Expression of TLR10 in Peripheral B Cell Subsets of Patients with Primary Sjögren’s Syndrome

Primary Sjögren’s syndrome (pSS) is considered as a B cell-mediated disease, yet the precise role of B cells in the pathogenesis is not fully understood. Toll-like receptor 10 (TLR10) is highly expressed in human B cells, indicating that TLR10 probably plays a vital role in regulating B cell function as well as B cell-related diseases. However, the biology of TLR10 in pSS is less researched. Here, we examined the TLR10 expression in peripheral B cell subsets isolated from both pSS patients and healthy controls (HCs) and further analyzed the correlations between TLR10 expression and disease activity. We observed that TLR10 was highly expressed in switched memory B cells (CD19 + CD27 + IgD − ) in the pSS patients compared with the HCs. TLR10 expression in CD19 + B cells, memory B cells (CD19 + CD27 + ) and switched memory B cells in pSS patients was negatively correlated with serum levels of anti-SSA antibody and B cell-activating factor of TNF family (BAFF), respectively. A much lower proportion of high-activity pSS patients was observed in TLR10 high- as compared to low-expressed patients. TLR10 expression in CD19 + B cells, naïve B cells (CD19 + CD27 − IgD + ), memory B cells, and switched memory B cells was signicantly increased in low-activity pSS patients as compared with HCs and high-activity pSS patients. Our study concluded that TLR10 expression in CD19 + B, naive B, and memory B cells was negatively correlated with pSS disease activity, suggesting that TLR10 might play a critical in the progression of pSS.


Introduction
Primary Sjögren's syndrome (pSS) is a chronic autoimmune disorder affecting exocrine glands of the body, preferentially lacrimal and salivary glands. The global prevalence of pSS is approximately 0.4-1%, in which around 30-40% of pSS patients will progress to at least one additional systemic autoimmune complications [1] and nearly 5% of pSS patients may develop to B cell malignancies, most common salivary gland mucosa-associated lymphoid tissue lymphomas [2]. Clinical treatments for pSS easily fail due to the heterogeneity of clinical phenotype in pSS patients. Moreover, plenty of studies on pSS demonstrated that numerous factors seem to contribute to the progression of the pSS [3,4]. However, the mechanism by which pSS develops, although being widely researched, remains unclear.
Currently, pSS is considered as a B cell-mediated disease characterized by autoantibodies and hypergammaglobulinaemia in patients [4]. Besides the production of autoantibodies, B cells can activate T cells by presenting autoantigen and secreting multiple in ammatory cytokines upon Toll-Like Receptor (TLR) activation, thereby contributing to the development of pSS [5]. However, the precise role of B cells in the pathogenesis of pSS is still poorly understood. According to clinical e cacy for rituximab treatment in different trials, it was suggested that B cells at different developmental stages contribute to the broad clinical phenotypes in pSS patients, suggesting that understanding the abnormality of B cell development and differentiation is essential to uncover the pathogenesis of pSS [6][7][8].
TLRs are pattern recognition receptors that have crucial roles in the initiation of innate immunity and the activation of adaptive immunity. Many studies have proved that TLR signaling is required for human B cell activation and plays an important role in autoimmune diseases [9,10]. Among ten human TLRs (TLR1-10), TLR10 remains the least understood one because it presents in human beings, but not in the mice, and its ligand has not been identi ed yet [11]. TLR10 is expressed at the highest level in B cells, followed by plasmacytoid dendritic cells but not expressed in monocytes, natural killer cells, and T cells [12,13] [19]. Patients with other autoimmune or in ammatory diseases, severe renal or liver disease, or cancer were excluded. The disease activity of pSS patients was determined according to the ESSDAI. All the pSS subjects were divided into two groups: a high-activity group (ESSDAI≥5), a low-activity group (ESSDAI<5) [20]. In addition, 25 age-and sex-matched HCs were chosen for comparison. More information about age, gender, and treatments are shown in Table 1.
The data are expressed as n (%), mean ± standard deviation (SD). a The age of each group is proved to be normal distribution (sample K-S test, P>0.05). And T-test about age between each group ensured that there was no signi cant difference between them (P>0.05). b Anti-SSA antibody, anti-SSB antibody, anti-Ro52 antibody data were lacking in a few subjects. c The patient de ned as "unknown" was someone who was on her rst visit to our hospital, could not tell which medication to use.

Specimen Collection and Laboratory Testing
For analysis of the peripheral blood, 2 ml venous blood was collected from each participant into ethylene diamine tetra-acetic acid (EDTA)-containing collection tubes (Becton Dickinson). Samples were centrifuged to collect the upper serum layer and then frozen and stored at -80°C until use. After that, the remaining sample was processed to isolate peripheral blood mononuclear cells (PBMCs) by Ficoll density-gradient centrifugation. PBMCs were used for ow cytometry analyses.

Antibodies and Flow Cytometry Analysis
The immunophenotyping of B cells was performed in the peripheral blood samples using the following uorochrome-labeled anti-human antibodies: and human Fc receptor blocking solution (cat: 422302) were purchased from Biolegend; IgD-FITC (clone IA6-2) was purchased from BD Biosciences, and 7-amino-actinomycin D (7AAD) (cat:00-6993-50) was purchased from Invitrogen. Fresh isolated PBMCs were rst blocked with human Fc blocking reagents and stained with diluted antibodies at 4°C for 15 minutes in the dark. Then, PBMCs were washed twice with cold FACS buffer (1×PBS containing 2% FBS), resuspended in 0.3 ml of FACS buffer, and analyzed by ow cytometry. Approximately 300,000~500,000 events were collected per sample. The data were collected with a FACS Calibur (Beckman CytoFLEX, USA) and analyzed using the FlowJo software version 10.0.

Statistical analysis
The results are expressed as the means ± standard deviation (SD) and medians (interquartile range). Statistical comparisons were performed by Student's t-tests. Differences among the three groups were determined by the Kruskal-Wallis H nonparametric test. Correlation analyses between two parameters were performed by Spearman's correlation method. All statistical analyses were performed using the SPSS software version 20 (SPSS Inc., Chicago, Illinois, USA) and GraphPad Prism (v.8.0, CA). A p-value < 0.05 was considered statistically signi cant.

TLR10 expression in B cells in pSS patients
Both previous studies [12,13] and the Human Protein Atlas showed that TLR10 mRNA mainly enriched in B cells, less in dendritic cells and monocytes, and undetectable in other peripheral blood immune cells.
Therefore, to investigate the differences in TLR10 expression on the B cell surface between pSS patients and HCs, we isolated PBMC from the above two groups and detected them by ow cytometry. The proportion of CD19 + B cells in PBMC between pSS patients and HCs was comparable (Sup. Fig. 1).
Compared with the HCs, the pSS patients expressed relatively high levels of TLR10 on CD19 + B cells' surface when determined by the mean uorescence intensity (MFI) (Fig. 1a-b), but there were no statistical differences between the two groups. In addition, the proportion of TLR10 + CD19 + B cells was also similar between the two groups ( Fig. 1c-d).

Correlation between TLR10 expression in B cells and pSS related autoantibodies
Anti-SSA, anti-SSB, and anti-Ro52 are important clinical diagnostic indicators for pSS patients, and their concentrations are usually positively correlated with pSS progression [21][22][23]. We analyzed the TLR10 expression in B cells in pSS patients according to the extractable nuclear antigen pro le results.
Interestingly, the expression of TLR10 in CD19 + B cells from anti-SSA +++ and anti-Ro52 +++ pSS patients is signi cantly reduced as compared with the anti-SSA −/+ and anti-Ro52 +/− pSS patients respectively ( Fig. 2a-b), and anti-SSB +++ pSS patients show a moderate reduction in TLR10 expression as compared with anti-SSB −/+ pSS patients (Fig. 2c), indicating that the expression of TLR10 in B cells might be related with the production of autoantibodies in pSS patients. Further ELISA results showed that the expression of TLR10 in CD19 + B cells is negatively correlated with serum level of anti-SSA (r = -0.4599, p = 0.0138), anti-SSB (r = -0.4028, p = 0.0336) and ANA (r = -0.7855, p = 0.0011) in pSS patients (Fig. 2d-f). Moreover, the expression of TLR10 in CD19 + B cells was negatively correlated with BAFF (r = -0.4092, p = 0.0306) (Fig. 2g), which is important for survival and activation of B cells and presents excessive level in pSS patients [24]. These results suggested that the expression of TLR10 in CD19 + B cells might be correlated with pSS formation and/or progression.

TLR10 expression is mainly upregulated in switched memory B in pSS
Numerous studies have been reported that memory B cells, PB, and plasma cells are the key subsets of B cells involved in the pathogenesis of pSS [4]. We further analyzed the expression of TLR10 in peripheral B cell subsets, including CD19 + CD24 ++ CD38 ++ transitional B cells, CD19 + IgD + CD27 − naïve B cells, CD19 + CD27 + memory B cells, CD19 + IgD + CD27 + unswitched memory B cells, CD19 + IgD − CD27 + switched memory B cells, and CD19 + CD24 − CD38 ++ PB (Fig. 3a), obtained from both the pSS patients and HCs. The results showed that TLR10 expression was similar between the pSS patients and HCs in transitional B cells, naïve B cells, memory B cells, unswitched memory B cells, and PB (Fig. 3b). Interestingly, the expression of TLR10 in switched memory B cells was signi cantly increased in pSS patients compared with the HCs (Fig. 3b). Moreover, the expression of TLR10 in memory B cells and switched memory B cells was negatively correlated with serum level of anti-SSA (memory B, r = -0.4034, p = 0.0333; switched memory B, r = -0.3953, p = 0.0373) and BAFF (memory B, r = -0.3966, p = 0.0367; switched memory B, r = -0.3760, p = 0.0486) in pSS patients (Fig. 3c-f), respectively. The expression of TLR10 in memory B cells and switched memory B cells showed no signi cant correlation with the serum level of anti-SSB (memory B, r = -0.2288, p = 0.2416; switched memory B, r = -0.2376, p = 0.2234) in pSS patients (Sup Fig. 2a-b). In addition, the expression of TLR10 in transitional B cells, naïve B cells, unswitched memory B cells, and PB showed no obvious correlation with serum level of anti-SSA, anti-SSB, and BAFF in pSS patients, respectively (Sup Fig. 2c-e). These results further con rmed that the expression of TLR10 was increased in switched memory B cells, which might play an important role in pSS progression.

TLR10 expression in B cells is negatively correlated with pSS progression
Although TLR10 expression in CD19 + B cells showed no signi cant differences between the HCs and pSS patients, we wondered that whether TLR10 expression in B cells changed with pSS progression. Firstly, the expression of TLR10 in B cells in pSS patients with low-and high-activity evaluated by ESSDAI according to their clinical features [20] was analyzed by ow cytometry. As shown in Sup Fig. 3a, the proportion of CD19 + B cells among the HCs, low-and high-activity pSS patients was comparable.
Interestingly, the expression of TLR10 in CD19 + B cells in low-activity pSS patients signi cantly increased compared with the HCs, while decreased in high-activity pSS patients compared with low-activity pSS patients (Fig. 4a). Then we divided the pSS patients into TLR10 high-and low-expressed groups based on the average value of the TLR10 MFI of CD19 + B cells, and calculated the proportion of low-or highactivity patients between the two groups. As shown in Fig. 4h, the proportion of high-activity patients in TLR10 low-expressed pSS patients was signi cantly higher than that in TLR10 high-expressed pSS patients (76.19% vs 7.69%). Conversely, the proportion of low-activity patients in TLR10 low-expressed pSS patients was signi cantly lower than that in TLR10 high-expressed pSS patients (23.81% vs 92.31%). Moreover, correlation analysis showed that the pSS progression was closely related to TLR10 expression in CD19 + B cells (p <0.001) ( Table 2). These results suggested that TLR10 expression in CD19 + B cells was negatively correlated with pSS progression.

TLR10 expression in naive and memory B cells is upregulated in low-activity pSS patients
To clarify whether the TLR10 expression in B subsets changed during the pSS progression, we analyzed the expression of TLR10 in peripheral B subsets in the HCs, low-and high-activity pSS patients. Compared with the HCs, the high-activity pSS patients presented with a signi cantly increased proportion of transitional B cells, naive B cells, and PB proportion, respectively, but decreased proportion of memory B cells, as well as unswitched and switched memory B cells (Sup. Fig. 3b-g). Consistent with the above results, the expression of TLR10 in transitional B cells and PB was relatively comparable among these groups ( Fig. 4b and g). Notably, the expression of TLR10 in naïve B cells, memory and switched memory B cells was increased in low-activity pSS patients compared with the HCs (Fig. 4. c, d, and f). With the progression of pSS, the expression of TLR10 in naïve and memory (including unswitched and switched) B cells was signi cantly decreased in high-activity pSS patients compared with low-activity pSS patients ( Fig. 4c-f). These results suggested that TLR10 expression might suppress pSS progression by taking part in the process of B cell activation and differentiation.

Discussion
pSS is a systemic rheumatic autoimmune disease characterized by abnormal B cell biological function [5]. TLRs, as one kind of pattern recognition receptor, are well known for their signi cant roles in in ammation and innate immunity. Previous studies have proved that the expression of TLR7 and TLR9 in B cells may play an important role in the dysregulation of B cells in pSS [25,26]. TLR10, as the latest identi ed functional TLR in human beings, is mainly expressed in B cells [27]. However, few studies on TLR10 expression in pSS have been reported so far. Here, we found that TLR10 was highly expressed in switched memory B cells in the pSS patients compared with the HCs. TLR10 expression in CD19 + B cells, memory and switched memory B cells in pSS patients was signi cantly negatively correlated with the anti-SSA autoantibody and BAFF production, respectively. The TLR10 high-expressed pSS patients usually had a relatively lower proportion of high-activity condition as compared with the TLR10 lowexpressed pSS patients. Moreover, TLR10 was highly expressed in CD19 + B cells, naïve B cells, and memory B cells in low-activity pSS patients as compared with the high-activity pSS patients.
Today, most investigators agree that there are two main CD27 + memory B cell compartments in blood, IgM + IgD +, and IgM − IgD − B cells, the former is unswitched memory B cells exhibiting characteristics of marginal zone B cells, whereas the latter most likely represents class-switched B cells [28,29].
Importantly, we found that TLR10 was particularly highly expressed in CD27 + IgD − switched memory B cells in pSS patients compared with the HCs, while TLR10 expression in CD27 + IgD + unswitched memory B cells in pSS patients was slightly lower than the HCs. Our results indicate that TLR10 might play an important role in B cell germinal center reaction in patients with pSS. Further studies are necessary to uncover the role and detailed mechanism of TLR10 about B cell activation and class switch response.
We also assessed the proportion change of B cell subsets in the peripheral blood of patients with pSS. It was found that the pSS patients presented with a comparable proportion of CD19 + B cells, increased proportions of transitional B, naive B, and PB, but a reduced proportion of memory B, as compared with the HCs (unpublished data), which were consistent with previous studies [30,31]. Importantly, the proportion change of B cell subsets mentioned above was even more obvious between HCs and highactivity pSS patients, but showed no statistical difference between HCs and low-activity pSS patients, suggesting that the proportion change of B cell subsets focused on high-activity pSS patients. Notably, the expression of TLR10 was increased only in switched memory B cells in pSS patients compared with the HCs, whereas that was reduced in CD19 + B, naive B, memory B (including unswitched and switched memory B) in high-activity compared with low-activity pSS patients. Considering that the change of TLR10 expression mainly occurs in naïve B and memory B, TLR10 might play an important role in B cell activation and differentiation during the pSS progression. In addition, we observed that the TLR10 expression in B cells had a complex effect on the proportion change of patients' corresponding B subsets.
pSS disease activity is assessed by the physician according to ESSDAI from the patient's clinical manifestations [32]. Among all criterion scores, anti-SSA received the highest average weights [33].
Moreover, it has been reported that the presence and increased titers of anti-SSA, anti-SSB, and rheumatoid factor serum autoantibodies are correlated with the severity in pSS patients [34]. Intriguingly, we observed that TLR10 expression in CD19 + B cells in pSS patients was negatively correlated with anti-SSA, anti-SSB, and ANA autoantibodies. Moreover, TLR10 expression in memory and switched memory B cells in pSS patients was also negatively correlated with anti-SSA but not anti-SSB autoantibodies. When using anti-SSB autoantibody as the biomarker, different correlation results between TLR10 expression in CD19 + B and switched memory B cells were achieved, but understandable because the anti-SSB shows a weaker correlation with pSS progression as compared with anti-SSA according to the ESSDAI [33].
Importantly, the expression of TLR10 in CD19 + B, naïve B, and memory B cells in low-activity pSS patients was signi cantly increased compared with high-activity pSS patients. Thus, we postulated that TLR10 could inhibit pSS progression via negatively regulating B cell function.
Although the ligand and downstream signaling pathways of TLR10 remain unclear, it has been identi ed as an immunomodulatory receptor with inhibitory properties [35][36][37]. The high expression of TLR10 in human B cells suggests that TLR10 may regulate B cell function. Hess et al. reported that TLR10 can suppress responses mediated by a variety of B cell co-stimulatory signals and attenuate both T cellindependent and T cell-dependent antibodies production in a TLR10 knock-in mouse model [35]. BAFF, an important cytokine for B cell maturation, proliferation, and survival, is upregulated in salivary gland tissue and blood from pSS patients and plays a vital role in the pathogenesis of pSS [38,39]. We observed that TLR10 expression in CD19 + B cells, memory and switched memory B cells in pSS patients also was inversely correlated with BAFF level in pSS patients' serum ( Fig. 2d and Fig. 3d, f), which further supported our hypothesis that TLR10 inhibited pSS progression by suppressing B cell function.
In summary, for the rst time, we found that the protein level of TLR10 expression in peripheral switched memory B cells was increased in pSS patients and was signi cantly negatively correlated with both anti-SSA and BAFF production. Moreover, TLR10 expression in peripheral CD19 + B, naive B, and memory B is negatively correlated with pSS disease activity. These ndings suggest that TLR10 might participate in the pathogenesis of pSS by negatively regulating B cell function, and support the further investigation on TLR10 biological function in B cells.
Declarations FUNDING Figure 1 The