We previously established an IgG Fc receptor IIB (FcγRIIB)-deficient C57BL/6 (B6)-congenic mouse strain (KO1), which spontaneously develops rheumatoid arthritis (RA), but not systemic lupus erythematosus (SLE). Here, we show that when Y chromosome-linked autoimmune acceleration (Yaa) mutation was introduced in KO1 strain (KO1.Yaa), the majority of KO1.Yaa mice did not develop RA, but instead did develop SLE. This phenotype conversion did not depend on autoantibody specificity, since KO1.Yaa mice, compared with KO1, showed a marked increase in serum levels of both lupus-related and RA-related autoantibodies. The increase in frequencies of CD69+ activated B cells and T cells, and the spontaneous splenic GC formation with T follicular helper cell generation were manifest early in life of KO1.Yaa, but not KO1 and B6.Yaa, mice. Activated CD4+ T cells from KO1.Yaa mice showed upregulated production of IL-21 and IL-10, compared with the finding in KO1 mice, indicating the possibility that this aberrant cytokine milieu relates to the disease phenotype conversion. Thus, our model is useful to clarify the shared and the disease-specific mechanisms underlying the clinically distinct systemic autoimmune diseases RA and SLE.
IgG Fc receptor IIB (FcγRIIB) is a major negative regulator of BCR-mediated activation signals in B cells . We previously found that the Fcgr2b gene encoding FcγRIIB is polymorphic, and that autoimmune disease-prone mouse strains, such as NZB, BXSB, MRL, and NOD, all share deletion polymorphism in the AP-4-binding site in the Fcgr2b promoter region . Because of the defective AP-4 binding, mice with this autoimmune-type allele polymorphism show downregulation of FcγRIIB expression levels in activated GC B cells, resulting in upregulation of IgG auto-antibody production [3, 4]. These observations suggested that the autoimmune-type Fcgr2b confers the basis of susceptibility to autoimmune diseases. Consistent was our earlier finding that systemic lupus erythematosus (SLE) phenotypes in BXSB male mice carrying Y chromosome-linked autoimmune acceleration (Yaa) mutation were almost completely inhibited by the substitution of the autoimmune-type Fcgr2b for the wild C57BL/6 (B6)-type Fcgr2b . However, because BXSB female mice carrying the autoimmune-type Fcgr2b but lacking Yaa did not develop SLE, it is likely that the autoimmune-type Fcgr2b contributes to SLE susceptibility through a strong epistatic interaction with Yaa mutation.
To examine further the role of the downregulated expression of FcγRIIB in autoimmune diseases, we recently established an FcγRIIB-deficient B6 mouse strain, KO1, by gene targeting in 129-derived embrionic stem cells and selective backcrossing to a B6 background. Intriguingly, KO1 did not develop SLE, but instead developed severe rheumatoid arthritis (RA), as reported previously . This KO1 strain carried a 129-derived approximately 6.3 Mb interval distal from the null-mutated Fcgr2b gene within the Sle16 locus, which is shown to induce loss of self-tolerance in the B6 background . Boross et al.  reported that FcγRIIB-deficient B6 mice generated by gene targeting in B6-derived embryonic stem cells, thus lacking the 129-derived flanking Sle16 locus, fail to develop any sign of autoimmune diseases. Thus, the development of RA in KO1 mice may be due to the epistatic interaction of FcγRIIB deficiency and Sle16 locus.
Boross et al.  also reported that their FcγRIIB-deficient B6 mice develop lethal lupus nephritis in the presence of Yaa mutation, indicating the epistatsis between FcγRIIB-deficiency and Yaa in the development of full-blown autoimmune diseases. In addition, Subramanian et al.  reported that the strong epistatic interaction between Yaa and Sle1, which contains the autoimmune-predisposing Slam/Cd2 haplotype, contributes to severe lupus nephritis. The Sle16 locus also contains this autoimmune-predisposing Slam/Cd2 haplotype .
In contrast to the accelerated effect of Yaa on lupus nephritis, Jansson and Holmdahl  reported the suppressive effect of Yaa on collagen-induced arthritis. In the present study, we have introduced Yaa mutation into FcγRIIB-deficient RA-prone KO1 mice to examine how Yaa affects the disease phenotypes in these mice. We found that the majority of KO1.Yaa mice did not develop RA, but instead did develop severe SLE early in life, and that this phenotype conversion did not depend on the shift of autoantibody specificity from RA-related to lupus-related one.
Characteristic clinical features differ between RA and SLE; however, both diseases share aberrant activation of immune processes associated with the production of a variety of autoantibodies and subsequent immune complex-mediated tissue inflammation. Our model is useful to investigate the shared and the disease-specific factors contributing to the clinically distinct systemic autoimmune diseases RA and SLE.
Disease phenotype in Yaa-carrying FcγRIIB-deficient KO1 mice
KO1 mice developed arthritis after 4 months of age and the disease incidence and severity were increased with age. At 8 months of age, 67% of KO1 mice showed arthritis with marked swelling and stiffness of forepaws and hindpaws. In contrast, the incidence and severity of arthritis were markedly suppressed in KO1.Yaa mice and 88% of KO1.Yaa mice were free from arthritis (Fig. 1A). Representative macroscopic findings of forepaws and hindpaws of KO1 and KO1.Yaa mice at 8 months of age are shown in Figure 1B. Intriguingly, KO1 strain did not develop proteinuria; however, KO1.Yaa began to be positive for proteinuria at 2 months of age and the incidence of positive proteinuria reached 63% (Fig. 1C) with 46% mortality rate at 8 months of age (Fig. 1D).
Figure 1E shows a comparison of representative histopathological and immunofluorescent findings of renal glomeruli among B6, B6.Yaa, KO1, and KO1.Yaa mice at 4 months of age. In KO1.Yaa mice, glomeruli were significantly enlarged even at 4 months of age (Fig. 1F), because of a marked cellular proliferation in glomeruli and a large amount of IgG deposition in mesangial area and along glomerular capillary walls. These glomerular lesions were seldom observed in B6, B6.Yaa, and KO1 mice even at 8 months of age.
Serum levels of autoantibodies
To examine the relationship between the disease phenotype conversion from RA to SLE and the specificity of autoantibodies, we compared serum levels of lupus-related autoantibodies against dsDNA, chromatin, and RNP, and RA-related rheumatoid factor (RF), anti-type II collagen (CII), and -cyclic citrullinated peptide (CCP) antibodies at 2 and 6 months of age among B6, B6.Yaa, KO1, and KO1.Yaa mice (Fig. 2). KO1.Yaa mice showed higher serum levels of both lupus-related and RA-related autoantibodies than the other three strains of mice even at 2 months of age. The levels of all these antibodies were increased with age in KO1.Yaa mice. Age-associated increase was also observed in KO1 mice; however, the levels were remarkably higher in KO1.Yaa mice than those in KO1 mice at 6 months of age. Thus, the conversion of disease phenotype from RA to SLE was not explained by the shift of antibody specificity from RA-type to lupus-type.
Splenomegaly, subpopulation, and maturation/activation status of splenic lymphocytes
The spleen weight in B6, B6.Yaa, KO1, and KO1.Yaa mice was compared at 4 months of age. Splenomegaly was only observed in KO1.Yaa mice (Fig. 3A). Consistently, spontaneous GC formation was observed only in KO1.Yaa mice at 4 months of age (Fig. 3B).
Flow cytometric analysis of spleen cells from 4-month-old mice revealed that, while there were no differences in frequencies of B220+ B cells per total spleen cells among four strains of mice (Table 1), there was a significant decrease in frequencies of CD21+CD23− marginal zone B cells in Yaa-bearing B6.Yaa and KO1.Yaa mice (Fig. 4A and Table 1). This decrease is thought to be due to the effect of Yaa mutation, as reported previously , and not directly related to SLE phenotype. As for the activation/maturation status of B cells, frequencies of CD69+ activated B cells, peanut agglutinin (PNA)+ GC B cells, and CD138+ plasma cells were significantly higher in KO1.Yaa mice than those found in other strains of mice (Fig. 4B and Table 1). As for T cells, while total CD3+ T cells per total cells was significantly decreased in KO1.Yaa mice, the frequency of CD69+ activated T cells was markedly increased in KO1.Yaa mice (Fig. 4C and Table 1). This activation of T cells may reflect the increases in the CD4+/CD8+ T-cell ratio and in the frequency of TFH cells with PD1+ICOS+CXCR5+CD4+ phenotype (Fig. 4C and Table 1). Because the frequencies of PD1+ICOS+CXCR5+CD4+ TFH cells in B6, B6.Yaa, and KO1 mice were within normal range (Table 1), the observed abnormal increase in PD1+ICOS+CXCR5+CD4+ TFH cells in KO1.Yaa mice with overt SLE was thought to be due to the combined effect of the FcγRIIB-deficiency, Sle16 locus, and Yaa mutation. Table 1 also shows that the frequency of CD11b+ monocyte/macrophage population was significantly increased in KO1.Yaa mice with a comparable level observed in BXSB male mice .
Table 1. Subpopulations of splenocytes in KO1, KO1.Yaa, B6, and B6.Yaa mice at 4 months of agea
Results were obtained from six mice in each strain, and are shown as mean and SE.
The value is significantly different from B6 mice or KO1 mice (p < 0.005, Student's t-test).
The value is significantly different from other strains of mice (p < 0.05, Student's t-test).
Cytokine profile in spleen from KO1 and KO1.Yaa mice
To examine the difference in in vivo cytokine expression levels associated with phenotype conversion from RA to SLE, quantitative real-time PCR (qRT-PCR) analysis was performed to compare mRNA expression levels of notable cytokines in spleen between KO1 and KO1.Yaa mice at 4 months of age (Fig. 5A). The result showed that the expression of IL-6, IL-10, and IL-21 was significantly upregulated in KO1.Yaa mice compared with that in KO1 mice. Among these, the increase in IL-10 expression was prominent, with more than tenfold increase in KO1.Yaa mice. There was no significant difference in expression levels of other cytokines such as IL-2, IL-4, IL-17, IFN-γ, TNF-α, and IFN-α between two strains of mice.
We next examined the cellular source of IL-10 and IL-21, using flow cytometric analysis of PMA/ionomycin-stimulated spleen cells from 4-month-old KO1 and KO1.Yaa mice. Both IL-10 and IL-21 were secreted from CD4+ T cells and the frequencies of IL-10 and IL-21-secreting cells per total CD4+ T cells were significantly higher in KO1.Yaa than those in KO1 mice (mean ± SE of KO1 versus KO1.Yaa; IL-10: 7.56 ± 1.25 versus 14.74 ± 0.43, p < 0.01, IL-21: 5.09 ± 0.22 versus 9.91 ± 0.60, p < 0.01) (Fig. 5B), consistent with the results of qRT-PCR analysis. PD1 and ICOS expression levels were upregulated in in vitro stimulated CD4+ T cells. Most IL-10 and IL-21-secreting cells showed high PD1 expression levels; however, the ICOS expression level was broadly distributed in these cytokine-secreting cells (Fig. 5B). As shown in Figure 5C, in addition to IL-10 and IL-21 single producers, the significant frequency of CD4+ T cells secreted both cytokines.
The current study showed that introduction of Yaa mutation into RA-prone KO1 mice leads to conversion of disease phenotypes from RA to SLE. RA and SLE are both classified as systemic autoimmune diseases. Since features of RA are occasionally associated with the clinical pictures of SLE , it has long been suggested that certain shared genetic pathways, as well as disease-specific ones, underlie the pathogenesis of both RA and SLE . Our current model provided a clue to investigate this issue and suggested that, while the FcγRIIB deficiency and Sle16 locus in KO1 genetic background confers predisposition to RA , an additional epistatic effect of Yaa mutation induces conversion of the disease phenotype from RA to SLE.
It has been shown that an etiology of Yaa-mediated B-cell activation is the duplication of the Tlr7 gene [9, 15, 16]. The ligand for TLR7 is single-stranded RNA, thus suggesting that overexpression of TLR7 activates B cells by RNA-containing autoantigens, resulting in RNA-associated lupus autoantibody production. However, in the present study, Yaa-mediated disease phenotype conversion from RA to SLE was not explained by the shift of autoantibody specificity, and rather Yaa-mediated B-cell activation seems to be polyclonal in KO1.Yaa mice. This polyclonal B-cell activation may relate to the marked spontaneous GC formation and the TFH-cell generation that developed in the spleen early in life of KO1.Yaa mice. The formation of GC depends on intrafollicular localization of antigen, activated B cells and T cells [17, 18]. Among subsets of CD4+ T cells, TFH cells are the specialized subset to help B cells to generate affinity-matured antibodies . In addition to the B-cell help by TFH cells in GC reaction, it has been shown that the relationship between B cells and TFH cells is a reciprocal dependency, and that the cognate interaction with activated B cells is required for the maintenance of PD-1+ICOS+CXCR5+ TFH cells . This is consistent with the present study, in which the combined effect of FcγRIIB-deficiency, Sle16 locus, and Yaa mutation accelerated not only spontaneous PNA+ B-cell generation and GC formation but also TFH-cell generation in KO1.Yaa mice. As this vicious cycle of activated B cells and TFH cells promotes polyclonal B-cell activation, KO1.Yaa mice showed the marked increase in serum levels of both lupus-related and RA-related autoantibodies.
Anti-CCP antibodies are currently considered to be the most specific autoantibodies for RA patients, although some patients with SLE and Sjögren's syndrome were found to have these autoantibodies . Anti-CCP antibodies react with citrullinated proteins, which are the product of posttranslational modification. Citrullination of protein is a physiological process and is catalyzed by peptidyl arginine deiminase enzymes. Anti-CCP antibodies may thus gain the arthritogenicity when citrullinated proteins are increased, particularly in the arthritic region . In mouse models, an increased serum level of anti-CCP antibodies was observed in SLE-prone and arthritis-free bcl-2-transgenic (NZW × B6)F1 mice , as in the case of KO1.Yaa mice in the current study. Thus, it appears that this autoantibody specificity is not exclusively associated with inflammatory joint diseases.
In KO1.Yaa mice, there was a significant increase in the IL-21 expression level early in life compared with that in KO1 mice. IL-21 is a potent immunoregulatory cytokine produced by NKT cells and CD4+ T cells, and it has recently been shown that IL-21 is an autocrine growth factor for TFH cells [17, 22]. Many cell types express the receptors for IL-21, but the level of expression on B cells is the highest and drives terminal differentiation of B cells and plasma cells [19, 22]. Intriguingly, Bubier et al.  reported that SLE phenotypes including autoantibody production in BXSB male mice were almost completely inhibited in the mice with the deficient IL-21 receptor. Furthermore, Rankin et al.  recently reported that IL-21 receptor-deficient MRL/lpr mice were devoid of abnormal systemic accumulation of activated B cells and T cells. These findings suggest that IL-21-mediated signals play an essential role for the pathogeneses of SLE.
It has been shown that IL-21 is a potent regulator of IL-10, since IL-10 production decreases in IL-21 receptor knockout mice, while it increases in IL-21-transgenic mice . IL-10 was first described as a factor produced by Th2 cells, which inhibited the production of cytokines by Th1 cells . Accumulating evidence, however, have shown that IL-10 is actually produced by many types of cells, and that, although IL-10 shows antiinflammatory properties against T cells and macrophages through inhibiting the production of inflammatory cytokines, it promotes B-cell function to induce antibody production [27, 28]. Considering these dual effects with immunosuppressive and immunostimulatory properties, IL-10 may confer different effects on the disease progression processes of RA and SLE. Indeed, the hallmark of RA is the excess production of inflammatory cytokines by T cells and macrophages at inflammatory foci, while SLE is characterized by increased production of high-affinity autoantibodies and deposition of their immune complexes in a wide variety of tissues, particularly in renal glomeruli. Consistently, there are several reports indicating the suppressive effect of IL-10 on RA [27, 29] and the promoting effect of IL-10 on lupus pathogenesis . Further studies are needed to define the role of IL-10 in the conversion of disease phenotypes observed in the present study. These studies are underway in our laboratory.
Peripheral blood mononuclear cells from patients with active SLE show up-regulated expression of a group of type I IFN-induced genes [31-33]. Thus, IFN-α seems to be an important cytokine in SLE pathogenesis. In pristane-induced lupus model, the disease was shown to be associated with excess IFN-α production , as in the case of human SLE. However, overexpression of IFN-α is not likely to be involved in SLE pathogenesis in KO1.Yaa mice, since there were no differences in IFN-α expression levels between RA-prone KO1 mice and SLE-prone KO1.Yaa mice. Accumulating evidence shows that IL-6, IL-17, and TNF-α are important contributing cytokines to the pathogenesis of RA [35-37]. In the present studies, however, IL-6 expression levels were increased in SLE-prone KO1.Yaa mice compared with those in RA-prone KO1 mice, and there was no significant difference in expression levels of IL-17 and TNF-α between KO1 and KO1.Yaa mice. Thus, these cytokines are suggested to be unrelated to the observed phenotype conversion from RA to SLE in our model.
In conclusion, we introduced Yaa mutation into RA-prone KO1 strain and found that the disease phenotype converted from RA to SLE in KO1.Yaa mice. This phenotype conversion was likely to be due to the changes in cytokine milieus rather than the shift of autoantibody specificity from RA-related to lupus-related one. Further studies for the clarification and identification of the mechanism underlying this phenotype conversion are of paramount importance for shedding light on the mechanisms that control the development of clinically distinct systemic autoimmune diseases RA and SLE.
Materials and methods
FcγRIIB-deficient KO1 mice were generated by gene targeting in 129-derived embryonic stem cells and by backcrossing to B6 for over 12 generations . The Yaa mutation was introduced into KO1 mice by crossing with B6.Yaa mice. B6.Yaa mice were purchased from the Jackson Laboratory. All mice were housed under identical conditions. Experiments were performed in accordance with our institutional guidelines. Male mice were analyzed in the current study.
Incidence of arthritis
Ankle joint swelling was examined by inspection and arbitrarily scored as follows: 0, no swelling; 1, mild swelling; 2, moderated swelling; 3, severe swelling. Scores of both ankle joints are put together, and mice with scores over 2 were considered positive for arthritis.
Measurement of proteinuria
The proteinuria was monitored by biweekly testing and scored as previously descrobed . Briefly, urine samples (10 μL) spotted on filter paper were air dried, fixed in 70% ethanol and stained with bromophenol blue solution. A series of standard three-fold dilution of BSA were processed as the same way, and the degree of proteinuria was assessed by visually comparing the color intensity of urine spot with that of the spot of BSA standards. Scores are as follows; 0:<37 mg/100 mL 1:≧37 mg/100 mL, 2:≧74 mg/mL, 3:≧111 mg/100 mL, 4:≧333 mg/100 mL, 5:≧1000 mg/100 mL, and 6:≧3000 mg/100 mL. Mice with urinary protein levels of four or more in repeated tests were considered as positive for proteinuria.
Histopathology and tissue immunofluorescence
Tissues fixed in 4% paraformaldehyde and embedded in paraffin were sectioned at 2 μm thickness, and tissue sections were stained with periodic acid-Schiff and hematoxylin (PAS). For immunofluorescence, tissues were embedded in Tissue-Tek OCT compound, frozen in liquid nitrogen, and sectined at 4 μm thickness. Frozen kidney sections were stained with FITC-labeled polyclonal goat anti-mouse IgG for 60 min at room temperature. For analysis of splenic tissues, frozen sections were three-color stained with Alexa 488-labeled anti-CD4 and -CD8 mAbs, Alexa 647-labeled anti-B220 mAb, and Alexa 546-labeled PNA. Antibodies and PNA were purchased from BD Pharmingen (San Diego, CA) and Vector Laboratories Inc. (Burlingame, CA), respectively. The labeling of these reagents was performed in our laboratory. Color images were obtained using laser scanning microscopy (Zeiss LSM510, Carl Zeiss Co., Ltd., Germany).
Estimation of the severity of glomerular lesion
The extent of cellular proliferation in glomerular lesion was estimated by the measurement of glomerular size. Kidney section stained by PAS was photographed under a microscope (Biozero, KEYENCE, Osaka, Japan) with ×50 magnification. Ten glomeruli in each field were randomly selected in order of size, and the size of each glomerulus was calculated using BZ-II analyzer software (KEYENCE). Mean size of 10 glomeruli was used as an indicator of histological severity of lupus nephritis in each individual mouse.
Serum levels of autoantibodies
Serum levels of IgG anti-dsDNA and -chromatin antibodies were measured using ELISA, as previously described . Serum antibody levels are expressed in units, referring to a standard curve obtained by the serial dilution of a pooled serum of (NZB × NZW) F1 mice over 8 months, containing 1000 units/mL. Serum levels of IgG anti-RNP antibodies were measured by employing a commercially available kit (Alpha Diagnostic Intl. Inc., San Antonio, TX), and are expressed as relative units according to the manufacturer's instructions.
Serum levels of IgG RF and IgG anti-CCP antibodies were measured employing commercially available kits (Shibayagi Co. Ltd., Gunma, Japan and Cosmic Corporation, Tokyo, Japan, respectively), and are expressed as relative units according to the manufacturer's instructions. Serum levels of IgG anti-CII antibodies were measured using an ELISA plate precoated with bovine CII (Sigma-Aldrich, St. Louis, MO). CII-binding activities are expressed in units, referring to a standard curve obtained by serial dilution of a standard serum pool from KO1 mice hyper-immunized with CII, containing 1000 unit activities/mL.
Flow cytometric analysis
For the analysis of splenic lymphocytes, spleen cells were stained with the following reagents: FITC-conjugated anti-CD3, -CD21, -ICOS, -Foxp3, and -CD11b mAbs, Pacific blue-conjugated anti-B220, -CD4 mAbs, and PNA, PE-conjugated anti-B220, -CD138, -PD1, and -CD25 mAbs, and biotin-conjugated anti-CD69, -CD23, -CD8, and -CXCR5 mAbs, followed by streptavidin allophycocyanin. mAbs for CD25 and those for Foxp3, PD1, and ICOS were obtained from BioLegend (San Diego, CA) and eBioscience (San Diego, CA), respectively. Others were from BD PharMingen. Stained cells were four-color analyzed using a FACSAria cytometer and FlowJo software (Tree Star, Inc., Ashland, OR) with whole cell-gate excluding dead cells in forward and side scatter cytogram.
For intracellular cytokine staining of spleen cells, cells were stimulated with PMA (0.2 μg/mL)/ionomycin (2 μg/mL) in the presence of Golgi-Stop (BD Bioscience, San Jose, CA) for 5 h and stained with Pacific Blue-labeled anti-CD4 and biotin-labeled anti-PD1 or anti-ICOS mAbs followed by streptavidine allophycocyanin. Stained cells were then fixed and permeabilized using BD Cytofix/Cytoperm (BD Bioscience), followed by staining with FITC-labeled anti-IL-10, and PE-labeled anti-IL-21 mAbs. Stained cells were analyzed as above.
Total RNA was isolated from spleen and first-stranded cDNA was synthesized using an oligo(dT)-primer with Superscript II First-Strand Synthesis kit (Invitrogen, Carlsbad, CA). The cDNA product was used for each qRT-PCR sample. The data were normalized to β-actin reference. Primer pairs used were as follows: IL-2 (forward) 5′-AACCTGAAACTCCCCAGGAT-3′, (reverse) 5′-AGGGCTT GTTGAGATGATGC-3′; IL-4 (forward) 5′-CCTCACAGCAACGAA GAACA-3′, (reverse) 5′-AAGTTAAAGCATGGTGGCTCA-3′; IL-6 (forward) 5′-GACAAAGCCAGAGTCCTTCAGAGAG-3′, (reverse) 5′-CTAGGTTTGCCGAGTAGATCTC-3′; IL-10 (forward) 5′-CCAA GCCTTATCGGAAATGA-3′, (reverse) 5′-TGGCCTTGTAGACACCT TGG-3′; IL-17 (forward) 5′-TCTCTGATGCTGTTGCTGCT-3′, (reverse) 5′-GACCAGGATCTCTTGCTGGA-3′; IL-21 (forward) 5′-ATCCTGAACTTCTATCAGCTCCAC-3′, (reverse) 5′-GCATTTAGCT ATGTGCTTCTGTTTC-3′; IFNγ (forward) 5′-AAGACAATCAGGCC ATCAGC-3′, (reverse) 5′-ATCAGCAGCGACTCCTTTTC-3′; TNF-α (forward) 5′-GGCAGGTCTACTTTGGAGTCATTGC-3′, (reverse) 5′-ACATTCGAGGCTGCTCCAGTGAATTCGG-3′; consensus IFNα (forward) 5′-ATGGCTAGRCTCTGTGCTTTCCT-3′, (reverse) 5′-AG GGCTCTCCAGAYTTCTGCTCTG-3′; β-actin (forward) 5′-AGCCAT GTACGTAGCCATCC-3′, and (reverse) 5′-CTCTCAGCTGTGGTGG TGAA-3′. The quantity was normalized using the formula of the 2−∆∆CT method.
Statistical analysis was performed using Mann–Whitney's U test for disease phenotypes and Student's t-test for flow cytometric analysis. A value of p < 0.05 was considered as statistically significant.
The authors thank Dr. A. Sato-Hayashizaki, Mr. N. Ishihara, Ms. K. Kojyo, Ms. N. Ohtsuji, and Ms. T. Ikegami for excellent technical assistance. This work was supported in part by Grants-in-Aid for Scientific Research (C) from the Ministry of Education, Science, Technology, Sports and Culture of Japan, and Grant for Research on Intractable Diseases from the Ministry of Health, Labour, and Welfare of Japan.
Conflict of interest
The authors declare no financial or commercial conflict of interest.
type II collagen
cyclic citrullinated peptide
IgG Fc receptor IIB
systemic lupus erythematosus
quantitative real-time PCR
Y chromosome-linked autoimmune acceleration mutation