ARTHRITIS & RHEUMATOLOGY Vol. 67, No. 4, April 2015, pp 1012–1022 DOI 10.1002/art.38989 © 2015, American College of Rheumatology

Sialic Acid–Binding Immunoglobulin-Type Lectin H–Positive Plasmacytoid Dendritic Cells Drive Spontaneous Lupus-like Disease Development in B6.Nba2 Mice Laura M. Davison and Trine N. Jørgensen Objective. Patients with systemic lupus erythematosus (SLE) often present with elevated levels of interferon-␣ (IFN␣) in serum. Plasmacytoid dendritic cells (pDCs) have been suggested to be the primary source of IFN␣ in SLE due to their capacity to produce high levels of IFN␣. During viral infection, a subset of pDCs expressing sialic acid–binding immunoglobulintype lectin H (Siglec H) produces the majority of pDCderived IFN␣. The aim of this study was to provide evidence that Siglec H–positive pDCs are pathogenic in the IFN␣-dependent B6.Nba2 mouse model of lupus. Methods. B6.Nba2 blood dendritic cell antigen 2 (BDCA-2)–diphtheria toxin receptor (DTR)–transgenic (Tg) mice were treated intraperitoneally with DT 3 times weekly starting at 4 weeks or 12 weeks of age and analyzed at 12 weeks and 18 weeks of age, respectively. Lupus-like disease development was measured by the presence of elevated levels of autoantibodies in serum (as determined by enzyme-linked immunosorbent assay), increased expression of IFN-inducible genes (as determined by real-time reverse transcription–polymerase chain reaction), increased IgG immune complex deposition in kidney glomeruli (as determined by immunofluorescence staining), spontaneous lymphocyte activation, and differentiation of B cells into antibodyproducing plasma cells (as determined by flow cytometry).

Results. B6.Nba2 mice in which Siglec H–positive pDCs were depleted for 6–8 weeks displayed reduced levels of IFN␣-induced gene transcripts and decreased anti-chromatin autoantibody levels in serum, and significantly fewer activated splenic T cells and B cells, germinal center B cells, follicular helper T cells, and splenic plasma cells. In 18-week-old mice, IgG immune complex deposition in kidney glomeruli was similarly reduced. Conclusion. The development of lupus-like disease in congenic B6.Nba2 mice depends on Siglec H– positive pDCs. We suggest that depletion of Siglec H–positive pDCs represents a novel cellular target in SLE. Systemic lupus erythematosus (SLE) is an autoimmune disorder that potentially targets any organ in the body; therefore, clinical manifestations of SLE often include inflammation of the lungs, joints, kidneys, central nervous system, or skin (1). Serologic abnormalities such as lymphopenia and thrombopenia, along with the presence of autoantibodies targeting nuclear antigens such as DNA, RNA, and histones, are measurable in the majority of patients. Over the past 4–5 decades, it has become clear that a significant proportion of patients also display elevated levels of interferon-␣ (IFN␣)– regulated gene transcripts within peripheral blood mononuclear cells (PBMCs) (2,3). However, the cause and the source of elevated levels of IFN␣ in the serum of patients with SLE remain unknown. Although many cells are capable of producing IFN␣ in response to viral infection, plasmacytoid dendritic cells (pDCs) and neutrophils have most often been associated with IFN␣ production in patients with SLE (4–6). Plasmacytoid DCs express high levels of intracellular Toll-like receptor 7 (TLR-7), TLR-8, and TLR-9 specific for single-stranded RNA and unmethylated DNA, respectively. Although the specificity of these

Supported by Cleveland Clinic Seed Funding. Laura M. Davison, BS, Trine N. Jørgensen, PhD: Lerner Research Institute, Cleveland Clinic Foundation, and Cleveland Clinic Lerner College of Medicine of Case Western Reserve University, Cleveland, Ohio. Address correspondence to Trine N. Jørgensen, PhD, Department of Immunology, NE40, Lerner Research Institute, Cleveland Clinic Foundation, 9500 Euclid Avenue, Cleveland, OH 44195. E-mail: [email protected]. Submitted for publication October 23, 2014; accepted in revised form December 4, 2014. 1012

Siglec H–POSITIVE pDCs DRIVE DISEASE IN B6.Nba2 MICE

1013

TLRs was likely meant (evolutionarily) to evoke responses to viral and intracellular bacterial infections, cross-reactivity to self nucleotides released during apoptosis or NETosis (Neutrophil extracellular trap [NET] release by dying cells) is known to occur (7,8). Investigations into the specificity of IFN␣producing pDCs in mice revealed the existence of a specific cell subset expressing the surface marker sialic acid–binding immunoglobulin-type lectin H (Siglec H) (9–11). Most impressively, blocking antibodies against Siglec H prevented IFN␣ production in response to CpG-containing oligonucleotide (TLR-9) stimulation (9). Although Siglec H is a mouse-specific antigen, IFN␣-producing pDCs isolated from patients with SLE and healthy individuals have been shown to express blood dendritic cell antigen 2 (BDCA-2) and BDCA-4 (12). Based on this information, transgenic (Tg) mice expressing the diphtheria toxin receptor (DTR) under control of the human BDCA-2 promoter were created (13). Despite the human origin of the BDCA-2 promoter, treatment of B6.BDCA-2–DTR-Tg mice with DT specifically ablated Siglec H–positive pDCs in mice, making the system valuable for investigations into the role of Siglec H–positive pDCs during inflammation and autoimmunity (13). We backcrossed the BDCA-2–DTR transgene onto congenic B6.Nba2 mice in order to study the effect of Siglec H–positive pDCs on spontaneous development of lupus-like disease. Female B6.Nba2 mice develop a lupus-like disease recapitulating many of the clinical characteristics of SLE, including hypergammaglobulinemia, elevated levels of antinuclear autoantibodies (ANAs) in serum, lymphadenopathy, splenomegaly, IgG immune complex deposition in the kidneys, glomerulonephritis, and elevated levels of IFN␣ in serum (14–18). In fact, disease development in B6.Nba2 mice depends on functional IFN␣ signaling (19). Furthermore, we previously showed that pDCs accumulate in B6.Nba2 mice as they age (16). Using B6.Nba2.BDCA-2–DTRTg mice, we show that ablation of the Siglec H–positive subpopulation of pDCs in young female B6.Nba2 mice is sufficient to reduce serum IFN␣ levels and prevent disease initiation.

gensen. B6.Nba2.BDCA-2–DTR-Tg mice were generated inhouse by backcrossing the BDCA-2–DTR transgene onto the B6.Nba2 background. Expression of the Nba2 locus was established by Mit marker–assisted polymerase chain reactions (PCRs), as previously described (16), while genotyping for BDCA-2–DTR was performed according to the protocol provided by The Jackson Laboratory. All mice used were female. The mice were maintained in the Biological Research Unit at the Lerner Research Institute, in accordance with Cleveland Clinic Foundation Animal Research Committee guidelines, and all procedures were approved by the Institutional Animal Care and Use Committee of the Lerner Research Institute of the Cleveland Clinic Foundation and conducted in compliance with guidelines issued by the National Institutes of Health. Protocol for Siglec H–positive pDC depletion. B6.Nba2.BDCA-2–DTR-Tg mice were treated intraperitoneally with 100 ng DT in 200 ␮l sterile 1⫻ phosphate buffered saline (PBS) 3 times per week starting at either 4 weeks or 12 weeks of age. For all experiments, DT-treated B6.Nba2. BDCA-2–DTR-Tg–negative littermates and nonmanipulated, age-matched B6 mice were included as controls. All end point analyses were performed in mice that had received DT treatment ⱕ24 hours previously. Flow cytometric analysis. Splenic single-cell suspensions were prepared by gently separating single cells between the frosted areas of 2 microscope slides, and red blood cells were lysed using 1⫻ ACK lysing buffer (0.15M NH4Cl, 0.01M KHCO3, 0.1 mM EDTA). Cells were incubated with unlabeled anti-CD16/32 antibodies to reduce nonspecific Fc receptor– dependent binding of fluorescence-labeled antibodies, followed by staining with fluorescence-labeled antibodies specific for B220, CD3, CD4, CD8, CD11c, CD19, CD21/35, CD23, CD25, CD44, CD62L, CD69, CD86, CD93 (AA4.1), CXCR5, GL7, IgD, IgM, class II major histocompatibility complex (MHC), programmed death 1 (PD-1), PDC antigen 1, Siglec H (all from eBioscience), and CD138 (PharMingen). For detection of follicular DCs (fDCs), single-cell suspensions were incubated with fluorescence-labeled antibodies specific for CD11c, CD16/32, CD21/35, and CD86 without prior blocking of Fc receptors. Flow cytometry was performed using a FACSCalibur (BD Biosciences), and FlowJo version 9.7.5 (Tree Star) was used for all analyses. Gating strategies for all cell populations analyzed are shown in Supplementary Figure 1 (available on the Arthritis & Rheumatology web site at http://onlinelibrary.wiley.com/doi/10.1002/art.38989/abstract). Total spleen counts in each mouse were acquired and used to calculate the total numbers of cell subsets. Real-time reverse transcription–polymerase chain reaction (RT-PCR). Splenic conventional DC (cDC) and pDC populations (Siglec H–negative and Siglec H–positive) were sorted by high-speed cell sorting using a FACSAria I (BD Biosciences) (see Supplementary Figure 1), counted, and frozen at ⫺80°C until used. Similarly, PBMCs were obtained from DT-treated DTR-Tg–positive and DTR-Tg–negative mice and saved for analysis. Total RNA was isolated from frozen cells using an RNeasy Plus Micro Kit (Qiagen), and complementary DNA (cDNA) was prepared using qScript cDNA SuperMix (Quanta BioSciences). The presence of cytokine and IFN␣-inducible gene transcripts was determined using a SYBR Green system (Life Technologies) and the following primers: for pan Ifna, forward 5⬘-CTTCCACAGG-

MATERIALS AND METHODS Mice. C57BL/6 and C57BL/6-Tg(CLEC4C-HBEGF) 956Cln/J (B6.BDCA-2–DTR-Tg) mice (13) were purchased from The Jackson Laboratory. B6.Nba2 mice were originally generated by Dr. Brian Kotzin (University of Colorado Health Sciences Center) and subsequently transferred to Dr. Jør-

1014

ATCACTGTGTACCT-3⬘ and reverse 5⬘-TTCTGCTCTGACCACCTCCC-3⬘; for Il6, forward 5⬘-AGTTGCCTTCTTGGGACTGA-3⬘ and reverse 5⬘-CAGAATTGCCATTGCACAAC3⬘; for Il10, forward 5⬘-AGCCTTATCGGAAATGATCCAGT-3⬘ and reverse 5⬘-GGCCTTGTAGACACCTTGGT-3⬘; for Mx1, forward 5⬘-TTCCTGAAGAGGCGGCTTT-3⬘ and reverse 5⬘GGTTAATCGGAGAATTTGGCAA-3⬘; for Ifi202, forward 5⬘-GGTCATCTACCAACTCAGAAT-3⬘ and reverse 5⬘-CTC TAGGATGCCACTGCTGTTG-3⬘. Levels of ␤-actin were determined (forward 5⬘-TGGGAATGGGTCAGAAGGAC-3⬘ and reverse 5⬘-GGTCTCAAACATGATCTGGG-3⬘) and used for standardization. Real-time RT-PCR was performed with an ABI 7300 PCR system (Applied Biosystems). Enzyme-linked immunosorbent assay. Levels of total IgG and IgM in serum were measured as previously described (19). Serum samples obtained from DTR-Tg–positive and DTR-Tg–negative B6.Nba2 mice before and during DT treatment were diluted 1:300 in serum diluent (sterile filtered 0.5% bovine gamma globulin, 5% gelatin, 0.05 mM Tween in 1⫻ PBS) and analyzed for levels of anti-chromatin IgG autoantibodies. For measurements of total IgG and IgM levels, sera were diluted 1:100,000 in serum diluent. Briefly, microtiter plates (Immulon 2HD) were coated with total mouse immunoglobulin (SouthernBiotech) or purified chromatin overnight at 4°C, blocked in 5% gelatin/PBS for ⱖ2 hours, and incubated with diluted serum samples for 2 hours. Secondary horseradish peroxidase–conjugated anti-mouse IgG or IgM antibodies (Invitrogen) were added for 1.5 hours, and the plates were developed using 10 mg/ml ABTS in McIlwain’s buffer (0.09M Na2HPO4, 0.06M citric acid, pH 4.6) or a TMB Substrate Kit (Thermo Scientific). Colorimetric readings were obtained using a VICTOR3 plate reader (PerkinElmer). Immunostaining. For histologic analyses, half kidneys were harvested and fixed in 10% formalin. The glomerular structure was determined, and inflammatory cells were quantified in two 5-␮m sections 30 ␮m apart after hematoxylin and eosin staining (Newcomer Supply). For detection of IgG and C3, half kidneys were quick-frozen in OCT, and 5-␮m sections were prepared and stained using Texas Red– conjugated anti-mouse IgG (Invitrogen) and fluorescein isothiocyanate (FITC)–conjugated anti-mouse C3–specific antibodies (Immunology Consultants Laboratory). Images were obtained using a 20x/0.7 NA HC PlanApo objective lens on a Leica DMR upright microscope equipped with a Retiga EXi CCD camera (cooled model; QImaging). IgG deposition was quantified in 12 glomeruli using Image-Pro Plus software (Media Cybernetics) and is reported as arbitrary units representing the average intensity per area (glomerulus). Detection of germinal center (GC) B cells within spleens was accomplished using FITC-conjugated anti-B220 antibodies, biotinylated anti-GL7 antibodies (eBioscience), and Alexa Fluor 568– conjugated streptavidin (Invitrogen) on 5-␮m frozen sections. The mean area of GCs per mouse was determined by averaging the area of each GC present within a 10⫻ field of view (0–4 GCs/mouse) using Image-Pro software and a microscopespecific conversion of 1 linear pixel ⫽ 1.57 ␮m. All images were obtained using identical microscope settings for fluorescence gain and exposure time. Statistical analysis. Differences between cell populations were calculated using Student’s unpaired t-test with Welch’s correction. Differences based on fluorescence staining

DAVISON AND JØRGENSEN

(IgG deposition and GCs) were determined by Mann-Whitney nonparametric test. Differences between longitudinal data (serum antibodies) were determined using two-way analysis of variance. For all analyses, P values less than 0.05 were considered significant.

RESULTS Effect of Siglec H–positive pDC depletion on serum levels of IFN␣ in DTR-Tg B6.Nba2 mice. Lupuslike disease develops in congenic B6.Nba2 mice in an IFN␣-dependent manner (19). In addition, we previously showed that the population of splenic pDCs accumulates in the B6.Nba2 lupus-prone mouse model during aging (16). Other studies have suggested that a subset of pDCs expressing Siglec H (Siglec H–positive pDCs) is the primary source of IFN␣ (9,13); however, the specific role of this population of cells in B6.Nba2 mice has not been investigated. We generated B6.Nba2.BDCA-2–DTR-Tg mice, in which Siglec H–positive pDCs are selectively susceptible to DT-mediated depletion (13). To confirm the immediate effect of DT treatment in our system, we initially treated 6–10-week-old B6.Nba2.BDCA-2– DTR-Tg–positive mice with DT or PBS and, after 24-72 hours, performed flow cytometry to analyze the levels of splenic Siglec H–positive pDCs. The number of Siglec H–positive pDCs was significantly reduced in DTtreated DTR-Tg–positive B6.Nba2 mice compared with PBS-treated DTR-Tg–positive mice (Figure 1A), while all other cell populations remained unchanged during this period of time (Figures 1B and C, and data not shown). It should be noted that as in our previous study (16), all Siglec H–positive pDCs coexpressed PDCA-1 (see Supplementary Figure 1, available on the Arthritis & Rheumatology web site at http://onlinelibrary.wiley.com/ doi/10.1002/art.38989/abstract), and therefore, PDCA-1– positive pDCs were equally depleted in DT-treated DTR-Tg–positive B6.Nba2 mice (data not shown). Given the clear correlation between Siglec H– positive pDCs and IFN␣ production during viral infection (20), we next assessed whether the immediate and strong depletion of Siglec H–positive pDCs affected endogenous levels of IFN␣-induced gene transcripts in lupus-prone B6.Nba2 mice. The expression levels of 2 distinct IFN␣-induced genes, Mx1 and Ifi202, were significantly reduced in DT-treated DTR-Tg–positive mice compared with untreated mice (P ⬍ 0.001) (Figure 1D), suggesting that Siglec H–positive pDCs spontaneously produce IFN␣ in young B6.Nba2 mice.

Siglec H–POSITIVE pDCs DRIVE DISEASE IN B6.Nba2 MICE

1015

Figure 1. Depletion of sialic acid–binding immunoglobulin-type lectin H (Siglec H; SigH)–positive plasmacytoid dendritic cells (pDCs) is specific and immediately reduces serum interferon-␣ levels in B6.Nba2 mice. B6.Nba2 blood dendritic cell antigen 2 (BDCA-2)–diphtheria toxin receptor–transgenic (DTR⫹) mice were treated with DT (n ⫽ 6) or phosphate buffered saline (PBS) (⫺; n ⫽ 7). A–C, After 24–72 hours, absolute numbers of splenic Siglec H–positive pDCs (A), Siglec H–negative pDCs (B), and conventional DCs (cDCs) (C) were determined by flow cytometry. Each symbol represents an individual mouse; horizontal lines show the mean. ⴱⴱ ⫽ P ⬍ 0.01. D, Levels of Mx1 and Ifi202 gene transcripts in peripheral blood mononuclear cells obtained 24–72 hours after treatment were determined by real-time reverse transcription–polymerase chain reaction. Values are the mean ⫾ SEM expression levels after normalization to ␤-actin levels. The levels in PBS-treated mice were set at 1. ⴱⴱⴱ ⫽ P ⬍ 0.001 versus PBS.

Reduced measures of lupus-like disease in B6.Nba2 mice following long-term depletion of Siglec H–positive pDCs. A significant accumulation of pDCs occurs in B6.Nba2 mice between the ages of 8 weeks and 16 weeks (16). Nonetheless, the depletion of Siglec H–positive pDCs in young (6–10-week-old) B6.Nba2 mice reduced spontaneous levels of serum IFN␣ (see Figure 1D). To test whether Siglec H–positive pDCs (and possibly Siglec H–positive pDC-derived IFN␣) is involved in disease initiation and/or progression in B6.Nba2 mice, we treated female DTR-Tg–positive and DTR-Tg–negative B6.Nba2 mice with DT 3 times/week starting at either 4 weeks or 12 weeks of age. Mice were followed up for up to 8 weeks and 6 weeks, respectively, and killed 24 hours after administration of the last injection of DT. Upon harvesting, we initially assessed known characteristics of lupus-like disease in B6.Nba2 mice,

including elevated ANA levels in serum, glomerulonephritis, and IgG immune complex deposition. Levels of antichromatin IgG in serum were significantly reduced in both age groups treated with DT (Figure 2A). This effect was sustained throughout the experiment; however, interestingly, it did not appear to reduce the levels of already-existing anti-chromatin IgG antibodies in the older group of mice but rather prevented de novo generation of antibodies (Figure 2A). Consistent with previous observations (16,17,21), neither anti-histone IgG nor anti–double-stranded DNA IgG reached detectable levels in B6.Nba2.BDCA-2–DTR-Tg mice at these ages (data not shown). Total IgM in serum was also reduced in all Siglec H–positive pDC-depleted mice, while IgG levels in serum were significantly affected only in the group treated from 4 to 12 weeks of age (Figures 2B and C).

1016

Figure 2. Long-term depletion of Siglec H–positive pDCs results in reduced autoantibody levels in B6.Nba2 mice. Four-week-old or 12-weekold DTR-transgenic (DTRtg⫹) B6.Nba2 mice (n ⫽ 6 and n ⫽ 7, respectively) and DTR-transgenic–negative (DTRtg⫺) B6.Nba2 mice (n ⫽ 5 and n ⫽ 9, respectively) were treated with DT 3 times per week for 8 weeks or 6 weeks, respectively. Untreated B6 mice (controls) were bled at 12 weeks and 18 weeks of age. Serum was obtained before starting treatment and every 2 weeks thereafter, and levels of antichromatin IgG (A), total IgM (B), and total IgG (C) were determined by enzyme-linked immunosorbent assay. P values were calculated using two-way analysis of variance, comparing antibody levels in DT-treated DTR-transgenic– negative and DTR-transgenic–positive B6.Nba2 mice over time. Values are the mean ⫾ SEM. See Figure 1 for definitions.

Although B6.Nba2 mice do not succumb to renal failure, older mice develop glomerulonephritis and significant IgG immune complex deposition in kidney

DAVISON AND JØRGENSEN

glomeruli (16,19). We analyzed kidneys obtained from the group treated from 12 to 18 weeks of age for signs of reduced glomerulonephritis and/or reduced IgG immune complex deposition. Both glomerulonephritis and overall renal pathology were largely unchanged by the depletion of Siglec H–positive pDCs, with DTR-Tg– positive mice and DTR-Tg–negative B6.Nba2 mice showing comparably sized glomeruli (P ⫽ 0.2) and similar levels of cellular infiltrates in glomeruli (Figures 3A and B, and results not shown). In contrast, DTtreated DTR-Tg–positive B6.Nba2 mice displayed significantly less IgG deposition compared with DTR-Tg– negative B6.Nba2 mice (P ⫽ 0.05) (Figures 3C and D), although the levels did not reach those of B6 control mice (P ⫽ 0.06). Finally, upon completion of each treatment period, spleens from DTR-Tg–positive, DTR-Tg–negative, and control age-matched B6 mice were harvested and processed for analysis. Depletion of Siglec H–positive pDCs in DTR-Tg–positive B6.Nba2 mice from 4–12 weeks of age resulted in significantly reduced splenomegaly compared with that in DTR-Tg–negative B6.Nba2 mice (P ⬍ 0.05), while treatment from 12 to 18 weeks of age had no effect on spleen weight and the numbers of recovered splenocytes (see Supplementary Figure 2, available on the Arthritis & Rheumatology web site at http://onlinelibrary.wiley.com/doi/10.1002/ art.38989/abstract). Prevention of early spontaneous activation of T cells and B cells in B6.Nba2 mice by depletion of Siglec H–positive pDCs. It was previously shown that B cells accumulate in B6.Nba2 mice, and that these cells displays a spontaneously activated phenotype as indicated by the expression of CD69 (18). Both DTR-Tg– positive and DTR-Tg–negative B6.Nba2 mice had more B220⫹ B cells compared with B6 control mice (see Supplementary Figure 2), and depletion of Siglec H–positive pDCs did not significantly affect the levels of splenic B220⫹ B cells. We next analyzed numbers and percentages of activated B lymphocytes in DT-treated DTR-Tg–positive and DTR-Tg–negative B6.Nba2 mice to evaluate whether Siglec H–positive pDCs are involved in this process. The percentages of B220⫹ B cells expressing CD69, CD86, or class II MHC were increased in young DTR-Tg–negative B6.Nba2 mice compared with control B6 mice (P ⬍ 0.01 for all) as well as compared with Siglec H–positive pDC-depleted mice (P ⫽ 0.06, P ⫽ 0.23, and P ⬍ 0.01, respectively) (Figure 4A, and results not shown). In contrast, depletion of Siglec H–positive pDCs in 12–18-week-old mice had no effect on the activation status of B220⫹ B cells (Figure 4A, and results not shown).

Siglec H–POSITIVE pDCs DRIVE DISEASE IN B6.Nba2 MICE

1017

Figure 3. Long-term depletion of Siglec H–positive pDCs results in decreased IgG immune complex deposition in B6.Nba2 mice. Kidneys from 18-week-old B6, DTR-transgenic–negative (DTR⫺) B6.Nba2, and DTR-transgenic–positive (DTR⫹) B6.Nba2 mice were obtained after the last injection of DT (12–18 weeks of age). A and B, Glomerulonephritis and glomerular size were unchanged in Siglec H–positive pDC–depleted mice. C and D, Glomerular IgG immune complex deposition, but not complement factor C3 fixation, was diminished in DTR-transgenic–positive B6.Nba2 mice. Arrows in A and C indicate glomeruli. In B and D, each symbol represents an individual mouse; horizontal lines show the mean. ⴱ ⫽ P ⬍ 0.05. Original magnification ⫻ 20 in A and C. H&E ⫽ hematoxylin and eosin (see Figure 1 for other definitions).

Although it has been shown that CD4⫹ T cells are involved in lupus pathogenesis in B6.Nba2 mice (21), the phenotype of T cells in B6.Nba2 mice has not been studied. The total number of CD4⫹ T cells, but not CD8⫹ T cells, was increased in 12-week-old but not 18-week-old control DTR-Tg–negative B6.Nba2 mice compared with age-matched B6 mice (see Supplementary Figure 2, available on the Arthritis & Rheumatology web site at http://onlinelibrary.wiley.com/doi/10.1002/art.

38989/abstract). Interestingly, 12-week-old DT-treated DTR-Tg–positive B6.Nba2 mice displayed reduced numbers of both CD4⫹ T cells (P ⬍ 0.05) and CD8⫹ T cells (P ⬍ 0.01), which suggests that Siglec H–positive pDCs may play a role in the general maintenance of T cells. Further analyses of splenic CD4⫹ T cell subsets revealed that depletion of Siglec H–positive pDCs significantly reduced the percentage of activated CD69⫹CD4⫹ T cells in both age groups (Figure 4B).

1018

DAVISON AND JØRGENSEN

Figure 4. Depletion of Siglec H–positive pDCs prevents early spontaneous activation of T cells and B cells in B6.Nba2 mice. The spleens of DTR-transgenic–positive (DTR⫹) B6.Nba2 mice (n ⫽ 6–9) and DTR-transgenic–negative B6.Nba2 mice (n ⫽ 5–7) were harvested 24 hours after the last injection of DT and analyzed for levels of activated CD69⫹ B cells (A) and CD69⫹CD4⫹ T cells (B), and the levels of naive versus effector/memory CD4⫹ cells (C). Age-matched B6 mice (n ⫽ 3–5) served as controls. Each symbol represents an individual mouse; horizontal lines show the mean. ⴱ ⫽ P ⬍ 0.05; ⴱⴱ ⫽ P ⬍ 0.01; ⴱⴱⴱ ⫽ P ⬍ 0.001. See Figure 1 for definitions.

The percentage of CD62L⫹ naive CD4⫹ T cells was also significantly increased in young, but not old, DTtreated DTR-positive B6.Nba2 mice (Figure 4C); however, there was no significant difference in either the percentage of CD44highCD62Llow effector/memory CD4⫹ T cells (mean ⫾ SEM 1.4% ⫾ 0.49 versus 1.95% ⫾ 0.31; P ⫽ 0.07) or the ratio of naive-to-effector/ memory CD4⫹ T cells between DTR-Tg–negative and DTR-Tg–positive B6.Nba2 mice in either age group (Figure 4C, and results not shown). Taken together, these data indicated that sustained depletion of Siglec H–positive pDCs starting at 4 weeks of age significantly reduced spontaneous T cell and B cell activation in lupus-prone B6.Nba2 mice in vivo. Siglec H–positive pDC depletion results in increased numbers of marginal zone (MZ) B cells but reduced numbers of DCs. Lupus-prone B6.Nba2 mice have been reported to harbor reduced numbers of MZ B cells (18). We therefore analyzed DT-treated DTRTg–negative and DTR-Tg–positive B6.Nba2 mice and control B6 mice for the proportion of splenic B cell subsets. In contrast with previously reported results, we observed that the overall numbers of follicular mature (FM) and MZ B cells were significantly increased in 12-week-old DTR-Tg–negative B6.Nba2 mice compared with age-matched B6 controls, although there were no significant differences in the percentages of FM, MZ, and T1 B cells (see Supplementary Figure 2, available on

the Arthritis & Rheumatology web site at http:// onlinelibrary.wiley.com/doi/10.1002/art.38989/abstract). At 18 weeks of age, the overall number of B220⫹ B cells and both the percentages and numbers of FM and MZ B cells were increased in DTR-Tg–negative B6.Nba2 mice compared with B6 mice (see Supplementary Figure 2). Ablation of Siglec H–positive pDCs led to an additional increase in the percentage of MZ B cells in young DTR-Tg–positive B6.Nba2 mice (P ⬍ 0.05), although this effect was not measurable in 18-week-old Siglec H–positive pDC-depleted B6.Nba2 mice (P ⫽ 0.23). Finally, to determine whether Siglec H–positive pDC depletion affected levels of other DC subsets, we analyzed young and old DT-treated DTR-Tg–positive and DTR-Tg–negative B6.Nba2 mice for levels of Siglec H–negative pDCs and cDCs. Consistent with a more dramatic effect in young mice, we observed significantly fewer Siglec H–negative pDCs and cDCs in young, but not old, DTR-Tg–positive B6.Nba2 mice (see Supplementary Figure 2, available on the Arthritis & Rheumatology web site at http://onlinelibrary.wiley.com/doi/10.1002/ art.38989/abstract). This reduction in splenic DC levels was not attributable to off-target effects of DT, because neither cell population changed significantly for up to 72 hours posttreatment (Figures 1B and C). Reduction in spontaneous GC formation in B6.Nba2 mice by Siglec H–positive pDC depletion. During immune responses, CD4⫹ T cells provide essen-

Siglec H–POSITIVE pDCs DRIVE DISEASE IN B6.Nba2 MICE

1019

Figure 5. Spontaneous germinal center (GC) formation in young B6.Nba2 mice is reduced by depletion of Siglec H–positive pDCs. A, Levels of splenic follicular helper T (Tfh) cells, GC B cells, and CD138⫹IgM⫺ plasma cells in 12- and 18-week-old DTR-transgenic–positive and DTR-transgenic–negative B6.Nba2 mice were determined by flow cytometry. Age-matched B6 mice served as controls. B, GCs were identified in spleen sections after immunofluorescence staining for B220 (B cells) and GL7 (GCs). The size of GCs was calculated. Symbols represent the average GC area (n ⫽ 2–5 GCs) per mouse. Original magnification ⫻10. C, Intrasplenic levels of interleukin-21 (IL-21) were determined by enzyme-linked immunosorbent assay. Values are the mean ⫾ SEM. In A and B (bottom), each symbol represents an individual mouse; horizontal lines show the mean. ⴱ ⫽ P ⬍ 0.05; ⴱⴱ ⫽ P ⬍ 0.01; ⴱⴱⴱ ⫽ P ⬍ 0.001. See Figure 1 for other definitions.

tial antigen-specific help to differentiating B cells in specialized compartments known as GCs (22). We analyzed levels of CD35⫹ fDCs, follicular helper T (Tfh) cells (CXCR5⫹PD-1⫹CD4⫹), GC B cells (B220⫹GL7⫹IgM⫺CD38low), and CD138⫹ plasma cells in Siglec H–positive pDC-depleted and control mice (for gating strategies, see Supplementary Figure 1, available on the Arthritis & Rheumatology web site at http://onlinelibrary.wiley.com/doi/10.1002/art.38989/ abstract). Overall, the numbers of GC-related cell populations were significantly increased in both 12-week-old and 18-week-old DTR-Tg–negative B6.Nba2 mice compared with age-matched B6 mice, suggesting that GC formation occurs spontaneously and at high frequencies in B6.Nba2 mice (Figure 5A and data not shown). Remarkably, depletion of Siglec H–positive pDCs significantly reduced the numbers of all GCrelated cell populations in mice treated for 4–12 weeks (P ⬍ 0.05), although only the numbers of GC B cells were reduced enough to reach the much lower levels observed in B6 mice (Figure 5A). In contrast, only the numbers of Tfh cells were significantly affected by Siglec H–positive pDC depletion in older mice (P ⬍ 0.05)

(Figure 5A). Confirming these findings, depletion of Siglec H–positive pDCs in young, but not old, B6.Nba2 mice significantly reduced the numbers and average size of GCs (Figure 5B). Finally, we determined the levels of IL-21 in spleens from DT-treated DTR-Tg–negative and DTR-Tg–positive B6.Nba2 mice, as IL-21 is produced by Tfh cells and is critical for GC formation (23). Correlating with the reduced numbers of Tfh cells and smaller GCs in young mice treated for 4–12 weeks, depletion of Siglec H–positive pDCs resulted in slightly reduced intrasplenic IL-21 levels (P ⫽ 0.2). There was no difference in the levels of IL-21 in mice treated for 12–18 weeks. Spontaneous expression of elevated levels of both Ifna and Il6 messenger RNA (mRNA) in Siglec H– positive pDCs from B6.Nba2 mice. Multiple cytokines, including IL-6, IL-10, IL-21, and IFN␣, have been shown to control GC reactions (24,25). Based on the results of numerous studies, we anticipated that Siglec H–positive pDCs would be the main source of IFN␣ in B6.Nba2 mice; however, whether this population would also produce IL-6 and/or IL-10 was less predictable. We performed cell sorting of cDCs, Siglec H–negative

1020

DAVISON AND JØRGENSEN

Figure 6. Siglec H–positive pDCs in B6.Nba2 mice spontaneously produce Ifna and Il6, while Siglec H–negative pDCs produce Il10. Conventional DCs (cDCs; CD11chighB220⫺), Siglec H–negative pDCs (CD11cintermediateB220⫹ Siglec H–negative), and Siglec H–positive pDCs (CD11cintermediateB220⫹ Siglec H–positive) were isolated by high-speed cell sorting. Gene expression levels of Ifna (A), Il6 (B), and Il10 (C) were determined by real-time reverse transcription– polymerase chain reaction analysis. Values are the mean ⫾ SEM. ND ⫽ not detectable (see Figure 1 for other definitions).

pDCs, and Siglec H–positive pDCs (⬎98% pure) from nonmanipulated 8-week-old female B6 and B6.Nba2 mice and assessed the expression of Ifna, Il6, and Il10 mRNA. As expected, only Siglec H–positive pDCs expressed high levels of Ifna mRNA (Figure 6A). Interestingly, the level of Il6 mRNA expression was highest in Siglec H–positive pDCs (Figure 6B), while only Siglec H–negative pDCs produced Il10 mRNA (Figure 6C). Thus, depletion of Siglec H–positive pDCs in B6.Nba2 mice likely reduces not only IFN␣ levels but also IL-6 levels in vivo. DISCUSSION Plasmacytoid DCs are rare cells with the propensity to produce extremely high levels of IFN␣ upon viral infection (for review, see ref. 26). Patients with SLE have elevated levels of IFN␣-inducible gene transcripts, suggesting higher levels of endogenous IFN ␣ (4,17,27,28). Plasmacytoid DCs from SLE patients have

been shown to spontaneously secrete IFN␣, IL-6, and IL-10 (4,16,25). Thus, pDCs have been repeatedly suggested as the primary dysregulated IFN␣-producing cell subset in lupus, although this suggestion has been based predominantly on correlative data (for review, see ref. 29). Transgenic mice expressing DTR under control of either murine Siglec H or the human pDC-specific promoter BDCA-2 have been generated, and the specificity of DT-induced pDC depletion has been analyzed (13,30). Taking advantage of this system, we generated B6.Nba2.BDCA-2–DTR-Tg mice, with the goal of establishing the role of Siglec H–positive pDCs during initiation (4–12 weeks of age) and progression (12–18 weeks of age) of lupus-like disease. We observed a significant decrease in the number of serum antichromatin autoantibodies in mice of both age groups but limited effects on splenic cell abnormalities in the older group of mice. Although kidney morphology was largely unaffected by Siglec H–positive pDC depletion, IgG immune complex deposition in kidney glomeruli was significantly reduced in the older group of Siglec H–positive pDC-depleted B6.Nba2 mice. Thus, despite having minor effects on lymphocyte activation, Siglec H–positive pDC depletion affected both autoantibody production and renal pathology in mice in which disease development was already ongoing, making a pDC-targeting approach clinically relevant. Studies investigating the effect of Siglec H–positive pDC depletion in older female (B6.Nba2 ⫻ NZW)F1 mice, in which renal failure develops in an IFN␣-dependent manner, are ongoing (19,31). Although the main product of Siglec H–positive pDCs has been shown to be IFN␣, the total pDC population (including Siglec H–positive and Siglec H– negative pDCs) is known to produce a variety of cytokines and chemokines that potentially affect the pathogenesis of lupus (32–34). For example, pDCs have been shown to produce IL-6 and IL-10, the levels of which are increased in patients with lupus (35), and both of which are involved in B cell differentiation into antibodyproducing plasma cells (25). We previously showed that B6.Nba2 mouse–derived pDCs spontaneously secrete IFN␣ and IL-10, thereby driving B cell differentiation in vitro (16); however, the relative contribution by Siglec H–positive and Siglec H–negative pDCs was not defined. Here, we confirmed the expression of all 3 cytokines in pDC subsets, showing that Siglec H–positive pDCs specifically produce IFN␣ and IL-6, while Siglec H–negative pDCs produce IL-10. Because long-term depletion of Siglec H–positive pDCs also affected the

Siglec H–POSITIVE pDCs DRIVE DISEASE IN B6.Nba2 MICE

1021

numbers of Siglec H–negative pDCs and cDCs (see Supplementary Figure 2, available on the Arthritis & Rheumatology web site at http://onlinelibrary.wiley.com/ doi/10.1002/art.38989/abstract), it remains to be determined whether the pathogenic effect of Siglec H– positive pDCs is uniquely and directly mediated by IFN␣, or whether altered levels of IL-6 and/or IL-10 also play a role. Corresponding with a reduction in pDCdependent expression of Ifna, Il6, and Il10, we observed significantly diminished GCs in the spleens of DTtreated DTR-Tg–positive B6.Nba2 mice. This effect was significant only in the 4–12-week-old group and correlated with reduced numbers of fDCs, Tfh cells, GC B cells, and plasma cells, and reduced intrasplenic levels of IL-21. Interestingly, although Tfh cell levels (and the number of anti-chromatin autoantibodies in serum) remained significantly reduced in DT-treated 12–18-weekold DTR-Tg B6.Nba2 mice, fDCs, GC B cells, plasma cells, and GCs were not affected by Siglec H–positive pDC depletion in these mice. Whether the lack of a full effect in older mice will translate into a lack of effect in proteinuric mice, or whether the significant reduction in the number of circulating antibodies and IgG immune complex deposition will persist will be important to measure in upcoming studies focusing on the therapeutic potential of pDC depletion during advanced disease. As we were preparing this manuscript, a new supporting study by Rowland and associates at Washington University (St. Louis, Missouri) was published (36). In their investigation of the effect of Siglec H– positive pDCs in the BXSB mouse model of lupus, those investigators also observed that depletion of Siglec H– positive pDCs in BXSB.BDCA-2–DTR-Tg mice abrogated disease development by reducing levels of IFN␣inducible gene transcripts (Mx1, Ifit1), decreasing the numbers of serum ANAs (anti-RNP), and preventing early activation and differentiation of T lymphocytes and B lymphocytes (36). While we studied the effect of constitutive pDC depletion during 6–8-week intervals (4–12 weeks of age or 12–18 weeks of age), Rowland et al depleted Siglec H–positive pDCs for 3 weeks (8–11 weeks of age) and studied the effect at either 11 weeks of age or 19 weeks of age (36). Although many of the significant effects observed following Siglec H–positive pDC depletion in 11-week-old BXSB mice were less pronounced in 19week-old mice, in which the Siglec H–positive pDC population had recovered, T cell activation, select serum autoantibody numbers, and renal involvement were clearly reduced even in the older mice (36).

These data thus showed that depletion of Siglec H–positive pDCs early in the course of disease development can have a modest long-term beneficial effect, although the data did not address whether depletion of Siglec H–positive pDCs at later stages of disease would similarly affect lupus pathogenesis. Our data for 18week-old DTR-positive B6.Nba2 mice suggest that even though the effect of Siglec H–positive pDC depletion is less pronounced in older mice than in younger mice, such an approach may still be applicable to SLE patients, because both serum antichromatin IgG levels and IgG immune complex deposition in kidney glomeruli were significantly reduced at the later time point. Nonetheless, early depletion of Siglec H–positive pDCs seems to be required for maximum beneficial effects on disease parameters in both BXSB and B6.Nba2 lupus-prone mice. In summary, depletion of Siglec H–positive pDCs prevented the initiation and limited the progression of disease in lupus-prone B6.Nba2 mice. Thus, Siglec H– positive pDCs represent a valid cellular target for intervention and future treatment of SLE. ACKNOWLEDGMENTS We thank Lauren Liegl, Joana Dimo, and Abhishek Trigunaite for their assistance with tissue processing and sectioning, Jennifer Powers for flow cytometry–based cell sorting, Dr. Judith Drazba for assisting with microscopy, and Dr. Xiaoxia Li for critically reading this manuscript. AUTHOR CONTRIBUTIONS All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Jørgensen had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study conception and design. Davison, Jørgensen. Acquisition of data. Davison. Analysis and interpretation of data. Davison, Jørgensen.

REFERENCES 1. Kotzin BL, West SG. Systemic lupus erythematosus. In: Rich R, Shearer WT, Kotzin BL, Schroeder Jr HW, editors. Clinical immunology: principles and practice. 2nd ed. London: Mosby; 2001. p. 60.1–60.24. 2. Bennett L, Palucka AK, Arce E, Cantrell V, Borvak J, Banchereau J, et al. Interferon and granulopoiesis signatures in systemic lupus erythematosus blood. J Exp Med 2003;197:711–23. 3. Baechler EC, Batliwalla FM, Karypis G, Gaffney PM, Ortmann WA, Espe KJ, et al. Interferon-inducible gene expression signature in peripheral blood cells of patients with severe lupus. Proc Natl Acad Sci U S A 2003;100:2610–5. 4. Blanco P, Palucka AK, Gill M, Pascual V, Banchereau J. Induction

1022

5.

6.

7. 8. 9.

10.

11. 12.

13.

14.

15. 16.

17. 18.

19.

20.

of dendritic cell differentiation by IFN-␣ in systemic lupus erythematosus. Science 2001;294:1540–3. Puga I, Cols M, Barra CM, He B, Cassis L, Gentile M, et al. B cell-helper neutrophils stimulate the diversification and production of immunoglobulin in the marginal zone of the spleen. Nat Immunol 2011;13:170–80. Palanichamy A, Bauer JW, Yalavarthi S, Meednu N, Barnard J, Owen T, et al. Neutrophil-mediated IFN activation in the bone marrow alters B cell development in human and murine systemic lupus erythematosus. J Immunol 2014;192. Marshak-Rothstein A. Toll-like receptors in systemic autoimmune disease. Nat Rev Immunol 2006;6:823–35. Kaplan MJ, Radic M. Neutrophil extracellular traps: double-edged swords of innate immunity. J Immunol 2012;189:2689–95. Blasius A, Vermi W, Krug A, Facchetti F, Cella M, Colonna M. A cell-surface molecule selectively expressed on murine natural interferon-producing cells that blocks secretion of interferon-␣. Blood 2004;103:4201–6. Zhang J, Raper A, Sugita N, Hingorani R, Salio M, Palmowski MJ, et al. Characterization of Siglec-H as a novel endocytic receptor expressed on murine plasmacytoid dendritic cell precursors. Blood 2006;107:3600–8. Blasius AL, Colonna M. Sampling and signaling in plasmacytoid dendritic cells: the potential roles of Siglec-H. Trends Immunol 2006;27:255–60. Blomberg S, Eloranta ML, Magnusson M, Alm GV, Ronnblom L. Expression of the markers BDCA-2 and BDCA-4 and production of interferon-␣ by plasmacytoid dendritic cells in systemic lupus erythematosus. Arthritis Rheum 2003;48:2524–32. Swiecki M, Gilfillan S, Vermi W, Wang Y, Colonna M. Plasmacytoid dendritic cell ablation impacts early interferon responses and antiviral NK and CD8(⫹) T cell accrual. Immunity 2010;33: 955–66. Drake CG, Rozzo SJ, Hirschfeld HF, Smarnworawong NP, Palmer E, Kotzin BL. Analysis of the New Zealand Black contribution to lupus-like renal disease: multiple genes that operate in a threshold manner. J Immunol 1995;154:2441–7. Gubbels MR, Jorgensen TN, Metzger TE, Menze K, Steele H, Flannery SA, et al. Effects of MHC and gender on lupus-like autoimmunity in Nba2 congenic mice. J Immunol 2005;175:6190–6. Jorgensen TN, Alfaro J, Enriquez HL, Jiang C, Loo WM, Atencio S, et al. Development of murine lupus involves the combined genetic contribution of the SLAM and Fc␥R intervals within the Nba2 autoimmune susceptibility locus. J Immunol 2010;184:775–86. Rozzo SJ, Allard JD, Choubey D, Vyse TJ, Izui S, Peltz G, et al. Evidence for an interferon-inducible gene, Ifi202, in the susceptibility to systemic lupus. Immunity 2001;15:435–43. Atencio S, Amano H, Izui S, Kotzin BL. Separation of the New Zealand Black genetic contribution to lupus from New Zealand Black determined expansions of marginal zone B and B1a cells. J Immunol 2004;172:4159–66. Jorgensen TN, Roper E, Thurman JM, Marrack P, Kotzin BL. Type I interferon signaling is involved in the spontaneous development of lupus-like disease in B6.Nba2 and (B6.Nba2 x NZW)F(1) mice. Genes Immun 2007;8:653–62. Gilliet M, Cao W, Liu YJ. Plasmacytoid dendritic cells: sensing nucleic acids in viral infection and autoimmune diseases. Nat Rev Immunol 2008;8:594–606.

DAVISON AND JØRGENSEN

21. Stohl W, Xu D, Metzger TE, Kim KS, Morel L, Kotzin BL. Dichotomous effects of complete versus partial class II major histocompatibility complex deficiency on circulating autoantibody levels in autoimmune-prone mice. Arthritis Rheum 2004;50:2227–39. 22. Victora GD, Nussenzweig MC. Germinal centers. Annu Rev Immunol 2012;30:429–57. 23. Zotos D, Coquet JM, Zhang Y, Light A, D’Costa K, Kallies A, et al. IL-21 regulates germinal center B cell differentiation and proliferation through a B cell-intrinsic mechanism. J Exp Med 2010;207:365–78. 24. Eto D, Lao C, DiToro D, Barnett B, Escobar TC, Kageyama R, et al. IL-21 and IL-6 are critical for different aspects of B cell immunity and redundantly induce optimal follicular helper CD4 T cell (Tfh) differentiation. PLoS One 2011;6:e17739. 25. Jego G, Palucka AK, Blanck JP, Chalouni C, Pascual V, Banchereau J. Plasmacytoid dendritic cells induce plasma cell differentiation through type I interferon and interleukin 6. Immunity 2003;19: 225–34. 26. Colonna M, Trinchieri G, Liu YJ. Plasmacytoid dendritic cells in immunity. Nat Immunol 2004;5:1219–26. 27. Preble OT, Black RJ, Friedman RM, Klippel JH, Vilcek J. Systemic lupus erythematosus: presence in human serum of an unusual acid-labile leukocyte interferon. Science 1982;216:429–31. 28. Kirou KA, Lee C, George S, Louca K, Peterson MG, Crow MK. Activation of the interferon-␣ pathway identifies a subgroup of systemic lupus erythematosus patients with distinct serologic features and active disease. Arthritis Rheum 2005;52:1491–503. 29. Cao W. Pivotal functions of plasmacytoid dendritic cells in systemic autoimmune pathogenesis. J Clin Cell Immunol 2014;5:212. 30. Swiecki M, Wang Y, Riboldi E, Kim AH, Dzutsev A, Gilfillan S, et al. Cell depletion in mice that express diphtheria toxin receptor under the control of SiglecH encompasses more than plasmacytoid dendritic cells. J Immunol 2014;192:4409–16. 31. Jorgensen TN, Thurman J, Izui S, Falta MT, Metzger TE, Flannery SA, et al. Genetic susceptibility to polyI:C-induced IFN␣/ ␤-dependent accelerated disease in lupus-prone mice. Genes Immun 2006;7:555–67. 32. Sozzani S, Allavena P, D’Amico G, Luini W, Bianchi G, Kataura M, et al. Differential regulation of chemokine receptors during dendritic cell maturation: a model for their trafficking properties. J Immunol 1998;161:1083–6. 33. Birmachu W, Gleason RM, Bulbulian BJ, Riter CL, Vasilakos JP, Lipson KE, et al. Transcriptional networks in plasmacytoid dendritic cells stimulated with synthetic TLR 7 agonists. BMC Immunol 2007;8:26. 34. Piqueras B, Connolly J, Freitas H, Palucka AK, Banchereau J. Upon viral exposure, myeloid and plasmacytoid dendritic cells produce 3 waves of distinct chemokines to recruit immune effectors. Blood 2006;107:2613–8. 35. Chun HY, Chung JW, Kim HA, Yun JM, Jeon JY, Ye YM, et al. Cytokine IL-6 and IL-10 as biomarkers in systemic lupus erythematosus. J Clin Immunol 2007;27:461–6. 36. Rowland SL, Riggs JM, Gilfillan S, Bugatti M, Vermi W, Kolbeck R, et al. Early, transient depletion of plasmacytoid dendritic cells ameliorates autoimmunity in a lupus model. J Exp Med 2014;211: 1977–91.

Copyright of Arthritis & Rheumatology is the property of John Wiley & Sons, Inc. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.