TRIM21 Dysfunction Enhances Aberrant B-Cell Differentiation in Autoimmune Pathogenesis

TRIM21 is one of the autoantigens that reacts with an anti-SS-A antibody (Ab) present in patients with systemic lupus erythematosus (SLE) and Sjögren's syndrome. TRIM21 is thought to play a role in B-cell proliferation and apoptosis, among other activities. Here we examined a pathological role of TRIM21 in SLE. Trim21-deficient MRL/lpr mice were generated by backcrossing Trim21-deficient C57BL/6 mice to MRL/lpr mice. The levels of serum anti-dsDNA Ab and urine protein at 28 weeks of age were significantly higher in Trim21-deficient MRL/lpr mice as compared to wild-type MRL/lpr mice (p = 0.029 and 0.003, respectively). Resting B cells from Trim21-deficient mice showed significantly higher abilities to differentiate into plasmablasts and to produce Ab as compared with control mice. Due to the reduction of TRIM21-mediated ubiquitylation, IRF5 protein expression was increased in Trim21-deficient MRL/lpr mice (p = 0.021), which correlated with increased plasmablast generation and immunoglobulin production. B cells from SLE patients with anti-TRIM21 Ab seropositivity also showed a significantly higher ability to differentiate into plasmablasts as compared with those without anti-TRIM21 Ab or healthy controls. These results suggest that TRIM21 dysfunction contributes to SLE pathogenesis by promoting B-cell differentiation, for which anti-TRIM21 Ab may be partly responsible.


INTRODUCTION
Systemic lupus erythematosus (SLE) is a chronic, systemic autoimmune disease of unknown cause. The etiology of SLE is thought to be multifactorial and to include contributions from both environmental factors, such as hormones (1), ultraviolet light (2), and infection (3), and genetic susceptibility which has been investigated by genome-wide association studies (4). As the disease is characterized by B-cell hyperactivity with the presence of increased immunoglobulin and a variety of autoantibodies in patients' sera, modulating B-cell function has been considered as a promising therapeutic approach (5,6).
Tripartite motif containing 21 (TRIM21), a ∼52-kDa protein also known as Ro52/SS-A, is one of major autoantigens recognized by anti-SS-A autoantibody present in sera of autoimmune diseases such as SLE and Sjögren's syndrome (7)(8)(9). TRIM21 belongs to the TRIM family, in which proteins are commonly comprised of a RING-finger domain, one or two B-box domains and a coiled-coil domain (10,11).
Previous studies have revealed that TRIM21 is expressed ubiquitously but predominantly in hematopoietic cells and induced by type I and type II interferons (IFNs) (9,11,12). TRIM21 ubiquitylates IFN regulatory factors (IRFs) via its E3 ubiquitin ligase activity on the RING-finger domain and regulates the production of type I IFN and IL-12p40 (13)(14)(15)(16)(17)(18). It has been reported that the TRIM21 overexpression in a B-cell line leads to reduced growth and increased apoptosis after activation (19,20).
Two groups have disrupted the murine Trim21 gene to analyze the function of TRIM21. We showed that NF-κBdependent proinflammatory cytokine production is increased in Trim21-deficient C57BL/6 mice (17). Another group reported that IL-23-dependent Th17 differentiation and production of proinflammatory cytokines and type I IFN are enhanced and thereby contact dermatitis is evoked in Trim21-deficient C57BL/6 mice (16). These results suggest that TRIM21 has physiological roles for suppressing the excess inflammatory response and keeping homeostasis.
It has been shown that TRIM21 is expressed at higher levels in peripheral blood mononuclear cells (PBMCs) in SLE patients as compared to healthy controls (21). Although the role of TRIM21 in SLE pathogenesis is still unknown, previous studies strongly suggest that TRIM21 can act as a suppressor for autoimmune and inflammatory response and its dysfunction can cause the pathological state of SLE.
In this study, we generated and analyzed MRL/lpr lupusprone mice in which Trim21 gene is deleted to investigate the function of TRIM21 in autoimmune pathogenesis. The mice showed worsening SLE pathology with significantly increased autoantibody production and urine protein relative to wild-type MRL/lpr mice. We found the aberrant B-cell differentiation is accompanied by increased expression of transcription factors, IRF5 and BLIMP-1, which are crucial for B-cell differentiation and antibody (Ab) production (22)(23)(24). These factors have also been identified by SLE genome-wide association studies (25,26). Similar to the results from mouse gene disruption studies, B cells from SLE patients with seropositivity of anti-TRIM21 Ab also indicated significantly higher ability to differentiate into plasmablasts and to produce Ab as compared with controls. Together, our results point that TRIM21 dysfunction promotes aberrant B-cell differentiation and Ab production in some SLE patients, which may be associated with anti-TRIM21 Ab.

Mice
Trim21 −/− C57BL/6 mice have been described previously (17). C57BL/6JJmsSlc, MRL/MpJJmsSlc +/+ (MRL/+) and MRL/MpJJmsSlc lpr/lpr (MRL/lpr) mice were purchased from Japan SLC (Hamamatsu, Japan). Trim21 −/− C57BL/6 mice were backcrossed to MRL/lpr mice for more than 10 generations to produce Trim21 −/− MRL/lpr mice. All mice were maintained under specific pathogen-free conditions within the animal facility at Yokohama City University, and female mice were used in all experiments. A comprehensive mouse genotyping examination was performed by ICLAS monitoring Center (Kawasaki, Japan). All experiments of skin-draining lymph nodes (sdLNs) were performed using bilateral axillary and inguinal lymph nodes. All animal experiment protocols were approved by the animal protocol ethics committee of Yokohama City University.

Patients
Seventeen patients with SLE (16 women and one man), who fulfilled the revised 1997 American College of Rheumatology criteria for SLE (27), and five healthy controls (4 women and one man) were enrolled in the study. The study was conducted in accordance with the Declaration of Helsinki, and informed consent was obtained from all patients and healthy controls before study enrollment. The study design was approved by the ethics committee of Yokohama City University Hospital (B100701027).

Primer
Sequence References

T Cell Preparation
Mice CD3 + T cells were isolated from the spleen and sdLN of 8-week-old mice using Mouse CD3ε MicroBead Kit (Miltenyi Biotec), according to the manufacturer's protocols. Gene expression of T-cell transcription factor in the isolated CD3 + T cells were evaluated using quantitative real-time RT-PCR (qPCR). For human PBMCs, CD43 − resting B cells, and cultured B cells, the FcR were blocked with anti-human FcR Ab (Human TruStain FcX; BioLegend) for 10 min at 4 • C. The cell staining was performed according to the method of standardizing immunophenotyping for the human immunology project (36).
Zombie Aqua dye (BioLegend) or Zombie Green dye (BioLegend) were used for detecting dead cells in samples, according to the manufacturer's protocols. All gates were set using fluorescence minus one control. Stained cells were analyzed on BD FACSCanto II instrument (BD Biosciences, Franklin Lakes, NJ, USA), and data were analyzed using FlowJo software version 10 (FlowJo, Ashland, OR, USA).

Multiple Soluble Analyte Immunoassay
The level of cytokines and immunoglobulins in cell supernatants were measured by bead-based immunoassay.

Measurement of Proteinuria and Serum Autoantibodies
Urine was collected from individual mice every 2 weeks from 8 weeks of age for over 16 h in metabolic cages (Shinano Manufacturing Co., Tokyo, Japan). The concentration of urine albumin and creatinine were measured using DC Protein Assay Reagent (Bio-Rad Laboratories, Hercules, CA, USA) and Creatinine Parameter Assay Kit (R&D Systems, Minneapolis, MN, USA), respectively, and the urine albumin-creatinine ratio was quantified.

Ubiquitylation Assay
Mice CD43 − resting B cells were stimulated with 1 µg/ml anti-mouse IgM Ab and 100 µg/ml poly(I:C) for 72 h. Cultures were treated with 25 µM MG-132 (Merck Millipore, Darmstadt, Germany) for 2 h before harvest. Extracts were immunoprecipitated with an anti-IRF5 Ab (Santa Cruz Biotechnology, Dallas, TX, USA) bound to protein G agarose beads (Cell Signaling Technology, Danvers, MA, USA). Immunoprecipitants were analyzed by Western blot using anti-ubiquitin Ab.

TRIM21 Deficiency Promotes Exacerbation of Lupus-Like Symptoms in MRL/lpr Mice
To verify the pathological role of TRIM21 in the autoimmune response, we first examined whether the expression of TRIM21 is influenced by an autoimmune condition. MRL/lpr mice, which are model mice spontaneously developing lupus-like symptoms, including anti-TRIM21 Ab production (37,38), expressed the transcript of Trim21 ubiquitously as shown in the previous study using C57BL/6 mice (17). However, the Trim21 transcript tended to be expressed more preferably in lymphoid tissues such as thymus and spleen in MRL/lpr mice as compared to their control, MRL/+ mice ( Figure 1A).
The genetic background of Trim21 +/− MRL/lpr mice was confirmed by genome scanning based on the microsatellite short tandem repeat analysis (Figure 2). The mice were genotyped by PCR analysis (Figure 1B), and Trim21-deleted allele was confirmed by the expression of an enhanced GFP (EGFP) reporter gene replacing Trim21 exons in cells from the immune tissues using flow cytometric analysis ( Figure 1C). As previously reported (39), Fas lpr mutation shortened life-span ( Figure 3A). Both Trim21 +/+ MRL/lpr and Trim21 −/− MRL/lpr mice started to die at around 24 weeks of age and ∼90% dead by 60 weeks of age. There was no significant difference in survival between them. Although splenomegaly and LN swelling were observed in both Trim21 +/+ MRL/lpr and Trim21 −/− MRL/lpr mice, there was no significant difference in the weights of spleens and sdLNs between the two genotype mice (Figure 3B). On the other hand, the titer of serum anti-dsDNA Ab and daily urine protein were significantly increased over time in Trim21 −/− MRL/lpr mice as compared with Trim21 +/+ MRL/lpr mice (Figures 3C,D). In particular, two severe cases of urine protein appeared in Trim21 −/− MRL/lpr mice at 28 weeks while no such severe cases were seen in Trim21 +/+ MRL/lpr mice ( Figure 3D).
Taken together, these results indicate that TRIM21 deficiency promotes exacerbation of systemic autoimmunity in MRL/lpr mice.

TRIM21 Deficiency Increases B Cells in the Spleen of Lupus Model Mice
It has been reported that autoantibody production and urine protein due to glomerulonephritis in MRL/lpr mice depend on B cell-intrinsic MyD88 signal (40). To examine whether the autoimmune pathogenesis in Trim21 −/− MRL/lpr mice is associated with B-cell abnormalities, we investigated the populations of leukocyte subsets in spleens and sdLNs by flow cytometric analysis. The number of B220 + B19 + B cells was significantly higher in spleens from Trim21 −/− MRL/lpr mice as compared to those from Trim21 +/+ MRL/lpr mice (p = 0.008; Figure 4A). The difference is mainly reflected by the increased number of IgM + IgD + CD19 + mature B cells in Trim21 −/− MRL/lpr mice (Figure 4A). Although there was no significant difference in the number of B220 + B19 + B cells in sdLNs between Trim21 +/+ MRL/lpr mice and Trim21 −/− MRL/lpr mice, the number of CD19 − CD138 + plasma cell was significantly higher in Trim21 −/− MRL/lpr mice as compared to Trim21 +/+ MRL/lpr mice (p = 0.009; Figure 4B). There were no significant differences in the activation markers, CD80 and CD86, in B cells from spleens and sdLNs between Trim21 +/+ and Trim21 −/− MRL/lpr mice (Figures 4A,B).
Dendritic cells (DCs) are also known to be important for autoimmune pathology like SLE. There were no significant differences in the proportions of conventional DCs (cDCs), CD4 + DCs, CD8 + DCs, and plasmacytoid DCs (pDCs), and the expression levels of activation markers, CD80 and CD86, in   cDCs from spleens and sdLNs between Trim21 +/+ MRL/lpr and Trim21 −/− MRL/lpr mice (Figures 5C,D).
These results suggested that B cells, rather than DCs and T cells, are essential in the autoimmune pathology of Trim21 −/− MRL/lpr mice.

TRIM21 Deficiency Promotes Aberrant B-Cell Differentiation and Ab Production in MRL/lpr Mice
Although we found that mature B and plasma cells were increased in lymphoid tissues of Trim21 −/− MRL/lpr mice, it is unknown whether they are due to an intrinsic abnormality of B cells or some abnormal state which can affect B-cell differentiation indirectly. To verify B cell-intrinsic function in Trim21 −/− MRL/lpr mice, we isolated resting state B cells, which are CD43 negative, from spleens and stimulated them by BCR signal (anti-IgM Ab), CD40L signal (anti-CD40 Ab), and/or TLR3/7 ligands. Surprisingly, resting B cells from Trim21 −/− MRL/lpr mice showed differentiation to plasmablasts in a significantly higher rate at 24 or 72 h after stimulation as compared to those from Trim21 +/+ MRL/lpr mice (Supplemental Figures 1A,B and Figures 6A,B). Not only in the absence but also in the presence of CD40L signal, known to be related to T cell-dependent B-cell activation, the Bcell differentiation into plasmablasts was enhanced by TRIM21 deficiency. The results suggest that TRIM21 has roles in B-cell differentiation both in T cell-dependent and T cell-independent manners [ Figure 6B; (42,43)].
There could be the possibility that the rate of plasmablasts was influenced by cell life-span because some previous reports showed that TRIM21 acts on apoptosis-related molecules such as BCL-2 and c-FLIP (19,20). However, there was no significant difference in the survival of B cells at 24 h after stimulation between Trim21 +/+ MRL/lpr and Trim21 −/− MRL/lpr mice (Supplemental Figure 1C).
Next, we investigated the effect of TRIM21 deletion in Ab production by stimulating the resting B cells with anti-IgM Ab, anti-CD40 Ab, and/or TLR3/7 ligands. The concentrations of IgG1, IgG2a, and IgA in the cell supernatants, in which the stimulated resting B cells were cultured for 24 or 72 h, were significantly higher in Trim21 −/− MRL/lpr mice as compared to Trim21 +/+ MRL/lpr mice in multiple stimulation conditions (Supplemental Figure 2 and Figure 7).
Based on the above results, it is conceivable that TRIM21 deficiency leads to an intrinsic abnormality of B cells that induce aberrant B-cell differentiation and Ab production in MRL/lpr mice.

TRIM21 Deficiency Induces
Overexpression of Blimp-1 and Suppression of Bcl6 via Overexpression of IRF5 Due to the Decreased Ubiquitylation IRF5 is the transcription factor known as one of the most essential factors for B-cell differentiation (23,24). IRF5 promotes the differentiation of the late B cell and regulates transcription factors necessary for differentiation into activated B cells, plasmablasts and plasma cells (24). In the differentiation into antibody-producing cells (APCs) such as plasmablasts and plasma cells, IRF5 regulates the expression of IRF4, consequently promotes the expression of BLIMP-1, which drives mature B cells to differentiate into APCs, and thereby suppresses the expression of BCL-6, which is known as an inhibitor of APCs development (44). Furthermore, it has been reported that TRIM21 ubiquitylates IRF5 protein and promotes its degradation from the overexpression experiments using human embryonic kidney (HEK) 293 cell line (45).
Although the E3 ubiquitin ligase activity of TRIM21 for IRF5 is not known in primary B cells, we hypothesized  that TRIM21 regulates plasma cell development via IRF5 ubiquitylation. In the ubiquitylation assay, there was no significant difference between Trim21 +/+ and Trim21 −/− MRL/lpr mice in the total amount of ubiquitylated proteins in resting B cells at 72 h after stimulation with anti-IgM Ab and TLR3 ligand ( Figure 8A). On the other hand, ubiquitylated IRF5 was remarkably decreased in Trim21 −/− MRL/lpr mice as compared to Trim21 +/+ MRL/lpr mice ( Figure 8A). Instead, the unubiquitylated IRF5 protein was significantly increased in Trim21 −/− MRL/lpr mice as compared to Trim21 +/+ MRL/lpr mice (p = 0.021; Figures 8A,B). Furthermore, there was no significant difference in Irf5 mRNA expression in resting B cells after stimulation with anti-IgM and TLR3 ligand for 72 h (Figure 8C), also suggesting that TRIM21 negatively regulates IRF5 protein expression via post-translational modifications.
We further investigated the change in factors important for B-cell differentiation downstream region of IRF5 by Trim21 deletion. IRF4 and IRF8 are also known to be essential transcription regulators in B-cell differentiation (44,46,47). Irf4 has already been reported as one of the target genes of IRF5 (24). Both mRNA and protein of IRF4 showed significantly higher expression levels in B cells from Trim21 −/− MRL/lpr mice at 72 h after stimulation as compared to those from Trim21 +/+ MRL/lpr mice (Figures 8B,C). On the other hand, there were no significant differences in the mRNA and protein expression levels of IRF8 between the two genotype mice (Figures 8B,C). IRF4 has already been shown to suppress BCL6 expression and promote BLIMP1 expression (48,49). The mRNA expression of Blimp1 was significantly increased, and the expression of Bcl6 mRNA was significantly decreased in B cells from Trim21 −/− MRL/lpr mice as compared to those from Trim21 +/+ MRL/lpr mice ( Figure 8C). The flow cytometric analysis using intracellular staining showed significantly higher expression of BLIMP-1, and significantly lower expression of BCL-6 at the protein level in B cells from Trim21 −/− MRL/lpr mice as compared to Trim21 +/+ MRL/lpr mice (p = 0.014 and 0.049, respectively; Figures 9A,B).
Together with the previous papers, these results suggest that TRIM21 negatively regulates the expression of IRF5 protein, decreases IRF4 expression, and thereby induces both BLIMP-1 suppression and BCL-6 expression, resulting in suppression of aberrant B-cell differentiation and Ab overproduction during lupus-like pathological condition.

Aberrant B-Cell Differentiation and Ab Production in SLE Patients With Anti-TRIM21 Ab
Our previous study showed that TRIM21 mRNA and protein in PBMCs of patients with SLE were significantly higher than those of healthy controls and that the expression level of TRIM21 mRNA was well-correlated with disease activities of SLE (21). It is conceivable that the high expression of IFN signature in SLE may be responsible for increased expression of TRIM21 as its suppression mechanism. Furthermore, our previous study showed that a negative correlation is observed between the expression of IFN signature mRNA and TRIM21 mRNA in healthy controls and SLE patients without anti-TRIM21 Ab who are considered to have normal function of TRIM21, but the correlation disappeared in SLE patients with anti-TRIM21 Ab (21). In this context, it was suggested that the function of TRIM21 is impaired by anti-TRIM21 Ab in patients with SLE. Therefore, we examined the difference of function of B cells between SLE patients with and without anti-TRIM21 Ab. Seventeen patients with SLE and 5 healthy controls were enrolled in the study. Eleven and 6 of the patients showed seronegative and seropositive for anti-TRIM21 Ab, respectively. There were no significant differences in the background parameters including female ratio, age, disease duration, SLE activity, treatment, and laboratory data ( Table 2).
There was no significant difference in the proportion of CD19 + B cells in PBMCs among healthy controls, SLE patients without anti-TRIM21 Ab and SLE patients with anti-TRIM21 Ab ( Figure 10A). The proportion of CD19 + B cells after magnetically activated cell sorting of CD43 − resting B cell was more than 90% in all three groups ( Figure 10A).
CD43 − resting B cell was stimulated by BCR signal (anti-IgM Ab), CD40L, and/or TLR3/7 ligands for 24 h and then tested for differentiation into CD19 + CD20 − CD38 + plasmablasts. As shown in Figures 10B,C, the percentage of plasmablasts was the highest in SLE patients with anti-TRIM21 Ab. Further, the percentage of plasmablasts was significantly higher in SLE patients without anti-TRIM21 Ab than that in healthy controls. Accordingly, the amounts of immunoglobulins produced in the culture was the highest in SLE patients with anti-TRIM21 Ab, followed by B-cell cultures from SLE patients without the Ab, and the lowest in healthy controls (Figure 11).
These results indicate that B cells in patients with SLE, especially in those with anti-TRIM21 Ab, are defective in plasmablasts differentiation and Ab production. Anti-TRIM21 Ab may be related to the B-cell abnormalities via inhibition of TRIM21 function.

DISCUSSION
TRIM21 has been shown to regulate the expression of various proinflammatory cytokine genes, type I IFN genes and IFNinducible genes by modulating the signals from the innate immunoreceptor such as TLR via ubiquitylation of various IRF proteins (12,45,50). Previous gene disruption studies have not shown the pathological role of TRIM21 in an autoimmune state such as SLE (16,17). Therefore, we made Trim21 −/− MRL/lpr mouse widely used as a lupus mouse model and examined the phenotype change and the reactivity of immune cells to various TLR ligands (37). This is the first report that shows the pathological role of TRIM21 dysfunction in the lupus mouse model.
Flow cytometric analyses showed that B cells were significantly increased in the spleens and sdLNs by Trim21 gene disruption in lupus-prone mice. Interestingly, Brauner et al. reported that the prognosis was reduced when the Trim21 gene expression was decreased in patients with diffuse large B-cell lymphoma (51). The report suggests that the dysfunction of TRIM21 promotes abnormal activation and proliferation of B cells. Lazzari et al., showed by the overexpression experiments using cell lines that TRIM21 regulates the expression of IRF5 by ubiquitylation (45). IRF5 has been shown to be necessary for  B-cell differentiation and Ab production of autoreactive B cells (23,24,52). The knockdown of IRF5 by siRNA transfection in human primary naïve B cells reduces plasmablast differentiation and Ab production (24). Our data are well-consistent with these previous studies because our ubiquitylation assay showed that TRIM21 deficiency leads to decreased ubiquitylation and increased expression of IRF5 in B cells. On the other hand, a recent study showed that the Irf5 risk alleles for SLE intrinsically affects the function of myeloid cells rather than that of B cells (53). Although the overexpression experiments are considered as somewhat artificial in general, not the risk haplotype but the levels of IRF5 expression may be important in B cell function.
We found the differences in B-cell differentiation and Ab production between the Trim21 −/− MRL/lpr mice and the Trim21 +/+ MRL/lpr mice. B-cell activation can occur both in T cell-dependent and T cell-independent manners and DCs also play a role in B-cell differentiation via several functions including the production of B cell-activating factor belonging to the tumor necrosis factor family (BAFF) (54,55). In our experiments, Trim21 deletion did not change the states of T cells and DCs. Moreover, we did not find any differences in the expression of CD80 and CD86, which are important for T cell-B cell interaction (56), in B cells between Trim21 +/+ and Trim21 −/− MRL/lpr mice. These results suggest that the enhanced B-cell proliferation and Ab production by Trim21 deletion is mainly caused by the abnormality of B cells themselves, and not by that of T cells or DCs.
Anti-SS-A Ab is observed in sera of patients with autoimmune diseases such as SLE and Sjögren's syndrome, and TRIM21 is one of the autoantigens that bind to anti-SS-A Ab (57,58). We have shown that negative correlation between TRIM21 and type I IFNs expression, which is seen in PBMCs from healthy controls, disappeared in those from SLE patients, especially those with anti-TRIM21 Ab seropositivity (21). In this study, the differentiation from resting B cells to plasmablasts and Ab production were significantly increased in SLE patients with anti-TRIM21 Ab seropositivity as compared to those negative for anti-TRIM21 Ab. In combination with previously reported in vitro data that TRIM21 has a suppressive role in human B-cell proliferation (19,20,59), these results strongly suggest that the disruption of TRIM21 fu0nction can be caused partly by anti-TRIM21 Ab in patients with SLE. Consistent with this hypothesis, B-cell differentiation and Ab production were increased in the Trim21 −/− MRL/lpr mice as compared with the Trim21 +/+ MRL/lpr mice in our experiments. On the other hand, type I IFNs induce TRIM21 expression while TRIM21 itself suppresses the production of type I IFNs (14,15,17,60). This suggests that the increased TRIM21 expression in SLE patients seems due to an increase in serum type I IFN concentration, and it can be considered as a negative feedback reaction to suppress type I IFN overproduction. Therefore, further increasing TRIM21 expression and suppressing the production of anti-TRIM21 Ab can be new treatment strategies for SLE.
How the extracellular anti-TRIM21 Ab can affect the function of intracellular TRIM21 proteins is an important point that has not been studied sufficiently yet. Further studies are needed to clarify this issue. It is possible that the extracellular anti-TRIM21 Ab directly invades the cell and binds to intracellular TRIM21, which may affect function. There were some reports that an anti-DNA Ab that recognizes DNA in the nucleus invades intracellularly from outside the cell via calreticulin or myosin-1 (61)(62)(63). Furthermore, it was recently reported that TRIM21 recruits incoming antibody-coated virus and targets it to the proteasome via its E3 ligase activity, resulting in degradation of virions in the cytosol of the infected cells (64). Another recent report shows that anti-TRIM21 autoantibodies inhibit the E3 ligase activity of TRIM21 by sterically blocking the E2/E3 interaction between TRIM21 and UBE2E1 (57). Thus, there is a possibility that serum anti-TRIM21 Ab enters the cytosolic compartment by some cue, interacts with TRIM21 protein, and interferes with its function.
In our experiment, Trim21 −/− MRL/lpr mice showed significant increases in urinary protein and serum dsDNA Ab titers as compared to wild-type MRL/lpr mice. Therefore, it is certain that the Trim21 gene deletion has an impact on nephritis and autoantibody production in lupus pathogenesis. On the other hand, Trim21 −/− MRL/lpr mice were not different from wild-type mice in the survival curve. The result suggests that various compensatory actions work against the changes associated with the Trim21 deletions to maintain life as a whole body. This may also be related to the fact that human SLE is a multifactorial disease.
During the preparation of this paper, Brauner et al. have reported that follicular B cells and Ab production were significantly increased by immunization of Trim21-deficient C57BL/6 mice with ovalbumin (65). Although the paper does not show an effect of TRIM21 deficiency on phenotype in disease model and the mechanism of B cell increase has not been investigated, their data are well-consistent with our results.
In conclusion, we showed that TRIM21 deficiency promotes aberrant B-cell differentiation and Ab production, leading to exacerbation of the lupus-like symptoms in lupus-prone mice. We also found that B cells from SLE patients with anti-TRIM21 Ab also have significantly higher abilities to differentiate into plasmablasts and to product Ab, suggesting that anti-TRIM21 Ab may be related to the TRIM21 dysfunction in human SLE pathogenesis. These findings suggest that TRIM21 and anti-TRIM21 Ab can be promising targets for SLE treatment.

DATA AVAILABILITY STATEMENT
The datasets generated for this study are available on request to the corresponding author.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by the ethics committee of Yokohama City University Hospital (B100701027). The patients/participants provided their written informed consent to participate in this study. This animal study was reviewed and approved by the Animal Protocol Ethics Committee of Yokohama City University. prepared figures, and devised the manuscript. YS, YKi, KO, and HN coordinated and critically discussed research work. All the above-listed authors edited the manuscript. The authors declare that this study received funding from GlaxoSmithKline K.K. and Bristol-Myers K.K. These funders were not involved in the study design, collection, analysis, interpretation of data, the writing of this article or the decision to submit it for publication.