Systemic lupus erythematosus (SLE) is a common autoimmune disease that is defined by the inappropriate activation of self-reactive T and B cells and the production of autoantibodies and immune complexes that can cause irreparable organ damage (1). The variety of SLE symptoms make both diagnosis and treatment difficult. Consequently, it is of great significance to further explore the mechanism underlying the pathogenesis of SLE.

B lymphocytes perform various functions in SLE, including autoantibody production, presentation of antigen to T cells, and activation of myeloid cells to produce various cytokines (2–4). B cells can be activated by diverse factors, including Toll-like receptors (TLRs). TLRs are pathogen pattern recognition receptors that sense conserved microbial components (5). TLRs are crucial for both the development of innate immune responses and adaptive immunity, and they are thought to bridge the innate and adaptive immune systems (6). Studies show that overactivation or impaired function of the TLRs that detect nucleic acids, TLR7 and TLR9, induces dysfunction of innate immunity and breakdown of immune tolerance, which can lead to the occurrence of autoimmunity (7, 8). Both TLR7 and TLR9 are strongly associated with interferon (IFN) production by plasmacytoid dendritic cells (pDCs) (9). In addition, they are also expressed in B cells, wherein they play crucial functions (10). Surprisingly, TLR7 and TLR9 appear to play opposing roles in SLE B cells (11). TLR7 contributes to the loss of germinal centre (GC) tolerance and drives the extrafollicular B-cell response implicated in SLE (12). Critically, TLR9 in B cells seems to protect against SLE even though it is needed to produce anti-double-stranded DNA (dsDNA) antibodies (13). In fact, TLR9 has a complex dual regulatory mechanism in B cells, which can both maintain immune tolerance and participate in autoimmunity (14). This may be due to the specific presence of signalling pathways in B cells (such as BCR, CD19 and related molecules)that play a role in these processes (14, 15). In addition, the participation of co-stimulatory molecules, pro-inflammatory factors and MyD88 (which is a key molecule downstream of TLR9) also determines the fate of B cells (16, 17). Additionally, pDCs and T cells can accelerate the progression of SLE by releasing cytokines to reinforce the sensitivity of TLR7 in B cells (18–21). These findings suggest that the onset of SLE is likely significantly influenced by dysregulation of TLR7 and TLR9 signalling in B cells. However, the findings of functional and genetic studies of TLR7 and TLR9 in SLE remain controversial.

In this review, the contribution of intracellular TLR signalling in B lymphocytes to SLE is discussed. First, we present the expression levels and functions of TLR7 and TLR9 in B cells during the development of SLE. Second, we emphasize that downstream signalling pathways connected to B cell proliferation and differentiation in SLE are activated by TLR7 and TLR9 in B cells. Third, we discuss the opposing effects of TLR7 and TLR9 in B cells and their interactions in SLE. Fourth, we discuss how pDCs and T cells accelerate the conversion of B cells into plasma cells in patients with SLE through TLR signalling. Finally, we look ahead to potential strategies for targeting TLRs to treat SLE.

Earlier studies have shown that the TLR7 gene is involved in SLE (22–24). The TLR7 gene is located on human chromosome Xp22.3. Interestingly, unlike other genes on the X chromosome, TLR7 is able to evade silencing in females (25). Therefore, it has a higher level of expression in monocytes, pDCs and B cells in females than in males (26). Similar to the case in humans, this female-biased increased expression of X-linked genes is also present in mice susceptible to autoimmune diseases (27). In addition, TLR7 may be coexpressed with other X-linked genes (e.g., CXorf21) to jointly promote the occurrence of autoimmune diseases with gender bias (28). With the increased use of genome-wide association studies (GWAS), researchers have found that SLE risk and clinical phenotype and the production of autoantibodies in SLE are influenced by TLR7 single-nucleotide polymorphisms (SNPs) (29, 30). This finding partly explains the heterogeneity of SLE. With the rapid development of gene sequencing technologies, TLR7 mRNA has been found to be upregulated in the peripheral blood mononuclear cells of SLE patients by using bioinformatics analysis (31). Another investigation found that the expression of TLR7 in CD19⁺ B cells in SLE patients was markedly elevated compared to that in healthy individuals (32). Another study showed that SLE patients had expansion of CD19⁺CD24^hiCD38^hi transitional B cells and increased release of autoantibodies due to overexpression of TLR7 (33). Similarly, transitional T1 B cells expansion and autoantibodies production can be found when overexpression of TLR7 in transgenic mice (34). In addition, Tlr7 transgenic mice spontaneously manifest SLE-like illnesses in a B-cell-dependent manner (35). Furthermore, Hwang et al. (36) demonstrated that upregulation of TLR7 in SLE-prone mice induces B-cell marginal zone impairment, elevated levels of antibodies targeting RNA/protein complexes in B cells, and progression of disease. Brown et al. (37) reported that a child with serious lupus had a novel de novo missense mutation in TLR7 (TLR7^Y264H). The variant selectively enhances sensing of guanosine and 2’,3’-cyclic guanosine monophosphate (2’,3’-cGMP) and triggers lupus-like disease when introduced in mice (generating so-called kika mice). Further studies have found that the Age-associated B cells (ABCs) of kika mice have higher TLR7 expression than those of WT mice (but lower TLR7 expression than Tlr7-double-positive Yaa mice). Nevertheless, the effects of the Y264H mutation were more serious than those of mutations of the Yaa allele, with kika mice showing higher levels of autoantibodies (37). Moreover, when wild-type (WT) mice of diverse genetic backgrounds were treated with the TLR7 agonists imiquimod and R848 for 4 weeks, they developed systemic autoimmune disease accompanied by an increased titre of autoantibodies against nucleic acids and multiple organ damage, including glomerulonephritis, carditis, hepatitis, and photosensitivity (38). According to additional research, focusing on the TLR7 pathway in lupus-prone models may provide insights for preventing the onset of serious disease (39). The above studies show that TLR7 promotes the occurrence and development of SLE at the genetic and molecular levels. These findings suggests that TLR7 is a candidate for targeted therapy for SLE.

TLR7 contributes to the loss of germinal centre tolerance and drives the extrafollicular B-cell response implicated in SLE ( Figure 1 ). In the periphery, GCs are a crucial place where B-cell–tolerance is assessed (40). Studies have shown that the GC response and production of antibodies are significantly decreased in Tlr7-null mice B cells upon RNA virus infection (41). Consistent with these findings, TLR7 in B cells is essential for GC production in the B^WAS−/− model, in which B cells lack the Wiskott-Aldrich syndrome protein (42). In another mouse model with two copies of Tlr7, GCs fail to eliminate markedly autoreactive B cells, and these cells undergo subsequent clonal expansion and somatic mutation and then enter the plasma-cell compartment (43). This finding suggests that B-cell-intrinsic TLR7 can disrupt GC tolerance. Another study showed that B6.Sle1b mice treated with a TLR7 agonist have increased numbers of spontaneous germinal centres (Spt-GCs) (44). Conversely, Tlr7 deficiency prevented the formation of Spt-GCs in autoimmune B6.Sle1b mice and reduced the production of autoantibodies (44). Additionally, a lack of TLR7/MyD88 impairs B-cell proliferation and survival after stimulation compared to B6 controls (44). These results demonstrate that the development of Spt-GCs and elevated autoantibody production in SLE are mediated by B-cell-intrinsic TLR7 signalling.

Furthermore, in cases of active SLE, TLR7 can promote extrafollicular B-cell differentiation, causing resting naive B cells to constantly differentiate into activated naive B cells, DN2 B cells, and ultimately Ab-secreting cells (12). Individuals with SLE have higher levels of activated naive B cells(CD11c⁺IgD⁺CD27⁻CD21⁻MTG⁺CD23⁻) and the DN2 subset of IgD⁻CD27⁻ double-negative B cells (DN2 B cells; IgD⁻CD27⁻CD11c⁺T-Bet⁺CD69⁺CD21⁻CD24⁻CD38⁻CXCR5⁻FCRL4⁻FCRL5⁺) (45). Moreover, SLE patients with increased abundance of DN2 B cells have more anti-RNA and anti-Sm/RNP autoantibodies and exhibit higher disease activity (12). DN2 cells are the predominant B-cell population in active SLE; they tend to develop into plasma cells and exhibit high levels of CD11c and T-Bet expression but low levels of CXCR5 and CD62L expression (12). Interestingly, the phenotype of DN2 cells is distinct from that of activated naive B cells. The two cell types share a similar transcriptional profile, with only 42 differentially expressed genes between them (12, 46). Furthermore, DN2 B cells are thought to be derived from activated naïve B cells (12). B lymphocyte-induced maturation protein 1 (BLIMP-1) and IFN regulatory factor 4 (IRF4) levels are higher in these two types of cells and control plasma cell differentiation and the expression of transcription factors (TFs) related to plasmablasts (47). DN2 cells express low levels of ETS proto-oncogene 1 (ETS1), a CD22-regulated TF that inhibits plasma cell differentiation (48). These features result in the accumulation of extrafollicular autoreactive B cells (49). SLE DN2 cells are hyperresponsive to TLR7 agonists, and TLR7-mediated upregulation of costimulatory molecule expression and suppression of the expression of CD72 (a receptor) inhibit the activation of endosomal TLR7 (50). Moreover, chronic TLR7 activation in lupus models is sufficient to cause CD11c⁺ B cells to differentiate and produce anti-Sm/RNP autoantibodies, whereas B-cell intrinsic TLR7 deficiency reduces the number of CD11c⁺ B cells and ameliorates lupus-like conditions (42). Indeed, TLR7 possesses stimulatory action in these autoreactive B-cell subsets and, in both humans and animals, can cause naive B-cell differentiation towards DN2 cells and plasma cells with the help of IL-21 and IFNγ (47, 51). In SLE patients, IL-21 promotes the growth of CD11c⁺T-Bet⁺CD21⁻CD38⁻B cells, which resemble DN2 B cells and are autoreactive and associated with clinical symptoms (18). Furthermore, TLR7 and B-cell receptor (BCR) agonists cooperate with IFNγ, IL-2, IL-21 and B-lymphocyte stimulator (BlyS) to stimulate naive human B cells in SLE, and an in vitro study reported that the produced IgD⁻CD27⁻CD11c⁺T-Bet^hiCD21⁻CXCR5⁻IRF4^intFcRL5⁺ B cells resemble DN2 B cells (52). Age-associated B cells (ABCs), the mouse equivalent of DN2 B cells, are another crucial B-cell subpopulation in SLE. These CD11b⁺CD11c⁺T-Bet ⁺ cells are increased in various mouse models of lupus, including Mer ^−/− and SWEF-deficient mice; these cells are also especially sensitive to TLR7 and differentiate into autoantibody-secreting cells (53–55). Another study found that ABCs in C57BL/6 mice can also present antigens to T cells (56). Specifically, the Tlr7-null lupus-prone mouse model lacks CD11c⁺T-Bet⁺CD21⁻ B cells. Moreover, with repeated TLR7 agonist stimulation, ABCs accumulate rapidly in mice based on a specific TLR7 signalling pathway in B cells (51). Brown et al. (37) identified that even though there was marked Spt-GC formation in Tlr7 ^Y264H mice, the lupus-like symptoms did not resolve with GC formation deficiency, thus indicating that extrafollicular ABCs may be the main reason for pathogenicity (37). In summary, the GC reactions and extrafollicular responses lead to B cell activation and the formation of plasma cells via B-cell-intrinsic TLR7 signalling in SLE. Additionally, TLR7 signalling restricts the proliferation of B10 cells, a regulatory B-cell subpopulation that generates the immunosuppressive cytokine IL-10, in an IFNγ-dependent manner (57).

TLR9 is a receptor that recognizes viral and bacterial DNA containing unmethylated cytosine–phosphate–guanine (CpG) motifs (58). One study showed that anti-dsDNA antibody levels and proteinuria in childhood-onset lupus nephritis (LN) are associated with the expression of TLR9. This result suggests that TLR9 is involved in the pathophysiology of LN (59). Consistent with these results, investigation has demonstrated that whole blood TLR9 mRNA levels are considerably greater in SLE patients than in healthy controls and are associated with the severity and type of renal pathology (60). Furthermore, the levels are positively correlated with SLEDAI grade, anti-dsDNA antibody titre, and interleukin-6 (IL-6) and other cytokine levels but negatively correlated with complement C3 levels (61). Notably, during a two-year follow-up period, patients with poor prognoses still had elevated TLR9 mRNA levels (62). Compared with the levels at SLE onset, the levels of TLR9 mRNA in patients with a favourable prognosis were obviously decreased (62). Strikingly, studies show that Tlr9 deficiency in B cells in murine models of lupus is sufficient to accelerate renal disease even though deplete anti-nucleosome antibodies. Moreover, overexpression of B-cell–specific Tlr9 induces remission of nephritis (13). In contrast, Tlr9 deficiency in other cell types, including pDCs, dendritic cells, and neutrophils, has no obvious effect on LN (13). These lines of evidence suggest that TLR9 in B cells plays a specialized role in restricting the formation of autoantibodies and preventing the onset of disease. That the expression of TLR9 in SLE is proportional to disease activity appears to contradict its protective function, suggesting a unique and complex role of TLR9 in SLE B cells.

Although B cells are crucial to the pathophysiology of SLE, the aberrant pathways that result in the loss of B-cell tolerance are unknown. A significant fraction of autoreactive B cells are eliminated throughout development (63), but self-reactive and polyreactive clonotypes still make up a significant portion of mature B-cell pools (40, 64). This indicates that there may be activation-associated checkpoints that can prevent these cells from differentiating into memory B cells and producing antibodies targeting self-antigens. However, patients with SLE likely have impaired peripheral and central B-cell tolerance checkpoints, leading to an accumulation of autoreactive B cells in their blood (65). TLR9 is expressed in the spleen and skeletal muscle in mammals and in peripheral blood mononuclear cells, mainly in pDCs and B cells. However, the biological effects are different in these cells. Endogenous or exogenous nonmethylated CpG-DNA can be recognized by BCR and internalized into B cells, and the process may involve phosphatidylinositol 3-kinases (PI3Ks) (66). Subsequently, the CpG-DNA is transported into early endosomes and ultimately transported into late endosomes containing TLR9, by which it is specifically identified, activating downstream signalling pathways and triggering inflammatory responses and immune effects (66). TLR9 can also promote B cell differentiation into plasma cells through germinal centre (GC) formation and extrafollicular B-cell response. However, unlike TLR7, TLR9 has a complex role in B cell differentiation and is affected by multiple elements. This results in TLR9 acting as both an immune checkpoint to maintain peripheral and central tolerance and to participate in autoimmunity (14). Rookhuizen and DeFranco discovered that TLR9-MyD88-dependent signalling in B cells increased GC output by increasing affinity maturation, the production of memory antibodies, a transition to the IgG2a subclass of antibodies and by specifically producing high-affinity antibodies (67). Additionally, the combined ligation of TLR9 and BCR has also been demonstrated to activate autoreactive B cells in previous studies (68–71). Moreover, these activated autoreactive B cells develop into extrafollicular plasma cells that can facilitate the generation of IgG2 isotype autoantibodies (72, 73). Eckl-Dorna and Batista (74) found that for both in vitro and in vivo studies, Ag-CpG conjugates are effective at promoting B cell proliferation and differentiation into short-lived extrafollicular plasma cells via intrinsic TLR9. In addition, ABCs do not respond to BCR ligation alone, but they quickly multiply when TLR9 is activated in aged mice (75). Activated B cells are prepared for an ABC fate by B cell-intrinsic TLR9 signals, but they also initiate a cell death program (53). Cancro’s group (14, 76) have posited that antigen taken up by the BCR is internalized and interacts with innate molecular pattern sensors. If TLR9 is activated (either alone or in conjunction with TLR7), the cell is also primed to adopt an effector or memory T-Bet⁺ ABC fate. However, the default pathway is cell cycle arrest and apoptosis, unless it is avoided by prosurvival signals or costimulation via cognate T cell assistance. In reality, studies have revealed that BCR can bind to TLR9 agonists on all premature murine B-cell subsets, and CD27^– human B cells can undergo apoptotic death after an initial proliferative burst that involves hyperphosphorylation of nuclear factor-κB (NF-κB) and p38 mitogen-activated protein kinase (MAPK) (16, 77). Moreover, p38 inhibits the persistent expansion and the survival of autoreactive B cells by inducing G1 cell cycle arrest and mitochondrial apoptosis (78, 79). However, B cells going through this stage can be saved, and the way they are saved influences what happens to them thereafter (16). In fact, impaired TLR9 function, BlyS stimulation, cognate T-cell-mediated effects or CD40 costimulation with IL-21 or IFN-γ can prolong the lifespan of autoreactive B cells (80, 81). Interestingly, a study using CpG to stimulate B cells and pDCs isolated from SLE patients who had not received hydroxychloroquine treatment found decreased expression of both B-cell activation molecules and numerous cytokines in B cells (15). However, normal reactions appear in pDCs from SLE patients; therefore, it appears that there is defective TLR9 function in SLE B cells (15). This impaired TLR9 function might prevent SLE-related autoreactive B-cell death and, consequently, harmful autoantibody generation.

All TLRs in B cells trigger similar signalling pathways that in turn activate NF-κB, MAPKs, extracellular signal-modulating kinase (ERK), p38, and Jun N-terminal kinase (JNK) (82). The common linker protein used by almost all TLRs is MyD88, which has death and toll interleukin 1 receptor (TIR) domains (83). When MyD88 binds to a TLR, it recruits members of the IL-1 receptor-associated kinase (IRAK) family, IRAK1 and IRAK4, by interacting with the death domains, thereby phosphorylating TNF receptor correlator 6 (TRAF6) (84). TRAF6 is phosphorylated by IRAKs and binds to Ubc13 and Uev1A to form a complex that activates a mitogen-activating protein kinase kinase (MAPKKK) called transforming growth factor β-activating kinase-1 (TAK-1). Activated TAK-1 phosphorylates MKKs, which are upstream kinases of p38 MAPKs and JNK (85). p38 is involved in many cellular physiological and pathological processes, including apoptosis, cellular stress, the cell cycle, and the inflammatory response (86). In addition, TAK-1 can activate the IκBα kinase (IKK) complex, consisting of the IKKα and IKKβ kinases and the hinge protein IKKγ (85). Phosphorylation of IκBα leads to its own degradation and the release of NF-κB ( Figure 2 ). The NF-κB signalling pathway is critical in innate and adaptive immunity because it modulates the transcription of various immune-inflammatory genes, such as TNF, IL-1, and IL-6, resulting in chemotaxis, granulocyte and macrophage aggregation, and lymphocyte infiltration (84, 87, 88). Although the downstream pathways triggered by many TLRs are similar, TLR7 and TLR9 perform different functions in controlling the behavior of SLE B cells. We therefore hypothesize that in SLE, B-cell-specific signalling impacts the response of TLRs.

pDCs are a dominant producer of IFN-I. The secretion of IFN-I by pDCs involves the TLR7/9-MyD88 signalling pathway (19). SLC15A4, which is a histidine transporter that controls lysosomal pH, is necessary for this process and impacts the production of autoantibodies (113–116). IFN-I signalling has a crucial role in SLE pathogenesis and specifically in the germinal centre response (117). By specifically upregulating TLR7 expression, pDC-derived IFN-I increases the TLR7 sensitivity of human naive B cells. The level of TLR9 expression, however, is unaltered (118). Soni et al. (119) experiments with mice showed that endosomal TLR7/TLR9 and IFN-I derived from pDCs promote the transformation of extrafollicular B-cells into short-lived antibody-forming cells (AFCs). Moreover, the optimal expansion of extrafollicular (ExFO) CD138⁺ B cells and ICOS^hiExFO-Th cells requires IFN-I (119). In addition, T cells participate in the proliferation and differentiation of SLE B cells. CD4⁺ effector T cells, including follicular helper T (Tfh) cells, are the main sources of TLR7-induced IFN-γ. Furthermore, IFN-γ signalling in SLE plays an indispensable role in TLR7-driven autoreactive B-cell development into antibody-secreting cells via germinal centre signalling (20, 21). The differentiation of SLE B cells into plasma cells is facilitated by IL-21 secreted by Tfh cells (18, 120).Additionally, T-Bet and CD11c expression in TLR-activated B cells is controlled by an interaction between IL-4, IL-21, and IFN-γ (121).. In addition, Soni et al. (119) showed that extrafollicular plasmablasts drive the autoantibody response in a T-cell-dependent manner. These autoantibodies bind to antigens to form immune complexes. On the one hand, immune complexes can cause damage to organs; on the other hand, they can be recognized by pDCs to produce IFN-I, thus forming a vicious cycle (119). Taken together, these findings suggest connections between pDCs, T cells, B cells, and endosomal TLRs ( Figure 1 ).

TLRs were originally identified as pattern recognition receptors and are now considered an important hub connecting innate and adaptive immunity. SLE is characterized by successive stimulation of the innate and adaptive immune systems by endogenous nucleic acids released by necrotic or apoptotic cells. TLR7 and TLR9, whose ligands are ssRNA and dsDNA, respectively, work as innate sensors for detecting viral infections. Endosomal TLR recognition of self-nucleic acids in B cells is considered a crucial step in the progression of SLE and leads to the formation of antinuclear antibodies. Both TLR7 and TLR9 mRNA are expressed to varying degrees in SLE B cells, but the two play divergent roles. TLR7 is a significant contributor to the pathogenesis of SLE in B cells, mediating GC responses and extrafollicular B-cell responses to promote the expansion of antibody-secreting cells and accelerate disease progression. However, B-cell-specific TLR9 appears to play a protective role in SLE. Moreover, the conventional views hold that impaired TLR9 function in SLE B cells leads to a breakdown of the balance of the TLR7 and TLR9 pathways, which manifests as enhanced TLR7 signalling, impaired apoptotic death after an initial proliferative burst, and increased conversion of B cells to plasma cells instead of Bregs, which eventually worsens the disease. However, recent research has found that B-cell-intrinsic TLR9 seems to have complex bidirectional regulatory roles in SLE, and its expression does not affect TLR7 signalling (17). These novel findings reveal unprecedented insights into TLR9 biology in environments beyond SLE. This unique dual regulatory role of TLR9 may be related to Toll–IL-1R (TIR) domain signalling from IL-1R family members, which have both proinflammatory and anti-inflammatory functions (136). Only 46% homology exists between the TIR domains of TLR7 and TLR9, explaining why these receptors have different functions. This may partly explain why TLR7 and TLR9 have different roles in SLE B cells. Ultimately, genetic assessment of the TIR domains in each protein and exploration at the biochemical level are needed to determine such differences.

Infection is known to be an important trigger exacerbating SLE, and TLR-induced innate immunity can kill pathogens to maintain homeostasis (137, 138). However, abnormal activation of TLR7 or impaired TLR9 function in B cells in patients with SLE leads to disruption of immune homeostasis and adaptive immune activation upon exposure to pathogens containing ssRNA and dsDNA. Therefore, therapy targeting B-cell TLR7/TLR9 has potential for SLE. Some drugs targeting TLR7 have been developed and applied in the clinic; however, drugs targeting B-cell TLR9 are still relatively lacking. We propose a method to enhance the function of TLR9 by upregulating the expression of CD19 in B cells, thereby restoring the postproliferative apoptosis pathway of TLR9. This is expected to prevent disruption of immune homeostasis in SLE patients upon infection with pathogens containing ssRNA and dsDNA, providing new ways to target B-cell TLR9 to treat SLE. Curiously, the expression of TLR9 in B cells of patients with SLE is upregulated and positively correlated with SLE disease activity. This seems to contradict the protective effects of TLR9 and may be the result of a feedback mechanism that upregulates the expression of TLR9 to induce protection. However, this hypothesis needs to be validated. Likewise, whether TLR9 expression and/or function differs between humans and mice remains unclear.

It should be noted that SLE is a disease involving multiple factors, so the study of SLE pathogenesis cannot be limited to B cells. A large number of studies have indicated that cytokines such as IFN and ILs derived from pDCs and T cells, the expression of CD19 and BCR, and the interaction of transporters such as UNC93B1 and SLC15A4 with TLRs operate in a complex network that controls autoreactive antibody responses to nucleic acid antigens and results in the complex pathogenesis and symptom heterogeneity of SLE. However, the interactions of these factors are still unclear and need to be further explored.

LW drafted the manuscript. BZ revised and created some of the figures. XW, RL, HF, and LH read and revised the manuscript. ZZ and XM performed the literature search. C-QC participated in the revision of words and sentences. XS designed, revised and approved the final manuscript. All authors contributed to the article and approved the submitted version.