Treg, . FH We undertook the current study to investigate whether Tregs al- Copyright Ó 2011 by The American Association of Immunologists, Inc. 0022-1767/11/$16.00 ter the physiology of PCs, using a murine autoantibody-mediated www.jimmunol.org/cgi/doi/10.4049/jimmunol.1002942 The Journal of Immunology 1547 disease model named K/BxN. The spontaneous development of purified using TRIzol reagent (Life Technologies). The level of FcgRIIb severe arthritis in this model depends on the production of auto- mRNA was measured by quantitative RT-PCR using an iCycler thermo- to glucose-6-phosphate isomerase (GPI) (20). Pre- cycler (Bio-Rad). Relative amounts of FcgRIIb transcripts were normal- ized to the amounts of b2-microglobulin transcript. The primer sequences viously, we found that despite the loss of self-tolerance, func- used were as follows: FcgRIIb forward, 59-GGA AGA CAC GGT GAC tionally intact Foxp3+ Tregs exist in the K/BxN mice (21). This ACT GA-39 ;FcgRIIb reverse, 59-TGC TCC ATT TGA CAC CGA TA-39; prompted us to consider the in vivo role of Tregs in autoantibody b2-microglobulin forward, 59-TGA CCAGCT TGT ATG CTA TC-39; b2- production. We show in this study that in addition to enhancing the microglobulin reverse, 59-CAG TGT GAG CCA GGA TAT AG-39. development of Th cell subsets, Treg deficiency results in changes BrdU incorporation assays intrinsic to PCs—as evidenced by their aberrant localization, mat- Mice were fed drinking water containing 0.8 mg/ml BrdU (Sigma-Aldrich) uration, and life span—which favor the persistence of humoral for 14–28 d. The water was protected from light and changed every 2 d. . Thus, our results demonstrating PC regulation by Tregs After staining for B220 and CD138, splenocytes and dLN cells extracted provide novel insight into how Tregs limit humoral memory. from the mice were fixed, permeabilized, treated with 50 Kunitz units DNase (Sigma-Aldrich), and then stained with FITC-conjugated anti-BrdU mAb according to the manufacturer’s instructions (BD Bioscience). BrdU Materials and Methods incorporation in B220loCD138+ cells was determined by FACS analysis. Mice Cell culture A cross between KRN TCR-transgenic mice on a C57BL/6 background To determine whether Tregs are cytotoxic to PCs, coculture experiments (K/B) and NOD mice generated arthritic transgenic progeny (K/BxN) and were carried out. CD4+CD25+ and B220+ cells from the spleens of C57BL/ nontransgenic littermates (BxN) (20). Female C57BL/6 mice bearing the 6 mice were purified by MACS (Miltenyi Biotec). Each cell fraction was Downloaded from scurfy (sf) allele (The Jackson Laboratory) were back-crossed to NOD for more than 97% pure. CD4+CD25+ cells were preactivated with 5 mg/ml more than six generations to generate sf carrier female mice on an NOD anti-CD3 mAb, 1 mg/ml anti-CD28 mAb, and 100 U/ml IL-2 (all from BD background (NODsf). NODsf female mice were crossed with K/B to Bioscience) for 2 d. B220+ cells were stimulated with 10 mg/ml LPS for 2 generate male K/BxN mice bearing the sf allele (referred to as K/BxNsf d, washed, and cocultured with equal numbers of preactivated CD4+CD25+ mice hereafter). All mice were maintained in a specific pathogen-free cells in the presence of 1 mg/ml anti-CD3 mAb, 1 mg/ml anti-CD28 mAb, barrier facility at Hanyang University. The study was approved by the 100 U/ml IL-2, and 1 mg/ml LPS, followed by FACS analyses. To de- institutional animal care and use committee. termine whether Tregs affect FcgRIIb expression in PCs, naive Tregs purified from BxN mice were cocultured for 24 h with PCs that had been http://www.jimmunol.org/ FACS analyses formed in vitro as above or purified from K/BxN mice by MACS. The cells Single-cell suspensions of spleen, joint-draining lymph node (dLN; a pool were then analyzed by FACS. To determine whether PCs are cytotoxic for of axillary, inguinal, and popliteal LNs), and BM cells were surface or immune complexes, GPI immune complexes were made by adding 10 mg/ ml GPI to serum from 8-wk-old K/BxN mice at a ratio of 1:1 (v:v). Normal intracellularly stained as previously described (22). The following mAbs lo + and reagents were purchased from BD Biosciences: anti-B220 (RA3-6B2), serum from BxN mice served as a negative control. B220 CD138 PCs anti-CXCR5 (2G8), anti-CD138 (281-2), anti-CD62L (MEL-14), anti–P- purified by MACS were incubated with the experimental and control sera selectin glycoprotein ligand 1 (anti–PSGL-1; 2PH1), anti-CXCR4 (2B11/ for 9 h, stained with allophycocyanin-conjugated annexin V and 7-AAD, CXCR4), anti–VLA-4 (R1-2), anti-FcgRII/III (2.4G2), 7-aminoacti- and assayed by FACS. nomycin D (7-AAD), and annexin V. Anti-ICOS mAb (C398.4A) was pur- chased from eBioscience. The mAbs were used as FITC, PE, PerCP, Results by guest on September 30, 2021 allophycocyanin, or allophycocyanin–Cy7 conjugates. Disease exacerbation in K/BxNsf mice is associated with Fluorescence microscopy enhanced activation of both extrafollicular and follicular Mouse spleens were embedded in OCT compound and snap-frozen in liquid pathways of Ab responses nitrogen. Cryosections (7mm thick) werefixedwith acetone,blockedwith 10% goat serum, and stained with appropriate combination of mAbs. The following To investigate the in vivo role of Tregs in humoral autoimmu- mAbs were purchased from BD Biosciences: anti-CD4–FITC, anti–GL-7– nity, we generated Treg-deficient K/BxNsf mice by congenically FITC, and anti-CD138–PE. The following mAbs and reagents were purchased combining the K/BxN mice with Foxp3 mutant scurfy mice. We from eBioscience: anti-CD4–PerCP–Cy5.5, anti-Foxp3–FITC, anti-Foxp3– found that K/BxNsf mice exhibited earlier onset and more ag- PE, anti-CD4–allophycocyanin, anti-B220–allophycocyanin, biotinylated anti- CD4, biotinylated anti–GL-7, and streptavidin–Cy-3. Fluorescence images gressive progression of arthritis than their K/BxN littermates, con- were acquired using an LSM 510 confocal microscope (Zeiss). sistent with previous findings (24). In particular, titers of serum Abs against GPI were detectable from 4 wk and peaked at ∼8wk ELISPOT assays in K/BxNsf mice, which were 10–14 d earlier than in K/BxN mice GPI-specific IgG1-secreting cells in spleen and BM were enumerated by (data not shown). ELISPOT assays as described (22). Aliquots of serial dilutions of cell T cell-dependent Ab responses can take place in either extra- suspensions from spleens and BM were added to GPI-coated plates and follicular or follicular areas. These two independent responses are incubated overnight at 37˚C, followed by sequential incubation with bio- + hi tinylated anti-mouse IgG1 mAb (BD Bioscience) and streptavidin–perox- governed by two distinct populations of Th cells, CD4 CD44 lo lo + 2 + + idase. After adding AEC substrate (BD Bioscience), numbers of spots of CD62L PSGL-1 TEFH cells and CD4 CD25 CXCR5 ICOS TFH GPI-specific IgG1 Ab-secreting cells were counted with a video-based auto- cells (4, 5). Both populations emerged in the spleen and dLNs of matic ELISPOT reader. 4-wk-old K/BxN mice and gradually expanded with age, sug- Cell migration assays gesting that both extrafollicular and follicular pathways of Ab responses are involved in the pathology of K/BxN mice (Fig. 1A, PC chemotaxis was quantified by ex vivo cell migration assays as described (23). Splenocytes (3 3 106) were added to the inserts of 5-mm microporous 1B). K/BxNsf mice contained much larger numbers of TEFH and Transwell plates (Corning) and incubated in RPMI 1640 medium con- TFH cells in their spleens and dLNs than those of K/BxN mice. taining 10% FBS for 1 h at 37˚C. After adding 100 ng/ml mouse CXCL12 These data suggest that Tregs limit the development of TEFH and (R&D Systems) to the wells, the cells were cultured for an additional 3 h. T cells in vivo, ultimately regulating B cell activation. Cells remaining in the inserts and those that had migrated to the wells were FH collected and assayed by FACS. In the central, follicular -rich region of the follicle, TFH cells help Ag-engaged B cells to form GCs. Because post-GC RT-PCR B cells give rise to -switched, high-affinity clones and B220+ and B220loCD138+ cells were sorted by MACS (Miltenyi Biotec). leave the GC as PCs and memory B cells, this GC-dependent path- The purity of resulting cell fractions was more than 97%. Total RNA was way of B cell differentiation is crucial for long-term humoral 1548 ROLE OF REGULATORY T CELLS IN PLASMA CELL

FIGURE 1. Enhanced development of

TEFH and TFH cells and GCs in K/BxNsf mice. A, Splenocytes (Sp) and dLN cells from BxN, K/BxN, and K/BxNsf mice were stained for CD4, CD44, CD62L, and PSGL-1. FACS profiles gated on CD4+CD44hi cells are shown. A representative result for 8-wk-old mice for the percentage of CD62LloPSGL-1lo cells among CD4+CD44hi cells is shown (top pan- els). Graphs are means 6 SEM of CD4+ CD44hiCD62LloPSGL-1lo cell numbers in spleen and dLNs (bottom panels). B, Spleno- cytes and dLN cells were stained for CD4, CD25, CXCR5, and ICOS. FACS profiles gated on CD4+CD252 cells are shown. A representative result for 8-wk-old mice for the percentage of CXCR5+ICOS+ cells among CD4+CD252 cells is shown (top panels). Graphs are means 6 SEM of CD4+CD252 CXCR5+ICOS+ cell numbers in spleen and dLNs (bottom panels). *p , 0.05; **p , Downloaded from 0.01 (Student t test). C, Immunofluorescence staining of spleen cryosections with anti- CD4–FITC (green), anti-B220–allophycocya- nin (blue), and anti–GL-7–Cy3 (red). D, Immunofluorescence staining of spleen cry- osections with anti-CD4–Cy3 (red), anti- http://www.jimmunol.org/ B220–allophycocyanin (blue), and anti–GL-7– FITC (green). E, Immunofluorescence staining of spleen cryosections with anti-CD4–allo- phycocyanin (blue), anti–GL-7–FITC (green), and anti-Foxp3–PE (red). Foxp3+ cells in the GC are indicated by the arrows. responses (25). We found that K/BxNsf mice developed GCs ∼2 extrafollicular area of the spleen and dLNs with a concordant loss wk earlier and in greater numbers than K/BxN mice (Fig. 1C,1D). of BM PCs. This suggests that the splenic accumulation of PCs is by guest on September 30, 2021 Along with data demonstrating the presence of Foxp3+ Tregs due to failure of BM homing, or vice versa. within the GC (Fig. 1E), these results provide robust evidence Splenic PCs from K/BxNsf mice are unresponsive to CXCL12, that Tregs affect the kinetics of GC formation and quantity of despite expressing its receptor normally GCs in vivo. PC migration to the BM is largely controlled by an interaction between CXCL12 and its receptor CXCR4 (26). After activation Treg deficiency results in aberrant accumulation of mature PCs within the GC, the expression of CXCR4 is enhanced on the in the spleen and LNs of K/BxN mice surface of post-GC B cells. CXCL12 is produced by BM stromal We tested whether the loss of Tregs affects the behavior of Ab- cells and guides the recruitment of CXCR4+ cells to the BM. To secreting PCs. PCs formed in the K/BxN mice resembled those address whether splenic PCs from K/BxNsf mice were responsive developing upon immunization with foreign protein Ags in several to CXCL12-induced migration, we carried out ex vivo chemotaxis ways. First, spleen and dLNs from K/BxN mice contained a small assays. The vast majority of K/BxNsf PCs did not migrate toward + but significantly larger fraction of CD138 PCs than that from CXCL12, whereas a large fraction of K/BxN PCs did (Fig. 3A). + hi normal BxN mice and consisted predominately of CD138 B220 Thus, PCs developing in a milieu lacking Tregs seem to lack the + lo early PCs (plasmablasts) rather than CD138 B220 mature PCs capacity to respond to CXCL12, and this affects their capacity for (Fig. 2A,2B). Second, a substantial fraction of autoantibody-se- BM homing. creting PCs resided in the BM (Fig. 2C). In contrast, K/BxNsf To determine whether this migratory defect resulted from re- mice contained a substantially higher fraction and number of duced expression of CXCL12 receptor, we measured the level of lo + B220 CD138 mature PCs in their spleens and dLNs, despite cell surface CXCR4 on PCs by FACS analysis. Early (B220hi) and + a significantly lower number of total B220 cells (Fig. 2A,2B). late (B220lo) PCs from K/BxNsf mice expressed even higher This phenomenon was not seen in older K/BxN mice (10 wk old) levels of CXCR4 than those from K/BxN mice (Fig. 3B). In ad- with full-blown arthritis, suggesting it represents a qualitative dition, VLA-4, an adhesion molecule that is involved in the re- change not simply a matter of kinetics. In ELISPOT assays mea- tention of PCs in the BM, was also enhanced on the surfaces of suring the frequency of autoantibody-secreting PCs, GPI-specific PCs from K/BxNsf mice. Thus, the defective responsiveness of ∼ IgG1 Ab-secreting cells were 5-fold more numerous in the PCs to CXCL12 in K/BxNsf mice is not due to lower expression spleen and ∼3-fold less numerous in the BM of K/BxNsf mice than of its receptor on the PC. that in their K/BxN counterparts (Fig. 2C). Immunofluorescence staining further confirmed the abundance of PCs in the spleen, Splenic PCs from K/BxNsf mice live longer and are less especially in the extrafollicular space of splenic red pulp (Fig. susceptible to cytostatic agents than those from K/BxN mice 2D). Taken together, our results demonstrate that Treg defi- During normal T cell-dependent responses, most PCs residing in ciency results in abnormal accumulation of mature PCs in the the spleen and LNs die within 3–4 d because these organs provide The Journal of Immunology 1549 Downloaded from http://www.jimmunol.org/ by guest on September 30, 2021

FIGURE 2. Aberrant accumulation of mature PCs in the spleen and dLNs of K/BxNsf mice. A, Splenocytes (Sp) and dLN cells were stained with mAbs to B220 and CD138 and assayed by FACS. Representative FACS profiles gated on live are shown. B, Numbers of total B220+ B cells, B220hi CD138+ plasmablasts (PB), and B220loCD138+ PCs from spleen. C, Splenocytes and BM cells from 8-wk-old K/BxN and K/BxNsf mice were assayed by

ELISPOT to determine numbers of GPI-specific IgG1 Ab-secreting cells. *p , 0.05; **p , 0.01; ***p , 0.001 (Student t test). D, Immunofluorescence staining of spleen cryosections with anti–CD4-PerCP–Cy5.5 (white), anti-B220–allophycocyanin (blue), anti–GL-7–FITC (green), and anti-CD138–PE (red). The data are representative of three independent experiments. F, follicle. only few survival niches for the cells. To determine whether Tregs BxNsf mice had fallen by only ∼50%, indicating that K/BxNsf can affect the life span of PCs in addition to their migratory be- PCs are less susceptible to the cytostatic effect of cyclophospha- havior, K/BxNsf and their counterpart littermates were fed BrdU mide than are K/BxN PCs (Fig. 4B). Thus, these results demon- for 14–28 d, and incorporation was assayed by FACS. Surpris- strate that Tregs act to reduce the development of cytotoxic drug- ingly, more than 95% of viable PCs from the spleens and dLNs of resistant, long-lived splenic PCs in vivo. K/BxNsf mice survived for at least 28 d without cell division, whereas 80–90% of those from K/BxN mice divided and were Tregs control the survival of PCs via direct and short-lived (Fig. 4A). FcgRIIb-mediated death mechanisms A hallmark of long-lived PCs is their resistance to cytostatic It is well established that Tregs elicit the death of T and B cells in treatment with immunosuppressants, such as cyclophosphamide a perforin/granzyme-dependent manner (14–16, 27), but the ex- (9). To determine whether splenic PCs from K/BxNsf mice were istence of such an effect on PCs has not been explored. Our data resistant to the cytotoxic effect of cyclophosphamide, K/BxN and showing Tregs in contact with splenic PCs in the extrafollicular K/BxNsf mice were treated with cyclophosphamide for a week, area (Fig. 5A) suggest direct action of Tregs on PCs. Indeed, we and viable PCs were assayed by FACS. Whereas the fraction of found that CD138+ PCs formed in vitro underwent increased cell PCs from K/BxN mice had decreased by ∼80%, that from K/ death, as evidenced by increased annexin V+7-AAD+ cell fraction 1550 ROLE OF REGULATORY T CELLS IN PLASMA CELL HOMEOSTASIS Downloaded from

FIGURE 4. Splenic PCs from K/BxNsf mice are long-lived and more resistant to the cytostatic effect of cyclophosphamide. A, K/BxN and K/ BxNsf mice of the indicated age were fed BrdU for 2–3 wk, and labeled cells were assayed by FACS. The FACS profiles gated on viable B220lo + +

CD138 PCs are shown. The numbers indicate the percentage of BrdU http://www.jimmunol.org/ lo + FIGURE 3. Splenic PCs from K/BxNsf mice are unresponsive to cells among B220 CD138 PCs. B, Four- to five-week-old K/BxN and K/ CXCL12. A, Spleen cells from 5- to 8-wk-old BxN, K/BxN, and K/BxNsf BxNsf mice were i.p. injected with cyclophosphamide (CyP) three times mice were subjected to ex vivo migration assays. Cells loaded in the upper a week with the indicated total doses. A week later, splenocytes were lo + Transwells (input) were inserted into lower wells containing medium or analyzed by FACS to measure viable B220 CD138 PCs. Numbers in- lo + 100 ng/ml CXCL12. After 3 h, the cells in the lower wells were collected dicate the percentage of B220 CD138 cells among viable lymphocytes. and assayed by FACS. B, Spleen cells from 5- to 8-wk-old BxN, K/BxN, One representative result of three independent experiments is shown. and K/BxNsf mice were stained with mAbs to B220, CD138, CXCR4, and VLA4. FACS profiles gated on either B220hiCD138+ plasmablasts or onstrating that the K/BxNsf PCs were not susceptible to the im- B220loCD138+ PCs are shown. The data are representative of three in- by guest on September 30, 2021 dependent experiments. mune complex-mediated that is dependent on the abun- dance of FcgRIIb. The reduced FcgRIIb expression on K/BxNsf PCs could mean and by decreased viable PC number when cocultured with Tregs that Tregs directly upregulate FcgRIIb expression on PCs. To test in wells but not when separated from the Tregs in Transwells this, we cocultured B220loCD138+ PCs purified from K/BxN (Fig. 5B,5C). Thus, these results demonstrate that Treg-mediated mice with syngeneic Tregs and found that FcgRIIb levels on the killing of PCs involves a contact-dependent mechanism. PCs were not influenced by the presence of Tregs (Fig. 6D). We Why are splenic PCs from K/BxNsf mice long-lived in an obtained the same result with PCs generated in vitro (data not -rich milieu? This question arises because K/ shown). Therefore, Tregs do not seem to directly signal PCs to BxNsf mice contain large numbers of immune complexes (28), induce FcgRIIb expression. and PCs readily undergo apoptotic death when FcgRIIb molecules expressed on them are cross-linked with immune complexes (29). Discussion To answer this question, we first measured the level of FcgRIIb on Although Treg deficiency has been shown previously to exacerbate the PCs and found that the surface level of FcgRIIb was signifi- the disease symptoms in the K/BxN model (24), the mechanism by cantly lower in the splenic PCs from K/BxNsf mice than in those which Tregs elicit their effects on components of humoral auto- from K/BxN mice or normal BxN mice (Fig. 6A). Accordingly, immunity remains to be explored. In the current study, we found FcgRIIb transcripts were significantly reduced in both B220+ cells that the absence of Tregs elicits quantitative and qualitative al- as a whole and in B220loCD138+ PCs from K/BxNsf mice (Fig. terations in diverse components of . Our data 6B). Therefore, these data demonstrate that the absence of Tregs showing more abundant numbers of TEFH and TFH cells in K/ results in downregulation of the death receptor FcgRIIb on PCs. BxNsf mice implicate Tregs in regulating the quantity of these To determine whether the reduced FcgRIIb expression affected cells in vivo. The abundance of TFH cells is accompanied by an immune complex-mediated cell death of the K/BxNsf PCs, puri- increase in the quantity of GCs and PCs. The increase in PCs far fied splenic PCs from K/BxN and K/BxNsf mice were incubated surpasses B cell insufficiency, which presumably stems from ab- with a mixture of GPI and K/BxN serum containing anti-GPI Abs normal in the BM, as reported recently in the or a mixture of GPI and normal BxN serum as a control. The K/ scurfy model (30). Importantly, the most prominent qualitative BxN PCs underwent greater apoptosis (annexin V+7-AAD2)in influence of Tregs is on the physiology of PCs. Contrary to the response to the GPI/K/BxN serum than to the GPI/BxN serum, behavior of normal PCs emerging in a Treg-replete milieu, PCs demonstrating that cross-linking of FcgRIIb by immune com- emerging in a milieu lacking Tregs are nonmigratory and long- plexes of GPI and anti-GPI Abs can induce PC apoptosis (Fig. lived. They survived at least for 4 wk without cell division, which 6C). In contrast, the GPI/K/BxN serum did not induce more ap- fits the criterion for long-lived PCs based on previous studies optosis of the K/BxNsf PCs than did the GPI/BxN serum, dem- using BrdU incorporation assays (31–34). However, because The Journal of Immunology 1551 Downloaded from

FIGURE 5. Cytotoxic effect of Tregs on PCs is elicited in a contact-de- pendent manner. A, Immunofluorescence staining of K/BxN spleen cry- osections with anti-CD4–PerCP–Cy5.5 (white), anti-Foxp3–FITC (green), http://www.jimmunol.org/ anti-B220–allophycocyanin (blue), and anti-CD138–PE (red). Foxp3+ Tregs in contact with CD138+ PCs in the extrafollicular zone of spleen FIGURE 6. Downregulation of FcgRIIb on splenic PCs from K/BxN are indicated by the arrows. B and C, Preactivated B220+ cells were mice. A, Splenocytes from 5- to 6-wk-old mice were stained with mAbs to cocultured with preactivated CD4+CD25+ cells in the presence of stimuli B220, CD138, and FcgRIIb/III and assayed by FACS. FACS profiles gated as described in Materials and Methods. The two populations were either on B220loCD138+ cells and mean fluorescence intensities are shown. B, cocultured in wells or incubated separately in Transwells (Tw). FACS Splenic whole B220+ cells and B220loCD138+ cells from the indicated profiles gated on B220loCD138+ cells after 8 h culture are shown (B). Viable mice were sorted by MACS and tested by real-time RT-PCR to quantify B220loCD138+ cell numbers after 48 h culture are shown as percentages (C). levels of FcgRIIb transcripts relative to b -microglobulin. C, B220lo The data are representative of three independent experiments. **p , 0.01 2 CD138+ PCs were incubated with GPI/K/BxN or GPI/BxN serum for 9 h (Student t test). by guest on September 30, 2021 and stained with annexin V and 7-AAD. The percentages of annexin V+7- AAD2 early apoptotic cells are shown. D, B220loCD138+ PCs were cul- tured in the presence or absence of Tregs for 24 h, stained with anti- others have reported that long-lived PCs can live for several FcgRII/III mAb, and assayed by FACS. Numbers indicate mean fluores- months (8, 9), it would be of interest to determine how long K/ cence intensities. The data are representative of at least two independent BxNsf PCs can actually survive in the spleen. Given the patho- experiments. *p , 0.05 (Student t test). genic characteristics of autoreactive long-lived PCs, our results suggest that the activity of Tregs is important for limiting the siveness to CXCL12 is prematurely terminated. It should also be persistence of humoral immune responses. noted that PC unresponsiveness to CXCL12 does not necessarily Previous studies have pointed to various kinds of cross-talk result in BM homing failure, as it has been shown that PCs can between regulatory cells and TFH cells. In autoimmune animals, home into the BM through a pathway independent of, albeit less the functions of TFH cells are attenuated by the induction of Tregs, efficient than, the CXCL12/CXCR4 axis (38). However, it is not but these cells retain LAP+Foxp32 and are thus distinct from likely that this alternative pathway is activated in the K/BxNsf Foxp3+ Tregs (35, 36). Moreover, although FOXP3+ Tregs iso- PCs, because they were barely to be seen in the BM. It is also lated from human tonsil limit certain functions of TFH cells in of interest that CXCR4 surface expression did not lead to vitro (12), it was not known whether they can suppress the devel- CXCL12-induced migration. We suspect that this uncoupling is opment of TEFH and TFH cells in vivo. Hence, our results pro- due to a defect in the signaling pathway downstream of CXCR4. vide new insight into the in vivo role of Foxp3+ Tregs in the de- In this context, our results imply that Tregs can affect the activity velopment of both TEFH and TFH cells. of PC-intrinsic factors involved in responsiveness to CXCL12. We found that the K/BxNsf PCs, which accumulated in ab- How this happens remains to be clarified. normally high numbers in the spleen and did not migrate into the Another molecular change found in K/BxNsf mice is the BM, were unresponsive to CXCL12 in ex vivo chemotaxis assays downregulation of FcgRIIb expression on splenic PCs. Because despite the fact that they expressed normal levels of its receptor, FcgRIIb is known to transduce a death signal to PCs upon en- CXCR4. Because CXCL12 is a major BM homing chemokine, this gagement with immune complexes (29), this downregulation pre- CXCL12 unresponsiveness may be responsible for the failure of the sumably permits them to escape from immune complex-medi- PCs to egress from the spleen. Our findings are in line with the ated cell death. Indeed, in agreement with this idea, we found observation that PC emigrants into the BM progressively lose their that K/BxNsf PCs were resistant to immune complex-medi- chemotactic responsiveness to CXCL12 as they become terminally ated cell death. This characteristic was also observed when PCs mature and long-lived, even though they continue to express from FcgRIIb-deficient mice were incubated with FcgRIIb- CXCR4 (37). It is conceivable that in the absence of Tregs, the specific cross-linking Abs (29), so confirming that the escape of K/ PCs age unusually rapidly in the spleen, so that their respon- BxNsf PCs from immune complex-mediated death arises from 1552 ROLE OF REGULATORY T CELLS IN PLASMA CELL HOMEOSTASIS their insufficiency of FcgRIIb. This escape cannot be attributed to 13. Lim, H. W., P. Hillsamer, A. H. Banham, and C. H. Kim. 2005. Cutting edge: a direct effect of Tregs, as we failed to detect upregulation of direct suppression of B cells by CD4+ CD25+ regulatory T cells. J. Immunol. 175: 4180–4183. FcgRIIb on PCs when Tregs were cocultured with PCs in vitro. 14. Zhao, D. M., A. M. Thornton, R. J. DiPaolo, and E. M. Shevach. 2006. Activated Hence, it is likely that factors extrinsic to, and influenced by, Tregs CD4+CD25+ T cells selectively kill B lymphocytes. 107: 3925–3932. 15. Iikuni, N., E. V. Lourenc¸o, B. H. Hahn, and A. La Cava. 2009. Cutting edge: affect FcgRIIb expression on PCs. We suspect that IL-4 may be regulatory T cells directly suppress B cells in systemic lupus erythematosus. J. involved in this phenomenon, because IL-4 has been known to Immunol. 183: 1518–1522. reduce FcgRIIb expression on B cells (39), and K/BxNsf mice 16. Grossman, W. J., J. W. Verbsky, W. Barchet, M. Colonna, J. P. Atkinson, and T. J. Ley. 2004. Human T regulatory cells can use the perforin pathway to cause contain more numerous IL-4–producing Th2 cells than do K/BxN autologous target cell death. Immunity 21: 589–601. mice in our observation (data not shown). 17. Seo, S. J., M. L. Fields, J. L. Buckler, A. J. Reed, L. Mandik-Nayak, S. A. Nish, Several studies have identified PCs with aberrant phenotypes in R. J. Noelle, L. A. Turka, F. D. Finkelman, A. J. Caton, and J. Erikson. 2002. The impact of T helper and T regulatory cells on the regulation of anti-double- models of SLE, and some of these phenotypes were evident among stranded DNA B cells. Immunity 16: 535–546. the PCs from the arthritic K/BxNsf mice. For instance, autoanti- 18. Morgan, M. E., R. P. Sutmuller, H. J. Witteveen, L. M. van Duivenvoorde, body-secreting long-lived PCs are abundant in the spleens of NZB/ E. Zanelli, C. J. Melief, A. Snijders, R. Offringa, R. R. de Vries, and R. E. Toes. 2003. CD25+ cell depletion hastens the onset of severe disease in collagen- W mice (9). Similarly, the majority of splenic PCs in the NZM2410 induced arthritis. Arthritis Rheum. 48: 1452–1460. mouse strain, another SLE model, are long-lived and defective 19. Scalapino, K. J., Q. Tang, J. A. Bluestone, M. L. Bonyhadi, and D. I. Daikh. in BM homing (23). Furthermore, there is a report showing the 2006. Suppression of disease in New Zealand Black/New Zealand White lupus- prone mice by adoptive transfer of ex vivo expanded regulatory T cells. J. absence of FcgRIIb expression on PCs from murine SLE models Immunol. 177: 1451–1459. (29). Our data prompted us to hypothesize that the aberrant phe- 20. Kouskoff, V., A. S. Korganow, V. Duchatelle, C. Degott, C. Benoist, and notypes of PCs in SLE are due to a numerical and/or functional D. Mathis. 1996. Organ-specific disease provoked by systemic autoimmunity. Downloaded from Cell 87: 811–822. deficit in Tregs. Given that a deficit in Tregs has been reported in 21. Kang, S. M., E. Jang, D. J. Paik, Y. J. Jang, and J. Youn. 2008. CD4+CD25+ patients with SLE (40, 41), this hypothesis seems to be worthy of regulatory T cells selectively diminish systemic autoreactivity in arthritic K/BxN further investigation. mice. Mol. Cells 25: 64–69. 22. Jang, E., S. H. Cho, H. Park, D. J. Paik, J. M. Kim, and J. Youn. 2009. A positive In conclusion, our study highlights the importance of Tregs in the feedback loop of IL-21 signaling provoked by homeostatic CD4+CD25- T cell regulation of humoral immunity. To our knowledge, this is the first expansion is essential for the development of arthritis in autoimmune K/BxN mice. J. Immunol. 182: 4649–4656. report showing that Tregs regulate the behavior of PCs as well as http://www.jimmunol.org/ 23. Erickson, L. D., L. L. Lin, B. Duan, L. Morel, and R. J. Noelle. 2003. A genetic the quantity of TEFH cells, TFH cells, and GCs in vivo. From the lesion that arrests plasma cell homing to the bone marrow. Proc. Natl. Acad. Sci. current observations, we conclude that the activity of Tregs is USA 100: 12905–12910. 24. Nguyen, L. T., J. Jacobs, D. Mathis, and C. Benoist. 2007. Where FoxP3- crucial for regulating PC homeostasis in autoimmune states, dependent regulatory T cells impinge on the development of inflammatory ar- thereby dampening humoral memory. thritis. Arthritis Rheum. 56: 509–520. 25. Wolniak, K. L., S. M. Shinall, and T. J. Waldschmidt. 2004. The germinal center response. Crit. Rev. Immunol. 24: 39–65. Acknowledgments 26. Hargreaves, D. C., P. L. Hyman, T. T. Lu, V. N. Ngo, A. Bidgol, G. Suzuki, We thank Dr. Diane Mathis for providing KRN mice, Dr. Hee Yong Chung Y. R. Zou, D. R. Littman, and J. G. Cyster. 2001. A coordinated change in chemokine responsiveness guides plasma cell movements. J. Exp. Med. 194: 45– for valuable discussion, and Dr. Julian Gross for editorial assistance.

56. by guest on September 30, 2021 27. Cao, X., S. F. Cai, T. A. Fehniger, J. Song, L. I. Collins, D. R. Piwnica-Worms, Disclosures and T. J. Ley. 2007. and perforin are important for regulatory T cell-mediated suppression of tumor clearance. Immunity 27: 635–646. The authors have no financial conflicts of interest. 28. Matsumoto, I., M. Maccioni, D. M. Lee, M. Maurice, B. Simmons, M. Brenner, D. Mathis, and C. Benoist. 2002. How antibodies to a ubiquitous cytoplasmic enzyme may provoke joint-specific . Nat. Immunol. 3: 360– References 365. 1. Shlomchik, M. J., J. E. Craft, and M. J. Mamula. 2001. From T to B and back again: 29. Xiang, Z., A. J. Cutler, R. J. Brownlie, K. Fairfax, K. E. Lawlor, E. Severinson, positive feedback in systemic autoimmune disease. Nat. Rev. Immunol. 1: 147–153. E. U. Walker, R. A. Manz, D. M. Tarlinton, and K. G. Smith. 2007. Fcgam- 2. Edwards, J. C., and G. Cambridge. 2006. B-cell targeting in rheumatoid arthritis maRIIb controls bone marrow plasma cell persistence and apoptosis. Nat. and other autoimmune diseases. Nat. Rev. Immunol. 6: 394–403. Immunol. 8: 419–429. 3. Martin, F., and A. C. Chan. 2004. Pathogenic roles of B cells in human auto- 30. Leonardo, S. M., J. A. Josephson, N. L. Hartog, and S. B. Gauld. 2010. Altered immunity; insights from the clinic. Immunity 20: 517–527. B cell development and anergy in the absence of Foxp3. J. Immunol. 185: 2147– 4. Kim, C. H., L. S. Rott, I. Clark-Lewis, D. J. Campbell, L. Wu, and E. C. Butcher. 2156. 2001. Subspecialization of CXCR5+ T cells: B helper activity is focused in a ger- 31. Shapiro-Shelef, M., K. I. Lin, D. Savitsky, J. Liao, and K. Calame. 2005. Blimp- minal center-localized subset of CXCR5+ T cells. J. Exp. Med. 193: 1373–1381. 1 is required for maintenance of long-lived plasma cells in the bone marrow. J. 5. Odegard, J. M., B. R. Marks, L. D. DiPlacido, A. C. Poholek, D. H. Kono, Exp. Med. 202: 1471–1476. C. Dong, R. A. Flavell, and J. Craft. 2008. ICOS-dependent extrafollicular helper 32. DiLillo, D. J., Y. Hamaguchi, Y. Ueda, K. Yang, J. Uchida, K. M. Haas, T cells elicit IgG production via IL-21 in systemic autoimmunity. J. Exp. Med. G. Kelsoe, and T. F. Tedder. 2008. Maintenance of long-lived plasma cells and 205: 2873–2886. serological memory despite mature and depletion during CD20 6. Radbruch, A., G. Muehlinghaus, E. O. Luger, A. Inamine, K. G. Smith, in mice. J. Immunol. 180: 361–371. T. Do¨rner, and F. Hiepe. 2006. Competence and competition: the challenge of 33. Duchez, S., R. Amin, N. Cogne´, L. Delpy, C. Sirac, V. Pascal, B. Corthe´sy, and becoming a long-lived plasma cell. Nat. Rev. Immunol. 6: 741–750. M. Cogne´. 2010. Premature replacement of mu with alpha immunoglobulin 7. Miller, J. J., III, and L. J. Cole. 1967. Resistance of long-lived lymphocytes and chains impairs lymphopoiesis and mucosal homing but promotes plasma cell plasma cells in rat lymph nodes to treatment with prednisone, cyclophospha- maturation. Proc. Natl. Acad. Sci. USA 107: 3064–3069. mide, 6-mercaptopurine, and actinomycin D. J. Exp. Med. 126: 109–125. 34. Neubert, K., S. Meister, K. Moser, F. Weisel, D. Maseda, K. Amann, C. Wiethe, 8. Slifka, M. K., R. Antia, J. K. Whitmire, and R. Ahmed. 1998. Humoral immunity T. H. Winkler, J. R. Kalden, R. A. Manz, and R. E. Voll. 2008. The proteasome due to long-lived plasma cells. Immunity 8: 363–372. inhibitor bortezomib depletes plasma cells and protects mice with lupus-like 9. Hoyer, B. F., K. Moser, A. E. Hauser, A. Peddinghaus, C. Voigt, D. Eilat, disease from nephritis. Nat. Med. 14: 748–755. A. Radbruch, F. Hiepe, and R. A. Manz. 2004. Short-lived plasmablasts and 35. Wu, H. Y., F. J. Quintana, and H. L. Weiner. 2008. Nasal anti-CD3 long-lived plasma cells contribute to chronic humoral autoimmunity in NZB/W ameliorates lupus by inducing an IL-10-secreting CD4+ CD25- LAP+ regulatory mice. J. Exp. Med. 199: 1577–1584. T cell and is associated with down-regulation of IL-17+ CD4+ ICOS+ CXCR5+ 10. Silverman, G. J., and S. Weisman. 2003. Rituximab therapy and autoimmune follicular helper T cells. J. Immunol. 181: 6038–6050. disorders: prospects for anti-B cell therapy. Arthritis Rheum. 48: 1484–1492. 36. Wu, H. Y., E. M. Center, G. C. Tsokos, and H. L. Weiner. 2009. Suppression 11. Sakaguchi, S., N. Sakaguchi, M. Asano, M. Itoh, and M. Toda. 1995. Immu- of murine SLE by oral anti-CD3: inducible CD4+CD25-LAP+ regulatory nologic self-tolerance maintained by activated T cells expressing IL-2 receptor T cells control the expansion of IL-17+ follicular helper T cells. Lupus 18: alpha-chains (CD25). Breakdown of a single mechanism of self-tolerance causes 586–596. various autoimmune diseases. J. Immunol. 155: 1151–1164. 37. Hauser, A. E., G. F. Debes, S. Arce, G. Cassese, A. Hamann, A. Radbruch, and 12. Lim, H. W., P. Hillsamer, and C. H. Kim. 2004. Regulatory T cells can migrate to R. A. Manz. 2002. Chemotactic responsiveness toward ligands for CXCR3 and follicles upon T cell activation and suppress GC-Th cells and GC-Th cell-driven CXCR4 is regulated on plasma blasts during the time course of a memory im- B cell responses. J. Clin. Invest. 114: 1640–1649. mune response. J. Immunol. 169: 1277–1282. The Journal of Immunology 1553

38. Nie, Y., J. Waite, F. Brewer, M. J. Sunshine, D. R. Littman, and Y. R. Zou. 2004. 40. Crispin, J. C., A. Martı´nez, and J. Alcocer-Varela. 2003. Quantification of reg- The role of CXCR4 in maintaining peripheral B cell compartments and humoral ulatory T cells in patients with systemic lupus erythematosus. J. Autoimmun. 21: immunity. J. Exp. Med. 200: 1145–1156. 273–276. 39. Rudge, E. U., A. J. Cutler, N. R. Pritchard, and K. G. Smith. 2002. 4 41. Miyara, M., Z. Amoura, C. Parizot, C. Badoual, K. Dorgham, S. Trad, D. Nochy, reduces expression of inhibitory receptors on B cells and abolishes CD22 and Fc P. Debre´, J. C. Piette, and G. Gorochov. 2005. Global natural regulatory T cell gamma RII-mediated B cell suppression. J. Exp. Med. 195: 1079–1085. depletion in active systemic lupus erythematosus. J. Immunol. 175: 8392–8400. Downloaded from http://www.jimmunol.org/ by guest on September 30, 2021

Top View