IRF4 Deficiency Abrogates Lupus Nephritis Despite Enhancing Systemic Cytokine Production

BASIC RESEARCH www.jasn.org

Maciej Lech,* Marc Weidenbusch,* Onkar P. Kulkarni,* Mi Ryu,* Murthy Narayana Darisipudi,* Heni Eka Susanti,* Hans-Willi Mittruecker,†‡ Tak W. Mak,‡ and Hans-Joachim Anders*

*Department of Nephrology, Medizinische Poliklinik, University of Munich, Munich, Germany; †Institute for Immunology, University Medical Center Hamburg-Eppendorf, Hamburg, Germany; and ‡The Campbell Family Institute for Breast Cancer Research, Ontario Cancer Institute, Princess Margaret Hospital, Toronto, Ontario, Canada

ABSTRACT The IFN-regulatory factors IRF1, IRF3, IRF5, and IRF7 modulate processes involved in the pathogenesis of systemic lupus and lupus nephritis, but the contribution of IRF4, which has multiple roles in innate and adaptive immunity, is unknown. To determine a putative pathogenic role of IRF4 in lupus, we crossed Irf4-deficient mice with autoimmune C57BL/6-(Fas)lpr mice. IRF4 deficiency associated with increased activation of antigen-presenting cells in C57BL/6-(Fas)lpr mice, resulting in a massive increase in plasma BASIC RESEARCH levels of TNF and IL-12p40, suggesting that IRF4 suppresses cytokine release in these mice. Neverthe- less, IRF4 deficiency completely protected these mice from glomerulonephritis and lung disease. The mice were hypogammaglobulinemic and lacked antinuclear and anti-dsDNA autoantibodies, revealing the requirement of IRF4 for the maturation of plasma cells. As a consequence, Irf4-deficient C57BL/6- (Fas)lpr mice neither developed immune complex disease nor glomerular activation of complement. In addition, lack of IRF4 impaired the maturation of Th17 effector T cells and reduced plasma levels of IL-17 and IL-21, which are cytokines known to contribute to autoimmune tissue injury. In summary, IRF4 deficiency enhances systemic inflammation and the activation of antigen-presenting cells but also prevents the maturation of plasma cells and effector T cells. Because these adaptive immune effectors are essential for the evolution of lupus nephritis, we conclude that IRF4 promotes the development of lupus nephritis despite suppressing antigen-presenting cells.

J Am Soc Nephrol 22: 1443–1452, 2011. doi: 10.1681/ASN.2010121260

Systemic autoimmunity in systemic lupus erythem- crotic lymphocytes and expose them to the im- atosus (SLE) involves a polyclonal expansion of mune system.4 Another group of susceptibility lymphocytes that are autoreactive to multiple nu- genes enhances the immune recognition of self clear autoantigens. This process can cause a broad nucleic acids by Toll-like receptors (TLRs) in spectrum of clinical manifestations ranging from mild fever, skin rashes, and arthralgia to severe in- Received December 13, 2010. Accepted April 5, 2011. 1 flammation of kidney, lungs, or brain. The patho- M.L. and M.W. contributed equally to the manuscript. genesis of SLE is based on variable combinations of Published online ahead of print. Publication date available at genetic polymorphisms that promote loss of toler- www.jasn.org. ance or tissue inflammation.2,3 For example, certain Correspondence: Dr. Hans-Joachim Anders, Medizinische Po- genes impair lymphocyte apoptosis and the clear- liklinik, Universita¨t Mu¨ nchen, Pettenkoferstr. 8a, D-80336 ance of dying cells via opsonization, phagocytosis, Mu¨nchen, Germany. Phone: 49-89-218075855; Fax: 49-89- and digestion of self-DNA, which all increase the 51603379; E-mail: [email protected] release of nuclear particles from secondary ne- Copyright © 2011 by the American Society of Nephrology

J Am Soc Nephrol 22: 1443–1452, 2011 ISSN : 1046-6673/2208-1443 1443 BASIC RESEARCH www.jasn.org

dendritic cells (DCs), which increases the production of of IRF4 in innate and adaptive immunity, we hypothesized a type I IFN5,6 and eventually the expansion of autoreactive functional contribution of IRF4 to SLE and lupus nephritis. lymphocytes.7 A third class of genetic lupus risk factors af- To test this concept, we generated Irf4-deficient C57BL/6lpr/lpr fects tissue inflammation.4 (B6lpr) mice and compared the phenotype with that of wild- IFN-regulatory factors (IRFs) form a group of transcription type B6lpr mice, an autoimmune mouse strain that develops factors that have the potential to contribute to all of the afore- lupus autoantibodies and SLE manifestations in kidneys mentioned pathomechanisms of SLE. IRF-1 is a proinflamma- and lungs.27 tory transcription factor that triggers the expression of proinflammatory cytokines in tubular epithelial cells and immune cells in the postischemic kidney8 as well as in mesangial cells during lupus nephritis of MRL-(Fas)lpr mice.9 IRF3 and IRF7 mediate type I IFN production upon immune recognition of viral and endogenous nucleic acids in DCs,10 including TLR7 signaling, which was shown to be an essential pathway in SLE.7,11–13 IRF5 is required for immune cell maturation and for TLR signaling, two mechanisms that contribute to SLE and lupus nephritis of Fc␥RIIBϪ/ϪYaa or Fc␥RIIBϪ/Ϫ mice14 and in lupus secondary to pristane injection.15 Unlike other IRFs, IRF4 is not regulated by IFNs and its expression is restricted to immune cells.16 IRF4 has multiple regulatory functions in adaptive immunity.16,17 For example, IRF4 is required for the maturation of B and T cells,18 plasma cell maturation and Ig isotype switching,19 the ability of regulatory T cells to suppress Th2 responses,20 and the induction of Th17 T cells.21,22 IRF4 also regulates innate immunity because it is required for M2 macrophage polarization.23 IRF4 also serves as an inhibitor of TLR signaling via binding to MyD88, which impairs its interaction with IRF5 and other downstream signaling elements.24 For example, Irf4-deficient mice develop an exag- gerated postischemic inflammatory re- sponse aggravating ischemic acute renal failure, which depends on the oxidative stress-driven induction of IRF4 in intrare- nal DCs.25 As another example, bacterial products specifically induce IRF4 in DCs of the intestinal wall, a mechanism that protects mice from experimental colitis.26 Hence, IRF4 suppresses innate immunity but fosters adaptive immunity. These am- Figure 1. Lack of IRF4 increases the activation of antigen-presenting cells. (A) Serum bivalent immunoregulatory roles are again Ϫ Ϫ cytokine levels were determined by ELISA in B6lpr mice (black bars) and B6lpr/Irf4 / unique among the members of the IRF mice (white bars) at 6 months of age. (B) Spleen monocytes were stimulated with 1 family. ␮g/ml LPS and the cytokine levels were determined by ELISA. (C-E) Spleen cells were Given the clear roles of IRF1, IRF3, quantified by flow cytometry using surface activation markers as indicated. Data IRF5, and IRF7 in the pathogenesis of SLE represent mean Ϯ SEM from 12 to 14 mice in each group. *P Ͻ 0.05, **P Ͻ 0.01 versus and lupus nephritis and the multiple roles B6lpr mice.

1444 Journal of the American Society of Nephrology J Am Soc Nephrol 22: 1443–1452, 2011 www.jasn.org BASIC RESEARCH

RESULTS presenting cells, we isolated splenocytes from B6lpr and B6lpr/Irf4Ϫ/Ϫ mice and assessed cytokine production 24 IRF4 Deficiency Increases Serum Cytokine Levels by hours after LPS stimulation in vitro. In fact, splenocytes Ϫ Ϫ Activating DCs and Macrophages in B6lpr Mice from B6lpr/Irf4 / mice produced much higher levels of We generated Irf4-deficient B6lpr mice by crossing Irf4-deficient IFN␥, IL-12p40, TNF, and MCP-1/CCL2 as compared with with Fas-deficient (lpr) C57BL/6 mice. Litters of B6lpr/Irf4Ϫ/Ϫ mice cells from B6lpr mice (Figure 1B). were bred along Mendelian ratios and revealed no differences in Next we determined the activation state of various DC sub- body weight gain as compared with B6lpr (Irf4 wild-type) mice sets in spleens of 6-month-old mice by flow cytometry. IRF4 (not shown). In phenotyping both strains at 6 months of age, deficiency increased the total numbers of CD11c/CD4-positive we first determined serum levels of various proinflammatory DCs and CD4/CD8 double-negative CD11c DCs (Figure 1C). cytokines by ELISA. B6lpr/Irf4Ϫ/Ϫ mice revealed significantly IRF4 was required to suppress DC activation because B6lpr/Irf4Ϫ/Ϫ higher serum levels of TNF and IL-12p40 than their age- mice displayed higher numbers of CD40-positive CD11c DCs and matched B6lpr counterparts (Figure 1A), suggesting that IRF4 MHCII-positive F4/80/CD11c DCs (Figure 1D). In addition, suppresses systemic cytokine release in B6lpr mice. Because lack of IRF4 increased the numbers of CD11b/F4/80-positive IRF4 has been described to suppress the activation of antigen- macrophages in spleens; most of them were MHCII positive, indicating a classically activated (M1) macrophage phenotype (Figure 1E). Alternatively activated CD206-positive macrophages remained in a minor splenocyte population in 6-month-old B6lpr/Irf4Ϫ/Ϫ mice (Figure 1E). Taken together, lack of IRF4 increases systemic cytokine production and the activation of antigen-presenting cells in B6lpr mice, indicating an immunosuppressive role for IRF4 in these aspects of innate immunity.

Lack of IRF4 Prevents Lupus Nephritis in B6lpr Mice Next we questioned whether the aggravated systemic inflammation in B6lpr/Irf4Ϫ/Ϫ mice is associated with an aggravation of lupus nephritis. At 6 months of age, B6lpr mice developed diffuse proliferative glomerulonephritis that was associated with diffuse mesangial matrix expansion, mesangial cell prolifera- tion, and occasional tuft adhesions (Figure 2). Tuft necrosis and crescent formation were absent. On immunostaining, diffuse proliferative glomerulonephritis was associated with ex- tensive glomerular IgM and IgG deposits presenting in mesangial (IgG) and capillary (IgM/IgG) staining patterns (Figure 2).

Figure 2. Lack of IRF4 abrogates renal pathology in 6-month-old Figure 3. Lack of IRF4 abrogates lupus nephritis in 6-month-old Ϫ Ϫ B6lpr mice. Renal sections from 6-month-old B6lpr and B6lpr/Irf4 / B6lpr mice. The indices of lupus nephritis disease activity and mice were stained with periodic acid–Schiff (PAS), anti-mIgM, chronicity were assessed by semiquantitative morphometry on anti-mIgG, or anti-complement factor C9 as indicated. Note that PAS-stained renal sections from 6-month-old mice of both mouse Ϫ Ϫ B6lpr but not B6lpr/Irf4 / mice develop diffuse proliferative glo- strains as described in Concise Methods. Plasma creatinine levels merulonephritis with IgM, IgG, and C9 deposits. Original magni- were determined at the same age. Data represent mean Ϯ SEM fication, ϫ400. from at least 12 mice per group.

J Am Soc Nephrol 22: 1443–1452, 2011 IRF4 in Lupus Nephritis 1445 BASIC RESEARCH www.jasn.org

Staining for complement factor 9 showed mainly mesangial positivity (Figure 2), al- together indicating diffuse proliferative lupus-like immune complex glomerulonephritis as the renal manifestation of spontaneous autoimmunity in B6lpr mice. Lack of IRF4 completely abrogated this phenotype because age-matched B6lpr/Irf4Ϫ/Ϫ mice developed hardly any glomerular Ig and complement deposits or glomerular abnormalities as shown by light microscopy (Figure 2). This was also evident by morphometrical assessment of the activity and chronicity Figure 4. Lack of IRF4 reduces renal chemokine mRNA expression in 6-month-old B6lpr mice. (A) Renal mRNA was isolated from 6-month-old B6lpr mice (black bars) and index for lupus nephritis (Figure 3). Some Ϫ Ϫ B6lpr/Irf4 / mice (white bars) and quantified by real-time reverse-transcriptase PCR. renal sections showed a slight increase in Ϯ lpr/Irf4Ϫ/Ϫ Data are expressed as mean of the ratio versus the respective 18S rRNA level SEM. activated mesangial cells in B6 (B) Mac2-positive macrophages were quantified in 20 glomeruli per section in both mice, but this was not a consistent finding. mouse strains at 6 months of age. Data represent mean Ϯ SEM from at least 12 mice Abnormalities of the vascular or tubuloin- per group. *P Ͻ 0.05 versus B6lprmice. terstitial compartment were absent in both mouse strains. Because of the moderate renal lesions, plasma creatinine levels were only mildly elevated in B6lpr mice and there was a nonsignificant trend toward lower levels in B6lpr/Irf4Ϫ/Ϫ mice (Figure 3). Albu- minuria was absent in both strains (not shown). The abrogated lupus nephritis phenotype in B6lpr/Irf4Ϫ/Ϫ mice was associated with lower mRNA expression levels of the proinflammatory chemokines CCL2, CCL5, CXCL2, and CXCL10 (Figure 4A), which correlated with a significant reduction of Mac2-positive glomerular macrophages Figure 5. Lack of IRF4 abrogates lung disease in 6-month-old B6lpr mice. (A) Lung Ϫ Ϫ (Figure 4B). Together, lack of IRF4 protects sections from 6-month-old B6lpr and B6lpr/Irf4 / mice were stained with PAS. Note that lpr Ϫ Ϫ B6 mice from lupus nephritis. B6lpr but not B6lpr/Irf4 / mice develop peribronchial and perivascular immune cell infitrates. Original magnification, ϫ100. (B) Mac2-positive glomerular cells were quan- Lack of IRF4 Prevents Autoimmune tified in 20 glomeruli per section. Data are mean cell counts per glomerulus Ϯ SEM of Lung Disease in B6lpr Mice 9 to 12 mice in each group. ***P Ͻ 0.001 versus B6lpr mice. Autoimmune lung disease is another man- pathogenesis of lupus nephritis; therefore, we next examined ifestation of SLE. At 6 months of age, B6lpr mice displayed focal the effect of the Irf4 genotype on lupus autoantibody forma- areas of peribronchial and perivascular lymphocyte infiltrates tion. At the age of 6 months, B6lpr mice displayed significant (Figure 5). Such infiltrates were not detected in age-matched Ϫ Ϫ hypergammaglobulinemia and antinuclear antibodies (ANAs) B6lpr/Irf4 / mice, indicating that lack of IRF4 protects B6lpr as well as autoantbodies specifically directed against double- mice not only from lupus nephritis but also from autoimmune stranded DNA (dsDNA) or the Smith antigen (Figure 6). By lung disease. contrast, IRF4-deficient B6lpr mice had low serum IgG levels, and ANAs, dsDNA, and Smith autoantibodies were absent IRF4 Is Required for the Production of Lupus (Figure 6). Thus, IRF4 is required for the production of lupus Autoantibodies in B6lpr Mice autoantibodies in B6lpr mice. Lupus nephritis is a manifestation of immune complex disease in systemic autoimmunity. Glomerular immune complex deposits can develop in situ when circulating autoantibodies bind Lack of IRF4 Reduces Plasma Cells in B6lpr Mice to nuclear particles that have deposited along the glomerular Serum IgG and pathogenic autoantibodies are derived from capillaries. Alternatively, circulating immune complexes get Ig-producing plasma cells. Given the low levels of circulating Ϫ Ϫ deposited along the glomerular filtration barrier.28 In any case, IgG and the absence of lupus autoantibodies in B6lpr/Irf4 / lupus autoantibodies represent an essential element in the mice, we performed spleen cell flow cytometry for CD138 and

1446 Journal of the American Society of Nephrology J Am Soc Nephrol 22: 1443–1452, 2011 www.jasn.org BASIC RESEARCH

of both genotypes (Figure 7B). We therefore conclude that lack of IRF4 abrogates lupus nephritis because B6lpr/Irf4Ϫ/Ϫ mice are plasma cell deficient and no longer pro- duce those autoantibodies that cause immune complex glomerulonephritis.

Lack of IRF4 Reduces Th1 and Th17 Effector T Cells in B6lpr Mice The various T cell subsets are essential reg- ulators of autoimmune disease. Regulatory T cells suppress the expansion of autoreactive T cells and autoantigen-specific Th1 or Th17 effector T cells that promote inflammatory tissue injury.29,30 We therefore quantified T cell subsets in the spleens of 6-month-old B6lpr and B6lpr/Irf4Ϫ/Ϫ mice by flow cytometry. The total number of CD3- positive T cells was not affected by the Irf4 genotype, but CD8 T cells were slightly reduced and CD4/CD25/Foxp3-positive “regulatory” T cells were increased in B6lpr/Irf4Ϫ/Ϫ versus B6lpr mice (Figure 8A). We identified Th1 and Th17 CD4 T cells by intracellular staining for IFN␥ or IL-17, respectively, and found that lack of IRF4 substantially reduced both of these subsets in B6lpr mice (Figure 8B). By contrast, IL-17 positivity in CD4/CD8 double-negative T cells was independent of the Irf4 genotype (not shown). Consistent with the latter finding, IL-17 and IL-21 were substantially reduced in the serum of 6-month-old B6lpr/Irf4Ϫ/Ϫ mice (Figure 8C). Together, IRF4 deficiency impairs the maturation of Th1 and Th17 T cells as well as that of autoantibody- producing plasma cells, which is associated with an abrogation of lupus nephritis in B6lpr mice. Figure 6. IRF4 deficient B6lpr mice lack hypergammaglobuminemia and autoantibody Ϫ Ϫ production. (A) B6lpr/Irf4 / (E) and B6lpr wild-type mice (F) were bled at the end of the study to determine serum levels of IgG, IgM, anti-Smith, and anti-dsDNA autoantibodies DISCUSSION by ELISA. (B) ANAs were detected by staining of Hep2 cells using plasma dilutions of 1:40 as described in Concise Methods. Note the homogenous nuclear staining pattern using IRFs contribute to SLE and lupus nephritis Ϫ Ϫ plasma from B6lprmice that was not detectable with plasma from B6lpr/Irf4 / mice. in various ways. IRF1 acts as a nonredundant transcription factor that promotes the ␬ light chains to quantify plasma cells in 6-month-old B6lpr and expression of many proinflammatory genes in immune and Ϫ Ϫ B6lpr/Irf4 / mice. Lack of IRF4 was associated with a drastic nonimmune cells.8,9 IRF3 and IRF7 mediate the expression of reduction of the absolute numbers of spleen plasma cells on type I IFNs upon activation of innate viral nucleic acid sensors flow cytometry (Figure 7A). By contrast, the total numbers of that, in lupus, can also be activated by endogenous nucleic mature B cells were not affected by the Irf4 genotype, although acids and lupus autoantigens.7,12 IRF5 is required for immune a shift from follicular- to marginal-zone B cells was observed in cell maturation and for TLR signaling, two mechanisms that B6lpr/Irf4Ϫ/Ϫ versus B6lpr mice (Figure 7A). The comparable contribute to human SLE and lupus nephritis of Fc␥RIIBϪ/ϪYaa numbers of mature B cells on flow cytometry were consistent or Fc␥RIIBϪ/Ϫ mice.14 Here we show that IRF4 contributes to with identical staining patterns for IgM-positive cells in mice SLE and lupus nephritis in a different manner. On one side,

J Am Soc Nephrol 22: 1443–1452, 2011 IRF4 in Lupus Nephritis 1447 BASIC RESEARCH www.jasn.org

tivation of DCs in postischemic kidneys, a process that limits local cytokine production and thereby prevents excessive renal pathology and acute renal failure.25 This immunosuppressive effect of IRF4 is related to its in- ducible expression in DCs, which blocks the interaction of IRF5 with the TLR adaptor MyD88 and thereby suppresses the local expression of NF␬B-dependent proinflammatory cytokines during bacterial or sterile forms of inflammation.24 Consistent with this concept, spleen DCs and macrophages in Irf4-deficient B6lpr mice were more activated and produced more proinflammatory cytokines as compared with their wild-type B6lpr counterparts. IRF4-deficient splenocytes also showed increased IFN-␥ production by natural killer (NK)1.1 cells, whereas IRF4 deficiency did not affect IFN-␥ producing T cells.31 This process should be sufficient to enhance autoimmunity and autoimmune tissue injury because enhanced antigen-pre- sentation and co-stimulation is usually sufficient to increase the expansion of autoreactive B and T cells in lupus.32 For example, lack of the TLR signaling inhibitor single immun- globulin IL-1R-related molecule was simi- larly associated with activated DCs and increased production of lupus autoantibodies and subsequent immune complex glomerulonephritis in B6lpr mice.33 In addition, we recently observed the same for IL-1 receptor- associated kinase M, a factor that suppresses TLR signaling in DCs at the kinase level (un- published observation). However, increased activation of antigen-presenting cells and mas- sively increased serum cytokine levels were not associated with more autoimmunity or more Figure 7. IRF4 is required for plasma cells’ maturation. Flow cytometry was used to severe lupus nephritis in B6lpr/Irf4Ϫ/Ϫ mice, determine the total number of distinct B and plasma cell subsets (A) in spleens of lpr lpr/Irf4Ϫ/Ϫ most likely because of the lack of autoimmunity 6-month-old B6 mice (black bars) and B6 mice (white bars). The histogram and immune complex disease. presents mean Ϯ SEM of 12 to 14 mice in each group. *P Ͻ 0.05, ***P Ͻ 0.001 versus The lack of lupus-like autoimmunity as B6lpr mice. (B) IgM immunostaining of spleens identifies mature B cell distribution in mice of both genotypes. Original magnification, ϫ100. indicated by the absence of ANAs and other lupus autoantibodies was the most promi- IRF4 suppresses innate immune recognition and therefore nent phenotype of B6lpr/Irf4Ϫ/Ϫ mice. Our analysis identified an IRF4 deficiency enhances the activation of antigen-presenting almost complete absence of plasma cells, which was also illus- cells including the production of NF␬B-dependent proinflamma- trated by hypogammaglobulinemia. IRF4 has a nonredundant tory cytokines. Although this process should enhance autoimmu- role in plasma cell maturation and Ig class switch recombina- nity and autoimmune tissue inflammation, Irf4-deficient tion while germinal center B cell formation remains intact.19 As B6lpr mice still remain protected from glomerulonephritis such, Irf4-deficient mice are generally unable to induce hu- because IRF4 has a nonredundant role in the maturation of moral immune responses to antigen exposure,18 which obvi- plasma cells and Th1 and Th17 effector T cells, which pro- ously includes the production of ANAs in experimental lupus. tects B6lpr mice from the evolution of autoimmunity and These data document that innate immune activation is neces- immune complex disease. sary but not sufficient to cause lupus nephritis because autoan- We have recently shown that IRF4 potently suppresses the ac- tibody-mediated immune complex disease is an essential com-

1448 Journal of the American Society of Nephrology J Am Soc Nephrol 22: 1443–1452, 2011 www.jasn.org BASIC RESEARCH

cells,18,31 especially for T cell priming toward Th17 cells, because IRF4 phosphorylation by ROCK2 is required for the synthesis of IL-17 Ϫ Ϫ and of IL-21.21 As such, B6lpr/Irf4 / mice lack the major effector T cell populations that are involved in autoimmune tissue injuries including lupus nephritis beyond the production of lupus autoantibodies by plasma cells. In summary, IRF4 has nonredundant biologic effects for evolution of immune complex glomerulonephritis like the other members of the IRF family, namely IRF1, IRF3, IRF5, and IRF7. However, the mechanisms by which IRF4 contributes to autoimmune tissue injury differ from those of the other IRF family members. Although IRF4 deficiency activates antigen-presenting cells and induces systemic inflammation, lack of IRF4 also severely impairs plasma cell maturation and subsequent autoantibody production as well as the maturation of Th1 and Th17 effector T cells. As a consequence, IRF4 deficiency protects from the evolution of autoimmunity and lupus-like immune complex glomerulonephritis. It is therefore very likely that these immunoregulatory roles of IRF4 contribute to other lupus disease models in a sim- ilar manner. Together, we conclude that IRF4 is essential for the development of lupus-like autoimmunity and immune complex glomerulonephritis despite its suppressive effect on innate immunity.

CONCISE METHODS

Animal Studies Irf4-deficient mice were generated, genotyped, and backcrossed to the C57BL/6J strain for ten generations as described previously.18 B6Irf4Ϫ/Ϫ and B6lpr mice (Charles River) were mated to generate B6lpr/Irf4Ϫ/ϩ mice, which were then mated Figure 8. Lack of IRF4 impairs the maturation of Th17 cells. Flow cytometry was used to among each other to generate B6lpr/Irf4ϩ/ϩ and determine (A) T cell subsets and IFN␥- and (B) IL-17-producing CD4 T cells in spleens of B6lpr/Irf4Ϫ/Ϫ mice as described.33 Littermate fe- lpr lpr/Irf4Ϫ/Ϫ 6-month-old B6 mice (black bars) and B6 mice (white bars). (C) Plasma IL-17 and males were used for all experimental procedures. IL-21 levels were determined by ELISA at 6 months of age. The histogram presents In each individual mouse, the genotype was as- Ϯ Ͻ Ͻ lpr mean SEM of 12 to 14 mice in each group. *P 0.05, **P 0.01 versus B6 mice. sured by PCR. Mice were housed in groups of five mice in sterile filter-top cages with a 12-hour dark/ Ϫ Ϫ ponent for the development of lupus nephritis. B6lpr/Irf4 / light cycle and unlimited access to autoclaved food and water. One cohort mice also lacked IFN␥-producing Th1 T cells and IL-17-pro- of mice was sacrificed at 24 weeks of age and one cohort was followed ducing Th17 T cells, both of which have been shown to con- until 52 weeks of age. All experimental procedures were performed ac- tribute to glomerulonephritis30 and lupus nephritis in lpr mu- cording to the German animal care and ethics legislation and were ap- tant lupus mice.34,35 IRF4 is required for the maturation of T proved by the local governmental authorities.

J Am Soc Nephrol 22: 1443–1452, 2011 IRF4 in Lupus Nephritis 1449 BASIC RESEARCH www.jasn.org

Flow Cytometry bodies, NUNC maxisorp ELISA plates were coated with poly-L- Anti-mouse CD3, CD4, CD8, and CD25 antibodies (BD Pharmingen, lysine (Trevigen, Gaithersburg, MD) and mouse dsDNA. After in- Heidelberg, Germany) were used to detect CD3ϩCD4-CD8- double- cubation with mouse serum, dsDNA-specific IgG and serum IgG negative T cells and CD4ϩCD25ϩ regulatory T cell populations in spleens. and serum IgM levels were detected by ELISA (Bethyl Laborato- Anti-CD11c was used to identify DCs, and the activation of CD11c-positive ries, Montgomery, TX). For anti-Smith antibodies, NUNC max- cells was assessed by co-staining for CD40 and MHCII (BD Pharmingen, isorp ELISA plates were coated with Smith antigen (Immunovi- Heidelberg, Germany). Anti-mouse B220, CD21, CD23, IgD, and IgM sion, Springdale, AR). A horseradish-peroxidase-conjugated goat antibodies (BD Pharmingen) were used to detect mature B cells anti-mouse IgG (Rockland, Gilbertsville, PA) was used for detec- (B220ϩIgDϩIgMϩ), marginal-zone B cells (B220ϩCD21highCD23low), tion. Horseradish-peroxidase-conjugated anti-mouse IgG was and follicular B cells (B220ϩCD21lowCD23high). Spleen B1 cells were iden- used as the secondary antibody. Anti-nuclear antibodies were de- tified as CD19highB220lowCD5ϩCD43ϩ and peritoneal B1 cells were identi- tected on Hep2 slides (1:50 diluted serum; BioRad Laboratories, fied as CD19highB220lowCD5ϩ. Anti-mouse CD19, CD5, and CD43 were Redmond, WA). Sections were mounted with Vectashield contain- procured from BD Biosciences. Plasma cells were identified using anti- ing DAPI (Vector Laboratories, Burlingame, CA). mouse antibodies for Ig ␬ light chain and CD138 (BD Pharmingen). To identify the macrophage population in spleen, we used anti-mouse F4/80, Real-Time Quantitative (TaqMan) CD11b, MHCII, and CD206 from BD Biosciences. Macrophages (F4/ Reverse-Transcriptase PCR 80ϩCD11bϩ), M1 macrophages (F4/80ϩCD11bϩMHCIIϩ), and M2 Real-time reverse-transcriptase PCR was performed on mRNA macrophages (F4/80ϩCD11bϩCD206ϩ) were identified as men- from mouse organs as described previously.33 The SYBR Green tioned. Anti-mouse IL17a and IFN␥ (from BD Biosciences) were Dye detection system was used for quantitative real-time PCR on a used along with anti-mouse CD3, CD4, CD8, and NK1.1 antibod- Light Cycler 480 (Roche, Mannheim, Germany). All of the techni- ies to evaluate Th1, TH17, and NK cells producing IFN␥. Respec- cal steps were performed according to the Minimum Information tive isotype antibodies were used to demonstrate specific staining for Publication of Quantitative Real-Time PCR Experiments of cell subpopulations.36 Respective isotype antibodies were used guidelines.41 Controls consisting of double-deionized water were to demonstrate specific staining of cell subpopulations. Quantifi- negative for target and housekeeper genes. Gene-specific primers cation of cell number was done using counting beads for FACS (300 nM, Metabion, Martinsried, Germany) were designed and (Invitrogen, Carlsbad, CA). further analyzed in silico to target all known possible transcripts of interest. Primers are listed in Table 1. Evaluation of Autoimmune Tissue Injury Lungs, spleens, livers, lymph nodes, and kidneys from all mice were fixed Statistical Analysis in 10% buffered formalin, processed, and embedded in paraffin. Sections One-way ANOVA followed by post hoc Bonferroni’s test was used (2 ␮m) for periodic acid–Schiff stains were prepared following routine for multiple comparisons using GraphPad Prism, version 4.03. protocols.37 The severity of the renal lesions was graded using the activity Single groups were compared by unpaired two-tailed t test. Data and chronicity indices for human lupus nephritis as described.38 were expressed as mean Ϯ SEM. Statistical significance was as- Autoimmune lung injury was scored semiquantitatively (0 to 4) by sumed at P Ͻ 0.05. assessing the extent of peribronchial, perivascular, or interstitial lymphocyte infiltrates as described.39 For Table 1. Primers used for real-time RT-PCR IgM staining of the spleen sections, anti-mouse IgM- Accession Gene Sequence ␮-chain specific antibodies (Vector, Burlingame, Number CA) were used. IRF4 NM013674 Forward primer: 5Ј-CAAAGCACAGAGTCACCTGG-3Ј Complement component C9 antibody (kindly Reverse primer: 5Ј-TGCAAGCTCTTTGACACACA-3Ј provided from Mohamed R. Daha, University of CCL2/MCP-1 NM011333 Forward primer: 5Ј-CCTGCTGTTCACAGTTGCC-3Ј Leiden, The Netherlands) was used at a 1:50 dilu- Reverse primer: 5Ј-ATTGGGATCATCTTGCTGGT-3´ tion, and secondary goat anti-rabbit biotinylated CXCL2/MIP2 NM009140 Forward primer: 5Ј-CGGTCAAAAAGTTTGCCTTG-3Ј antibody (Vector, Burlingame, CA) was used at a Reverse primer: 5Ј-TCCAGGTCAGTTAGCCTTGC-3Ј Ј Ј 1:300 dilution. Serum cytokine levels were deter- CCL5 NM013653 Forward primer: 5 -GTGCCCACGTCAAGGAGTAT-3 Ј Ј mined by ELISA following the manufacturer’s Reverse primer: 5 -CCACTTCTTCTCTGGGTTGG-3 CXCL10 NM009140 Forward primer: 5Ј-ATGGATGGACAGCAGAGAGC-3Ј protocols (IL-12p40: OptEiA, BD, Heidelberg, Reverse primer: 5Ј-GGCTGGTCACCTTTCAGAAG-3Ј Germany; TNF: BioLegend, San Diego, CA; IL-4: IFN-␥ NM008337 Forward primer: 5Ј-ACAGCAAGGCGAAAAAGGAT-3Ј OptEiA, BD, Heidelberg, Germany). Plasma cre- Reverse primer: 5Ј-ACAGCAAGGCGAAAAAGGAT-3Ј atinine levels were determined using a commercial TNF-␣ NM011609 Forward primer: 5Ј-CCACCACGCTCTTCTGTCTAC-3Ј assay kit (DiaSys). Reverse primer: 5Ј-AGGGTCTGGGCCATAGAACT-3Ј IL-6 NM031168 Forward primer: 5Ј-TGATGCACTTGCAGAAAACA-3Ј Ј Ј Autoantibody Analysis Reverse primer: 5 -ACCAGAGGAAATTTTCAATAGGC-3 Serum antibody levels were determined by 18s RNA NR003278 Forward primer: 5Ј-GCAATTATTCCCCATGAACG-3Ј Ј Ј ELISA as described.40 For anti-dsDNA anti- Reverse primer: 5 -AGGGCCTCACTAAACCATCC-3

1450 Journal of the American Society of Nephrology J Am Soc Nephrol 22: 1443–1452, 2011 www.jasn.org BASIC RESEARCH

ACKNOWLEDGMENTS IFN regulatory factor 5 is required for disease development in the FcgammaRIIBϪ/ϪYaa and FcgammaRIIBϪ/Ϫ mouse models of systemic lupus erythematosus. J Immunol 184: 796–806, 2010 The work was supported by a grant from the Deutsche Forschungs- 15. Savitsky DA, Yanai H, Tamura T, Taniguchi T, Honda K: Contribution of gemeinschaft (AN372/11-1 and GRK 1202) to H.J.A. The expert tech- IRF5 in B cells to the development of murine SLE-like disease through nical assistance of Dan Draganovic and Janina Mandelbaum is grate- its transcriptional control of the IgG2a locus. Proc Natl Acad Sci U S A fully acknowledged. Parts of this work were performed as a medical 107: 10154–10159, 2010 thesis project by M.W. at the Medical Faculty of the University of 16. Tamura T, Yanai H, Savitsky D, Taniguchi T: The IRF family transcription factors in immunity and oncogenesis. Annu Rev Immunol 26: Munich. 535–584, 2008 17. Shaffer AL, Emre NC, Romesser PB, Staudt LM: IRF4: Immunity. Ma- lignancy! Therapy? Clin Cancer Res 15: 2954–2961, 2009 18. Mittrucker HW, Matsuyama T, Grossman A, Kundig TM, Potter J, DISCLOSURES Shahinian A, Wakeham A, Patterson B, Ohashi PS, Mak TW: Require- None. ment for the transcription factor LSIRF/IRF4 for mature B and T lymphocyte function. Science 275: 540–543, 1997 19. Klein U, Casola S, Cattoretti G, Shen Q, Lia M, Mo T, Ludwig T, Rajewsky K, Dalla-Favera R: Transcription factor IRF4 controls plasma REFERENCES cell differentiation and class-switch recombination. Nat Immunol 7: 773–782, 2006 1. Kotzin BL: Systemic lupus erythematosus. Cell 85: 303–306, 1996 20. Zheng Y, Chaudhry A, Kas A, deRoos P, Kim JM, Chu TT, Corcoran L, 2. Goodnow CC: Multistep pathogenesis of autoimmune disease. Cell Treuting P, Klein U, Rudensky AY: Regulatory T-cell suppressor pro- 130: 25–35, 2007 gram co-opts transcription factor IRF4 to control T(H)2 responses. 3. Gregersen PK, Olsson LM: Recent advances in the genetics of auto- Nature 458: 351–356, 2009 immune disease. Annu Rev Immunol 27: 363–391, 2009 21. Biswas PS, Gupta S, Chang E, Song L, Stirzaker RA, Liao JK, Bhagat G, 4. Kanta H, Mohan C: Three checkpoints in lupus development: Central Pernis AB: Phosphorylation of IRF4 by ROCK2 regulates IL-17 and tolerance in adaptive immunity, peripheral amplification by innate IL-21 production and the development of autoimmunity in mice. J Clin immunity and end-organ inflammation. Genes Immun 10: 390–396, Invest 120: 3280–3295, 2010 2009 22. Brustle A, Heink S, Huber M, Rosenplanter C, Stadelmann C, Yu P, 5. Theofilopoulos AN, Baccala R, Beutler B, Kono DH: Type I interferons Arpaia E, Mak TW, Kamradt T, Lohoff M: The development of inflam- (alpha/beta) in immunity and autoimmunity. Annu Rev Immunol 23: matory T(H)-17 cells requires interferon-regulatory factor 4. Nat Im- 307–336, 2005 munol 8: 958–966, 2007 6. Hom G, Graham RR, Modrek B, Taylor KE, Ortmann W, Garnier S, Lee 23. Satoh T, Takeuchi O, Vandenbon A, Yasuda K, Tanaka Y, Kumagai Y, AT, Chung SA, Ferreira RC, Pant PV, Ballinger DG, Kosoy R, Demirci FY, Miyake T, Matsushita K, Okazaki T, Saitoh T, Honma K, Matsuyama T, Kamboh MI, Kao AH, Tian C, Gunnarsson I, Bengtsson AA, Rantapaa- Yui K, Tsujimura T, Standley DM, Nakanishi K, Nakai K, Akira, S: The Dahlqvist S, Petri M, Manzi S, Seldin MF, Ronnblom L, Syvanen AC, Jmjd3-Irf4 axis regulates M2 macrophage polarization and host re- Criswell LA, Gregersen PK, Behrens TW: Association of systemic sponses against helminth infection. Nat Immunol 11: 936–944, 2010 lupus erythematosus with C8orf13-BLK and ITGAM-ITGAX. N Engl 24. Negishi H, Ohba Y, Yanai H, Takaoka A, Honma K, Yui K, Matsuyama J Med 358: 900–909, 2008 T, Taniguchi T, Honda K: Negative regulation of Toll-like-receptor 7. Marshak-Rothstein A, Rifkin IR: Immunologically active autoantigens: signaling by IRF-4. Proc Natl Acad SciUSA102: 15989–15994, 2005 The role of toll-like receptors in the development of chronic inflam- 25. Lassen S, Lech M, Rommele C, Mittruecker HW, Mak TW, Anders HJ: matory disease. Annu Rev Immunol 25: 419–441, 2007 Ischemia reperfusion induces IFN regulatory factor 4 in renal dendritic 8. Wang Y, John R, Chen J, Richardson JA, Shelton JM, Bennett M, Zhou cells, which suppresses postischemic inflammation and prevents acute XJ, Nagami GT, Zhang Y, Wu QQ, Lu CY: IRF-1 promotes inflamma- renal failure. J Immunol 185: 1976–1983, 2010 tion early after ischemic acute kidney injury. J Am Soc Nephrol 20: 26. Watanabe T, Asano N, Murray PJ, Ozato K, Tailor P, Fuss IJ, Kitani A, 1544–1555, 2009 Strober W: Muramyl dipeptide activation of nucleotide-binding oli- 9. Reilly CM, Olgun S, Goodwin D, Gogal RM Jr, Santo A, Romesburg gomerization domain 2 protects mice from experimental colitis. J Clin JW, Ahmed SA, Gilkeson GS: Interferon regulatory factor-1 gene Invest 118: 545–559, 2008 deletion decreases glomerulonephritis in MRL/lpr mice. Eur J Immunol 27. Cohen PL, Eisenberg RA: Lpr and gld: Single gene models of systemic 36: 1296–1308, 2006 autoimmunity and lymphoproliferative disease. Annu Rev Immunol 9: 10. Yasuda K, Richez C, Maciaszek JW, Agrawal N, Akira S, Marshak- 243–269, 1991 Rothstein A, Rifkin IR: Murine dendritic cell type I IFN production 28. Glassock RJ: Human idiopathic membranous nephropathy—A mys- induced by human IgG-RNA immune complexes is IFN regulatory tery solved? N Engl J Med 361: 81–83, 2009 factor (IRF)5 and IRF7 dependent and is required for IL-6 production. 29. Sakaguchi S, Miyara M, Costantino CM, Hafler DA: FOXP3ϩ regula- J Immunol 178: 6876–6885, 2007 tory T cells in the human immune system. Nat Rev Immunol 10: 11. Hoshino K, Sugiyama T, Matsumoto M, Tanaka T, Saito M, Hemmi H, 490–500, 2010 Ohara O, Akira S, Kaisho T: IkappaB kinase-alpha is critical for inter- 30. Summers SA, Steinmetz OM, Li M, Kausman JY, Semple T, Edgtton feron-alpha production induced by Toll-like receptors 7 and 9. Nature KL, Borza DB, Braley H, Holdsworth SR, Kitching AR: Th1 and Th17 440: 949–953, 2006 cells induce proliferative glomerulonephritis. J Am Soc Nephrol 20: 12. Anders HJ: Pseudoviral immunity—A novel concept for lupus. Trends 2518–2524, 2009 Mol Med 15: 553–561, 2009 31. Lohoff M, Mittrucker HW, Prechtl S, Bischof S, Sommer F, Kock S, 13. Anders HJ, Krug A, Pawar RD: Molecular mimicry in innate immunity? Ferrick DA, Duncan GS, Gessner A, Mak TW: Dysregulated T helper The viral RNA recognition receptor TLR7 accelerates murine lupus. Eur cell differentiation in the absence of interferon regulatory factor 4. J Immunol 38: 1795–1799, 2008 Proc Natl Acad SciUSA99: 11808–11812, 2002 14. Richez C, Yasuda K, Bonegio RG, Watkins AA, Aprahamian T, Busto P, 32. Ohnmacht C, Pullner A, King SB, Drexler I, Meier S, Brocker T, Voeh- Richards RJ, Liu CL, Cheung R, Utz PJ, Marshak-Rothstein A, Rifkin IR: ringer D: Constitutive ablation of dendritic cells breaks self-tolerance

J Am Soc Nephrol 22: 1443–1452, 2011 IRF4 in Lupus Nephritis 1451 BASIC RESEARCH www.jasn.org

of CD4 T cells and results in spontaneous fatal autoimmunity. J Exp mer inhibition of CCL2/MCP-1 ameliorates lupus nephritis in MRL- Med 206: 549–559, 2009 (Fas)lpr mice. J Am Soc Nephrol 18: 2350–2358, 2007 33. Lech M, Kulkarni OP, Pfeiffer S, Savarese E, Krug A, Garlanda C, 38. Allam R, Pawar RD, Kulkarni OP, Hornung V, Hartmann G, Segerer S, Mantovani A, Anders HJ: Tir8/Sigirr prevents murine lupus by sup- Akira S, Endres S, Anders HJ: Viral 5Ј-triphosphate RNA and non-CpG pressing the immunostimulatory effects of lupus autoantigens. J Exp DNA aggravate autoimmunity and lupus nephritis via distinct TLR- Med 205: 1879–1888, 2008 independent immune responses. Eur J Immunol, 38: 3487–3498, 2008 34. Balomenos D, Rumold R, Theofilopoulos AN: Interferon-gamma is 39. Sunderrajan EV, McKenzie WN, Lieske TR, Kavanaugh JL, Braun SR, required for lupus-like disease and lymphoaccumulation in MRL-lpr Walker SE: Pulmonary inflammation in autoimmune MRL/Mp-lpr/lpr mice. J Clin Invest 101: 364–371, 1998 mice. Histopathology and bronchoalveolar lavage evaluation. Am J 35. Kyttaris VC, Zhang Z, Kuchroo VK, Oukka M, Tsokos GC: Cutting Pathol 124: 353–362, 1986 edge: IL-23 receptor deficiency prevents the development of lupus 40. Patole PS, Pawar RD, Lichtnekert J, Lech M, Kulkarni OP, Raman- nephritis in C57BL/6-lpr/lpr mice. J Immunol 184: 4605–4609, 2010 janeyulu A, Segerer S, Anders HJ: Coactivation of Toll-like receptor-3 36. Kulkarni OP, Sayyed SG, Kantner C, Ryu M, Schnurr M, Sardy M, Leban and -7 in immune complex glomerulonephritis. J Autoimmun 29: J, Jankowsky R, Ammendola A, Doblhofer R, Anders HJ: 4SC-101, a 52–59, 2007 novel small molecule dihydroorotate dehydrogenase inhibitor, sup- 41. Bustin SA, Benes V, Garson JA, Hellemans J, Huggett J, Kubista M, presses systemic lupus erythematosus in MRL-(Fas)lpr mice. Am J Mueller R, Nolan T, Pfaffl MW, Shipley GL, Vandesompele J, Wit- Pathol 176: 2840–2847, 2010 twer CT: The MIQE guidelines: Minimum information for publica- 37. Kulkarni O, Pawar RD, Purschke W, Eulberg D, Selve N, Buchner K, tion of quantitative real-time PCR experiments. Clin Chem 55: Ninichuk V, Segerer S, Vielhauer V, Klussmann S, Anders HJ: Spiegel- 611–622, 2009

1452 Journal of the American Society of Nephrology J Am Soc Nephrol 22: 1443–1452, 2011