NFS/sld mice were obtained from the Central Institute for Experimental Animals (Kawasaki, Japan). Neonatal thymectomy (Tx) was performed to female NFS/sld mice on day 3 after birth in order to establish the pSS model. At 4 weeks after the Tx surgery, peripheral blood mononuclear cells from the tail vein of these mice were analyzed by using flow cytometry. The completeness of the Tx surgery was confirmed by evaluating the proportion of CD90.2⁺ T cells (<7%). The control mice used in this study were sham (non)-thymectomized female NFS/sld mice. Female MRL/MpJ-Fas ^lpr/lpr (MRL/lpr) and aly/aly mice were obtained from Japan SLC Inc. (Hamamatsu, Japan). The total number of mice used in this study is 249 mice. Mice were housed at 21°C–23°C on a 12-h light/dark cycle in our facility under specific pathogen-free condition. The present study was designed and undertaken by following the “Fundamental Guidelines for Proper Conduct of Animal Experiments and Related Activities in Academic Research Institutions” under the jurisdiction of the Ministry of Education, Culture, Sports, Science, and Technology of Japan. The protocol was approved by the Committee on Animal Experiments of Tokushima University (permit number T2021-48). Experiments were performed under general anesthesia, and all efforts were made so as to minimize the suffering of the mice involved. Randomization was performed to allocate experimental units to all mice used in each experiment. In order to establish the pulmonary fibrosis model, the mice received a transbronchial instillation of 1.675 mg/kg (body weight) bleomycin (Nippon Kayaku Co.,Ltd.). Three weeks after the bleomycin administration, the mice were euthanized for the undertaking of the following analyses.

Lung tissues were fixed with 10% phosphate-buffered formalin (pH 7.2) and were prepared for histological examination. Sections were stained with hematoxylin and eosin (HE) or Masson’s trichrome staining. The focus score in the lung tissues was determined by counting as foci the visible aggregations containing 50 or more tightly aggregated lymphocytes. In addition, the number of foci infiltrated in the entire tissue section was counted per mm². The area of foci was measured using NIS Elements D (Nikon Corporation). For the semiquantitative evaluation of fibrosis in the lungs, histological analysis was performed by using the Ashcroft score (26). In brief, the degree of fibrosis was scored from 0 to 5 as follows: 0 = normal; 1 = presence of inflammation and fibrosis involving <5% of the lung parenchyma; 2 = presence of inflammation and fibrosis involving <25% of the lung parenchyma; 3 = lesions involving 25%–50% of the lung parenchyma, accompanied by moderate thickening of the bronchial wall without obvious damage to the lung architecture, or small fibrous masses; 4 = lesions involving >50% of the lung, parenchyma, accompanied by definite damage to the lung structure and formation of fibrous bands or small fibrous masses; 5 = lesions involving >50% of the lung parenchyma accompanied by severe distortion of the lung structure and a fibrous obliteration of fields. All evaluations were performed in a blind manner by three pathologists.

Lung tissue sections were deparaffinized in xylenes and were rehydrated by passage through serial dilutions of ethanol in distilled water. Heat-induced antigen retrieval was performed in an antigen retrieval solution (ImmuoActive; Matsunami Glass Ind., Ltd., Kishiwada, Japan) with the use of a microwave, twice, for 5 min. The sections were incubated with Blocking One Histo (Nacalai Tesque) in order to block non-specific reactions. Anti-mouse B220 (eBioscience), anti-mouse CD3 (Cell Signaling Technology) monoclonal antibodies (mAbs), and anti-mouse CD23 (Boster Biological Technology) polyclonal Ab were applied to the sections overnight, at 4°C. After washing with phosphate-buffered saline (PBS), the sections were incubated with an Alexa Fluor 488-conjugated anti-rat IgG (Invitrogen) or an Alexa Fluor 568-conjugated anti-rabbit IgG (Invitrogen) Ab. After washing thrice with PBS, the nuclear DNA was stained with 4′,6-diamidino-2-phenylindole (DAPI; Invitrogen, Waltham, MA), and the sections were observed under an optical microscope (ZEISS Axio Observer 7; Carl Zeiss) and were analyzed through the use of ZEN3.2 blue edition (Carl Zeiss).

Deparaffinized sections from lung tissues were used. Heat-induced antigen retrieval was performed in ImmunoActive, and the sections were incubated with Blocking One Histo. Subsequently, the sections were incubated with anti-mouse CD3 (Cell Signaling Technology), CD19 (Cell Signaling Technology, Danvers, MA), F4/80 (Cell Signaling Technology), CD11b (Cell Signaling Technology), or CD23 (Boster Biological Technology) Abs overnight, at 4°C. After washing with PBS, the sections were incubated with a horseradish peroxidase (HRP)-conjugated secondary Ab (Cell Signaling Technology). HRP reacted with the 3,3′-diaminobenzidine (DAB) substrate through the use of SignalStain DAB Substrate Kit (Cell Signaling Technology), and the sections were then counterstained with hematoxylin.

Under anesthesia, PBS was perfused into the left ventricle of each mouse in order to remove hematopoietic cells among the blood. For the isolation of lymphocytes from the lung and the salivary gland tissues, the tissues were minced into 1–3-mm pieces and were digested with collagenase (1 mg/ml; FUJIFILM Wako Pure Chemical Corp., Osaka, Japan), hyaluronidase (1 mg/ml; Sigma-Aldrich Co., St. Louis, MO), and DNase (10 μg/ml; Roche Diagnostics K.K., Tokyo, Japan) in Dulbecco’s modified Eagle’s medium (DMEM) containing 10% fetal bovine serum (FBS), at 37°C, for 30 min, by using gentleMACS Dissociators (Miltenyi Biotec, Bergish Gladbach, Germany). The spleen was homogenized in DMEM containing 2% FBS through the use of gentleMACS Dissociators (Miltenyi Biotec). By using 0.83% ammonium chloride, the red blood cells were removed from the tissues. In addition, both the viability and the number of the isolated cells were evaluated by a Luna II cell counter (Logos Biosystems, Anyang, South Korea) using trypan blue staining. Subsequently, the proportions of the suspended cells were analyzed by a flow cytometer. The absolute numbers of immune cells were calculated by using the data pertaining to their total cell number and their proportions. As for the lungs and the salivary glands, we used their lateral lobes in order to determine the cell numbers and the proportions of the immune cells in them. In the case of splenocytes, the whole spleen of each mouse was used in order to determine their numbers and their proportions.

Isolated lymphocytes were stained using Abs against fluorescein isothiocyanate (FITC)-conjugated anti-mouse CD21 (BD Biosciences, 7G6) CD4 (BioLegend, RM4-5), PE-conjugated anti-mouse CD23 (BioLegend, B3B4), CD19 (BioLegend, 6D5), PD-1 (BioLegend, San Diego, CA, 29F.1A12), CD45.2 (BioLegend, 104), PE-Cy7-conjugated anti-mouse CD45.2 (BioLegend, 104), CD4 (BioLegend, GK1.5), CXCR5 (Biolegend, L138D7), APC-conjugated anti-mouse CD3 (BioLegend, 17A2), CD19 (Biolegend, 6D5), Alexa flour 700-conjugated anti-mouse CD45.2 (BioLegend, 104), APC-Cy7-conjugated anti-mouse CD19 (BioLegend, 6D5), CD3 (BioLegend, 145-2C11), Brilliant Violet 510-conjugated CD4 (Biolegend, GK1.5), and Pacific Blue-conjugated CD45 (Biolegend, S18009F). A CytoFLEX S flow cytometer (Beckman Coulter, Brea, CA) was used to identify the cell populations according to expression profile. Viable cells were checked by gating on side scatter (SSC)/forward scatter (FSC), FSC-H/FSC-A, and 7-aminoactinomycin D (7AAD) staining solution (Invitrogen). Data were analyzed by using the FlowJo FACS Analysis software (BD Biosciences).

Total RNA from the spleen, the lungs, and the salivary glands was extracted with RNAiso Plus (TaKaRa Bio Inc., Kusatsu, Japan) according to the manufacturer’s instructions. Total RNA was then reverse-transcribed into cDNA using the PrimeScript RT Master Mix (TaKaRa Bio Inc., Cat. No. RR037A). The transcriptions of target genes and β-actin from the tissues were generated with a Light Cycler 96 System (Roche) using TB Green Premix Ex Taq II (TaKaRa Bio Inc., Cat. No. RR820) and the following primers: Gata3, forward, 5′-CCTACCGGGTTCGGATGTAA-3′ and reverse, 5′-CACACACTCCCTGCCTTCTGT-3′; Tbx21, forward, 5′-CCTGTTGTGGTCCAAGTTCAAC-3′ and reverse, 5′-CACAAACATCCTGTAATGGCTTGT-3′; Il4, forward, 5′-TCTCATGGAGCTGCAGAGACTCT-3′ and reverse, 5′-TCCAGGAAGTCTTTCAGTGATGTG-3′; Il13, forward, 5′-AGACCAGACTCCCCTGTGCA-3′ and reverse, 5′-TGGGTCCTGTAGATGGCATTG-3′; Il10, forward, 5′-GCTCTTACTGACTGGCATGAG-3′ and reverse, 5′-CGCAGCTCTAGGAGCATGTG-3′; Ifng, forward, 5′-AGCGGCTGACTGAACTCAGATTGTAG-3′ and reverse, 5′-GTCACAGTTTTCAGCTGTATAGGG-3′; Il6, forward, 5′-TCCTTCCTACCCCAATTTCC-3′ and reverse, 5′-GCCACTCCTTCTGTGACTC-3′; Ccl4, forward, 5′-TGTCTGCCCTCTCTCTCCTCT-3′ and reverse, 5′-AGCAAGGACGCTTCTCAGTGA-3′; Ccl6, forward, 5′-TTATCCTTGTGGCTGTCCTTG-3′ and reverse, 5′-CACGGGATCTGTGTGGCATA-3′; Ccl8, forward, 5′-ACGCTAGCCTTCACTCCAAAA-3′ and reverse, 5′-TTCCAGCTTTGGCTGTCTCTT-3′; Ccl12, forward, 5′-TCGAAGTCTTTGACCTCAACA-3′ and reverse, 5′-GGGAACTTCAGGGGGAAATA-3′; Ccl19, forward, 5′-AACAAAGGCAACAGCACCA-3′ and reverse, 5′-CACACTCACATCGACTCTCTA-3′; Ccl22, forward, 5′-TCATGGCTACCCTGCGTGTC-3′, and reverse, 5′-CCTTCACTAAACGTGGCAGAG-3′; Ccl28, forward, 5′-GCTGTGTGTGTGGCTTTTCAA-3′ and reverse, 5′-TACCTCTGAGGCTCTGATCCA-3′; Cxcl1, forward, 5′-CTGGGATTCACCTCAAGAACATC-3′ and reverse, 5′-CAGGGTCAAGGCAAGCCTC-3′; Cxcl2, forward, 5′-ACCACCAGGCTACAGCGGCT-3′ and reverse, 5′-TCCTGGGGGCCTCACACTCA-3; Cxcl12, forward, 5′-CTTCATCCCCATTCTCCTCA-3′ and reverse 5′-GACTCTGCTCTGGTGGAAGG-3′; Cxcl13, forward, 5′-CATCATGAGGTGGTGCAAAG-3′ and reverse, 5′-GGGTCACAGTGCAAAGGAAT-3′; Ccr1, forward, 5′-ATCCTCAAAGGCCCAGAAACA-3′ and reverse, 5′- TCCTTTGCTGAGGAACTGGTC-3′; Ccr2, forward, 5′-CCATGCAAGTTCAGCTGCCT-3′ and reverse, 5′-TGCCGTGGATGAACTGAGG-3′; Ccr3, forward, 5′-TTGATCCTCATAAAGTACAGGAAGC-3′ and reverse, 5′-CAATGCTGCCAGTCCTGCAA-3′; Ccr4, forward, 5′-GGCTACTACGCCGCCGAC-3′ and reverse, 5′-TACCAAAACAGCATGATGCC-3′; Ccr5, forward, 5′-AATTCTTTGGACTGAATAACTGCA-3′ and reverse, 5′-TCAGGATTGTCTTGCTGGAA-3′; Ccr7, forward, 5′-ATGCTGGCTATGAGTTTC-3′ and reverse, 5′-GCTGCTATTGGTGATGTT-3′; Cxcr5, forward, 5′-ACTCCTTACCACAGTGCAC-3′ and reverse, 5′-GGAAACGGGAGGTGAACCA-3′; Tgfb1, forward, 5′-GACCGCAACAACGCCATCTAT-3′ and reverse, 5′-GGCGTATCAGTGGGGGTCAG-3′; Il33, forward, 5′-ATTTCCCCGGCAAAGTTCAG-3′ and reverse, 5′- AACGGAGTCTCATGCAGTAGA-3′; Col1a2, forward, 5′-CCAAGGGTAACAGTGGTGAA-3′ and reverse, 5′-CCTCGTTTTCCTTCTTCTCCG-3′; Col3a1, forward, 5′-AACGGAGCTCCTGGCCCCAT-3′ and reverse, 5′-CCATCACTGCCCCGAGCACC-3′; Col4a1, forward, 5′-ATGCCCTTTCTCTTCTGCAA-3′ and reverse, 5′-GAAGGAATAGCCGATCCACA-3′; and β-actin, forward, 5′-GACGGCCAGGTCATCACTAT-3′, and reverse 5′-CTTCTGCATCCTGTCAGCAA-3′. Relative mRNA expression of each transcript was normalized against β-actin mRNA.

An anti-mouse CCL6 mAb (5 μg/mouse; R&D Systems) was administered every 2 days via intraperitoneal injections to pSS model mice from the sixth until the eight weeks of their lives. Moreover, an anti-mouse CXCL13 mAb (10 μg/mouse; R&D Systems) was administered twice a week via intraperitoneal injections to pSS model mice from sixth until the eight weeks of their lives. Finally, an Ultra-LEAF purified anti-mouse CD4 mAb (100 μg/mouse; BioLegend) was intraperitoneally injected twice a week into pSS model mice from the sixth until the eighth weeks of their lives. In addition, CD4 mAb was administered using the same protocol from the fourth until the sixth weeks of their lives. As a control, the Ultra-LEAF purified rat IgG2b κ isotype control Ab (5 μg/or 50 or 100 μg/mouse; BioLegend) and the Ultra-LEAF purified rat IgG2a κ isotype control Ab (10 μg/mouse; BioLegend) were administered into pSS model mice as the same regimens.

Dead cells among the lymphocytes isolated from lung tissues were removed by using the Dead Cell Removal Kit (Miltenyi Biotec) for in vitro differentiation into CD23⁺ B cells. CD19⁺ B cells were obtained through positive selection by using the Dynabeads (Invitrogen) and an anti-CD19 mAb (eBioscience). CD19⁺ B cells were dispensed in triplicates (3.5 × 10⁵ cells/well) from 8-week-old three control mice or pSS model mice using 48-well culture plate (Corning Incorporated). B cells were cultured for 1 week in Roswell Park Memorial Institute (RPMI) 1640 containing 10% FBS with an anti-CD40 mAb (5 µg/ml; BioLegend, HM40-3) and recombinant IL-4 (100 ng/ml; BioLegend) as described in a previous report (27).

Differences between individual groups were determined by using a Student’s t-test, or one or two-way analysis of variance, where p-values of <0.05 were considered as statistically significant. GraphPad Prism 8 software was used for all data analysis. Data are presented as mean ± standard deviation (SD).

We have established a mouse model of pSS by using NFS/sld mice treated with neonatal Tx, and the autoimmune lesions in the salivary and the lacrimal glands of these mice can be observed as early as the mice reach 6 weeks of age (28–31). The pathological analysis of the lung tissues in these pSS model mice was performed from the 4th to the 32nd week of their lives. When the mice became 8 weeks of age, a focal lymphocytic infiltration consisting of mononuclear lymphocytes was identified in the pSS model mice, while no inflammatory lesions were observed in the control mice ( Figure 1A ). Focal nodular infiltration was observed around the blood vessels and peribronchially in the pulmonary interstitium in the pSS model mice ( Figure 1B ). The pulmonary lesions in the pSS model mice resembled those of patients with pSS and those observed in bronchus-associated lymphoid tissue (BALT) forming during chronic inflammation in the lung (16, 32, 33). Interestingly, the BALT-like structure observed in these pSS model mice has not been confirmed in the pulmonary lesions of the other SS model mice, such as the MRL/lpr or the alymphoplasia (aly)/aly mice (34–37) ( Supplemental Figure 1A ). In fact, diffuse lymphocytic infiltration around bronchi and vessels was observed in the lesions of 20-week-old female MRL/lpr mice ( Supplemental Figure S1A ), while mild to moderate lymphocytic infiltration around bronchi and vessels was detected in the pulmonary lesions of 20-week-old female aly/aly mice ( Supplemental Figure S1A ). Histopathological analysis showed that pulmonary lesions exist in our pSS model mice as early as they reach their eighth weeks of age ( Figure 1C ). Interstitial fibrosis is known to occur in about half of the pSS patients, accompanied by pulmonary lesions and progressive fibrosing pneumonia that generally leads into severe respiratory failure (38). In the pSS model mice, pulmonary fibrosis is not conspicuous from the 8th until 32nd week of age. On the other hand, the construction of alveolar space and bronchi, thickness of alveolar wall, and any change in the alveolar epithelium were undetectable in the lung tissues of pSS model mice. In addition, no differences between them were observed in body weight at each age and their lifespan. Because fibrosis in this model was hardly observed, the BALT-like lesions may not cause the pulmonary failure showing severe symptoms. It is well known that autoimmune interstitial fibrosis shows more severe pulmonary failure (13, 14, 38). In addition, fibrosis-related genes, such as Tgfb1, Il33, Col1a2, Col3a1, and Col4a1 mRNA, were not upregulated, but rather downregulated, in the lung tissues of pSS model mice, compared with these of control mice ( Supplementary Figure S2 ), suggesting that inflammatory response may contribute to the pulmonary lesions in this model through independent pathway of fibrosis. Therefore, we tried to evaluate fibrosis by using bleomycin as an inducing factor of pulmonary fibrosis in our pSS model mice at 8 or 12 weeks of age. Pathological analysis has shown that prominent fibrosis was detectable at 3 weeks after the bleomycin administration in both the control and the pSS model mice at 11 and 15 weeks of age ( Figure 1D ). No differences were observed between these two groups ( Figure 1D ). In addition, no differences were observed in terms of the evaluation performed through the use of Masson’s trichrome staining between these groups at their 11th or 15th weeks of age ( Figure 1E ). The Ashcroft score was used in order to assess fibrosis, and there were no differences identified between them at their 11th or 15th weeks of age ( Figure 1F ). When lung tissue sections were analyzed with Azan staining, fibrosis in both MRL/lpr and aly/aly mice was hardly observed. Also in these models, fibrosis may be unrelated to the pulmonary pathology ( Supplementary Figure S1B ). These findings demonstrate that pulmonary inflammatory lesions with lymphocytic infiltration were confirmed in our pSS model mice.

T cells among various other immune cells, including B cells, macrophages, dendritic cells, and natural killer cells, are mainly infiltrating in the salivary glands during the early stages of pSS in patients (21). Based on the progression of the inflammatory lesions during the chronic stage of pSS, an increase in the numbers of B cells or plasma cells is also identified in the salivary gland lesions in pSS patients and of mouse models (21, 25). Immunohistochemical analysis has shown that CD19⁺ B cells are mainly infiltrating the pulmonary lesions of the pSS model mice once they reach 8 weeks of age, whereas CD3⁺ T cells were also found ( Figure 2A ). The use of immunofluorescence analysis has also demonstrated that B220⁺ B cells can be prominently observed in the pulmonary lesions of the pSS model mice ( Figure 2B ). In addition, the employed flow cytometry of lymphocytes within the interstitial tissues of the lung has revealed that both the proportion and the cell numbers of CD19⁺ B cells in pSS model mice are significantly increased compared with those of control mice ( Figures 2C–E ). On the other hand, the proportion of CD4⁺ T cells in the salivary gland tissues of the pSS model mice was found to be significantly higher than that of control mice, while there was no difference in terms of the proportion of CD19⁺ B cells between these two groups ( Figures 2C, D ). Our results suggest that B cells may play a key role in the formation of the autoimmune lesions observed in the lungs of the pSS model mice during the early stage of the disease, unlike the inflammatory lesions observed in the salivary glands.

Subsequently, we hypothesized that the peripheral B cells might migrate to the lung tissues through a chemokine network so as to form the inflammatory lesions observed in the pSS model mice. Quantitative reverse transcription–polymerase chain reaction (RT-PCR) analysis has shown that the mRNA expression of various chemokine genes, such as Ccl4, Ccl6, Ccl12, Ccl19, Ccl22, Cxcl2, and Cxcl13, in the lung tissues of pSS model mice was significantly higher than that in control mice ( Figure 3A ). Moreover, the mRNA expression of chemokine receptor genes, such as Ccr3, Ccr4, Ccr5, and Cxcr5, in the lung tissues of pSS model mice was significantly increased when compared with that of control mice ( Figure 3B ). Among these chemokines, CCL6 has been reported to contribute to the lung inflammation observed in allergic mouse models (39). In addition, CXCL13 is known to play a potent role in the B-cell migration (40, 41). Thus, the anti-CCL6 mAb or the anti-CXCL13 mAb was intraperitoneally administered to pSS model mice from their sixth till the eighth week of their lives. Flow cytometric analysis has revealed that the proportion of B cells in the lung tissues of anti-CCL6 or anti-CXCL13 mAb-treated mice was not significantly altered when compared with that observed in the lung tissues of isotype control mAb-treated mice ( Figure 3C ). Our results suggest that an abnormal chemokine network does not potentially contribute to the B-cell migration in the autoimmune pulmonary lesions in our pSS model mice. Instead, B-cell proliferation or differentiation in the lung may play a key role in the formation of the autoimmune lesions.

We then examined the B-cell phenotypes in the pulmonary lesions of the pSS model mice. CD23⁺ CD21⁻ follicular (F) and CD23⁻ CD21⁺ marginal zone (MZ) B-cell phenotypes among the mature CD19⁺ B cell subsets were analyzed by using flow cytometry on spleen cells and on lymphocytes obtained from the salivary glands and the lung tissues. The subset representing the CD23⁺ CD21⁻ FB cells in the lungs of the pSS model mice was found to be clearly increased, in contrast to that of control mice ( Figure 4A ). On the other hand, there were no differences among the proportion of the CD23⁺ CD21⁻ B cells from both the spleen and salivary glands between control and pSS model mice ( Figure 4A ). Both the proportions and the numbers of the CD23⁺ CD21⁻ B cells in the spleen and the salivary glands obtained from the pSS model mice were found unaltered when compared with those obtained from control mice ( Figures 4B, C ). Significantly increased proportions and numbers of CD23⁺ CD21⁻ B cells in the lungs of pSS model mice were identified when compared with those of control mice at 8 and 16 weeks of age ( Figure 4D ). Moreover, immunohistochemical and immunofluorescence analyses revealed that CD23⁺ FB cells were distributed in the center of the inflammatory foci of the pulmonary lesions in pSS model mice ( Figures 4E, F ). These findings suggest that CD23⁺ FB cells contribute to the formation of pulmonary lesions in pSS model mice.

CD23 upregulation in B cells requires signals of IL-4 and of CD40-CD40L (27). Therefore, we assessed mRNA expression of Il4 in the lung tissues of pSS model mice. The Il4 mRNA expression in the lung of pSS model mice was found to be significantly higher than that of control mice, while the expression levels of the spleen of pSS model mice were significantly lower than those in control mice ( Figure 5A ). Il13 and Il10 mRNA expression in the lung tissue from pSS model mice was significantly increased compared with that from control mice. IL-13 and IL-10 are well known to be key cytokines for Th2 response and IL-4 ( Supplementary Figure S2 ). By contrast, there were no differences in Ifg-γ and Il6 mRNA expression of the lung tissues between control and pSS model mice ( Supplementary Figure S2 ). In addition, we evaluated the mRNA expressions of Tbx21 and Gata3, key transcription factors controlling Th1 and Th2, respectively. A significantly decreased mRNA level of Gata3 in the spleen of pSS model mice was identified when compared with that of control mice ( Figure 5B , upper panels). The Gata3 mRNA expression in the lung tissues of pSS model mice was found to be significantly increased when compared with that of control mice, whereas no significant difference was found in the case of the salivary glands ( Figure 5B , upper panels). By contrast, Tbx21 mRNA expression levels between the control and the pSS model mice were not different in the spleen, the salivary gland, or the lung tissues examined ( Figure 5B , lower panels). In order to investigate the direct effect of IL-4 on the CD23 expression in B cells, the latter were isolated from the lung tissues of control and of pSS model mice and were cultured in vitro with an anti-CD40 mAb with IL-4, or with a combination of the two, according to a previous report (27). The number of CD23⁺ B cells isolated from pSS model mice was found to be significantly increased after a stimulation with the anti-CD40 mAb or IL-4 ( Figure 5C ). A significantly enhanced number of CD23⁺ B cells was also found after the in vitro exposure of pSS model mice-driven B cells to the combination of the anti-CD40 mAb and IL-4 ( Figure 5C ). However, no change was observed in the B cells isolated from the lung tissues of control mice ( Figure 5C ). These results suggest that the Th2 condition in the lungs of the pSS model mice may play an important role in the formation of autoimmune pulmonary lesions including the B-cell infiltration.

CD4⁺ T cells, including follicular T helper (Tfh) and follicular regulatory T (Tfr) cells, play potent roles in differentiation and activation of FB cells in germinal center reaction (42). We have reported that an increased Tfh cells in the periphery contributes to the pathogenesis of autoimmune lesions in the salivary glands of pSS model mice [43]. A significantly increased proportion of the T cells, including Tfh and Tfr cells, was confirmed by flow cytometric analysis using the lung tissues from control and pSS model mice ( Supplementary Figure S3 ). Therefore, the anti-CD4 mAb was intraperitoneally administered to pSS model mice from the sixth until eighth week in order to assess whether the CD23⁺ B cells in the pulmonary lesions are controlled by CD4⁺ T cells. The depletion of the CD4⁺ T cells in the spleen and in the lungs was confirmed as a result of the anti-CD4 mAb administration, whereas no changes were found with regard to the CD19⁺ B cells of the lungs in pSS model mice treated with the anti-CD4 mAb ( Figures 6A, B ). The undertaken pathological analysis revealed that the administration of the anti-CD4 mAb has failed to improve the pulmonary inflammation in the pSS model mice ( Figure 6C ). No difference was identified with regard to the focus number/mm² of the pulmonary lesions between the isotype control mAb- and the anti-CD4 mAb-treated mice ( Figure 6D ). On the other hand, the CD23⁺ FB cells in the pulmonary lesions were found to be clearly diminished as a result of the anti-CD4 mAb administration ( Figure 6E ). The proportion of the CD23⁺ FB cells was significantly reduced by the anti-CD4 mAb administration when compared with isotype control mAb administration ( Figures 6F, G ). Our findings suggest that the differentiation of CD23⁺ FB cells in the pulmonary lesions may be controlled by CD4⁺ T cells during the development of the autoimmune pathology observed in the lungs of pSS model mice.

To understand the pathogenesis during the onset of the CD23⁺ FB cell-associated pulmonary lesions dependent on CD4⁺ T cells, anti-CD4 mAb was intraperitoneally injected into the pSS model mice from fourth until sixth weeks of their lives. Pathological analysis of the pulmonary lesions showed that the inflammatory foci of anti-CD4 mAb-administered mice were considerably smaller than those of control mice ( Figure 7A ). BALT-like structure was not observed by anti-CD4 mAb administration ( Figure 7A ). The area of the foci of anti-CD4 mAb-treated mice was significantly reduced compared with that of isotype control mAb-treated mice ( Figure 7B ). The depletion of the CD4⁺ T cells in the spleen and in the lungs was confirmed as a result of the anti-CD4 mAb administration ( Figure 7C ). The cell numbers of CD4⁺ T cells, CD19⁺ B cells, and CD23⁺ FB cells of the lung tissues in anti-CD4 mAb-treated mice were significantly decreased compared with those in isotype control mAb-treated mice ( Figure 7D ). The results suggest that the onset of CD23⁺ FB cells-associated BALT-like lesions is dependent of CD4⁺ T-cell activation in the lung of pSS model mice.