Targeting alarmin release reverses Sjogren's syndrome phenotype by revitalizing Ca2+ signalling

Abstract

Background: Primary Sjogren's syndrome (pSS) is a systemic autoimmune disease that is embodied by the loss of salivary gland function and immune cell infiltration, but the mechanism(s) are still unknown. The aim of this study was to understand the mechanisms and identify key factors that leads to the development and progression of pSS.

Methods: Immunohistochemistry staining, FACS analysis and cytokine levels were used to detect immune cells infiltration and activation in salivary glands. RNA sequencing was performed to identify the molecular mechanisms involved in the development of pSS. The function assays include in vivo saliva collection along with calcium imaging and electrophysiology on isolated salivary gland cells in mice models of pSS. Western blotting, real-time PCR, alarmin release, and immunohistochemistry was performed to identify the channels involved in salivary function in pSS.

Results: We provide evidence that loss of Ca²⁺ signaling precedes a decrease in saliva secretion and/or immune cell infiltration in IL14α, a mouse model for pSS. We also showed that Ca²⁺ homeostasis was mediated by transient receptor potential canonical-1 (TRPC1) channels and inhibition of TRPC1, resulting in the loss of salivary acinar cells, which promoted alarmin release essential for immune cell infiltration/release of pro-inflammatory cytokines. In addition, both IL14α and samples from human pSS patients showed a decrease in TRPC1 expression and increased acinar cell death. Finally, paquinimod treatment in IL14α restored Ca²⁺ homeostasis that inhibited alarmin release thereby reverting the pSS phenotype.

Conclusions: These results indicate that loss of Ca²⁺ signaling is one of the initial factors, which induces loss of salivary gland function along with immune infiltration that exaggerates pSS. Importantly, restoration of Ca²⁺ signaling upon paquinimod treatment reversed the pSS phenotype thereby inhibiting the progressive development of pSS.

Keywords: Ca2+ signalling; ER stress; alarmins; immune cell activation; primary Sjogren's syndrome; salivary gland dysfunction.

FIGURE 1

Loss of salivary glands precedes immune cell migration: Part (A) shows hematoxylin and eosin (H&E) staining in salivary glands from control and IL14α Tg mice among various age‐groups (images shown are representative of 3–4 animals from each age‐group). The presence of immune cells is marked by a yellow border. (B) Representative confocal images showing apoptosis (using tunnel staining) in salivary glands from control and IL14α Tg mice. Age‐groups of animals tested are indicted in the figure. “*” Indicates cell death in acinar cells. Part (C) shows quantification of the tunnel⁺ cells in various conditions from 3 to 5 individual samples, *p ≤ .001 (Student's t test). (D) Confocal images showing the presence of CD11c (marker for dendritic cells), CD3 (marker for T cells) and B220 (marker for B cells) in salivary glands from control and IL14α Tg mice. Age‐groups used are indicted in the figure, and the images shown are representation of three separate experiments. (E) Lymphocytes were isolated from salivary glands of IL14α mice (1, 6 and 12 months old) and control littermates. Cells were stained with anti‐CD19 (PerCP‐Cy5.5), anti‐CD4 (phycoerythrin [PE]) and anti‐CCR4 (allophycocyanin [APC]) using multi‐colour flow cytometry. Gates were used to determine the percentage of infiltrating immune cells in the salivary gland tissues of the mice, CD19⁺ high (B‐cell marker) as indicated and the combined CD4⁺ high (T cell marker) and CCR4⁺ high (TH17 cell marker) for each mice group and control. Bar graphs represent the number of total number cells marked positively for the specific markers. Data shown are representative of three‐independent experiments with similar results. Bar graphs depict average ± SD for relative values, ***p ≤ .001, NS = non‐significant (Student's t test).

FIGURE 2

Age‐dependent decrease in salivary gland function is observed in IL14α Tg mice: (A) saliva secretion was induced by pilocarpine in 1‐month‐old control (wild‐type [WT]) and IL14α Tg mice and plotted. The data presented are representative of 6–8 animals in each group. Saliva was collected every 4 min, and total saliva secreted is plotted as line graph. (B) Quantification of total saliva (6–8 mice) secreted in 1‐month‐old control (WT) and IL14α Tg mice is shown as bar graph. Error bars represent means ± SE. NS indicates no significant difference between the two groups. (C) Saliva flow rate in 1‐month‐old control (WT) and IL14α Tg mice is shown as line graph, which showed a time‐dependent decrease in saliva secretion. (D) Quantification of saliva flow rate (8–10 mice) in 1‐month‐old control (WT) and IL14α Tg mice is shown as bar graph. Error bars represent means ± SE. NS indicates no significant difference between the two groups. (E) Saliva secretion in 6‐month‐old control (WT) and IL14α Tg mice (6–10 mice) and plotted. (F) Quantification of total saliva secreted (6–8 mice for each group) in 6‐month‐old control (WT) and IL14α Tg mice is shown as bar graph. Error bars represent means ± SE. *p ≤ .001 (Student's t test). (G) Saliva flow rate in 6‐month‐old control (WT) and IL14α Tg mice is shown. Quantification of saliva flow rate (6–8 mice) in 6‐month‐old control (WT) and IL14α Tg mice is shown as bar graph (H). Error bars represent means ± SE. *p ≤ .001 (Student's t test or analysis of variance [ANOVA]). (I) Secretion of saliva and its quantification is shown as bar graph in (J) in 12‐month‐old control (WT) and IL14α Tg mice (6 mice each). Error bars represent mean ± SE. *p ≤ .001 (Student's t test or ANOVA). (K) Saliva flow rate in 12‐month‐old control (WT) and IL14α Tg mice is shown. Quantification of saliva flow rate (8 mice) in 12‐month‐old control (WT) and IL14α Tg mice is shown as bar graph (L). Error bars represent mean ± SE. *p ≤ .001.

FIGURE 3

Age‐induced gradual loss of Ca²⁺ signalling in IL14α Tg mice: (A) principal component analysis (PCA) plot of normalized FPKM of RNA‐seq from duplicate samples of various age‐groups in control and IL14α Tg mice. (B) Heat map and clustering based on hierarchy showing differential expression of genes in control and IL14α Tg mice (at age 1, 6, 12 and 17 months). (C) Differential expression of genes in the salivary glands between control (6 months) and IL14α Tg mice (6 months). Volcano plot indicating up‐regulated and down‐regulated genes in control and IL14α Tg mice is shown as inset. Top gene ontology (GO) terms showing top 20 significant biological process and their molecular functions, in various pathways, are labelled. The data used are the −log10 p values from the differential expression (decreased) list of 1 versus 6 months (D) and increased in 12 versus 1 month (E) old IL14α Tg mice. (F) Representative plots of Ca²⁺ imaging performed on Fura‐2AM loaded primary acinar cells isolated from submandibular glands of 1‐month‐old control (wild‐type [WT]) and IL14α Tg mice (n = 3 mice each). Cells were stimulated by Tg (1 µM) followed by the addition of 1 mM external Ca²⁺. Bar graphs (shown in right) indicate quantification of the endoplasmic reticulum (ER) Ca²⁺ release and Ca²⁺ entry in data from more than 90–120 cells/areas in each condition and is plotted as mean ± SEM. NS indicates no significant difference between the two groups. (G) Individual traces showing changes in Ca²⁺ imaging in primary acinar cells isolated from submandibular glands of 3‐month‐old control (WT) and IL14α Tg mice (n = 3 mice each). Quantification of the ER Ca²⁺ release and Ca²⁺ entry from more than 120 cells in each condition is plotted as mean ± SEM in the bar graphs. *p < .05 indicates values that are significantly different from WT acinar cells (Student's t test). (H) Ca²⁺ imaging traces from primary acinar cells isolated from 6‐month‐old control (WT) and IL14α Tg mice. Bar graphs indicate quantification of the ER Ca²⁺ release and Ca²⁺ entry in data from more than 100 cells in each condition and is plotted as mean ± SEM. *p < .05 indicates values that are significantly different from WT acinar cells (Student's t test). (I) Ca²⁺ imaging trace in primary acinar cells isolated from submandibular glands of 12‐month‐old control (WT) and IL14α Tg mice. Quantification of the ER Ca²⁺ release and Ca²⁺ entry from more than 90 cells in each condition is plotted as mean ± SEM in the bar graphs. *p < .05 indicates values that are significantly different from WT acinar cells (Student's t test).

FIGURE 4

Transient receptor potential canonical‐1 (TRPC1)‐like currents are decreased over time in IL14α Tg mice: (A) Salivary gland (submandibular gland) lysates from control (wild‐type [WT]) and IL14α Tg mice samples (n = 4–6) from various age‐groups were subjected to SDS–PAGE and immunoblotted with respective antibodies as labelled. (B) Relative expression of TRPC1, TRPC3, TRPC4, TRPC5 and TRPC6 between WT mice and IL14α Tg mice (n = 3). mRNA levels of various TRPC genes in control and IL14α Tg mice (6 months). “*” Indicates significance p < .05. (C) Tg‐induced currents (at −80 mV) were evaluated in acinar cells isolated from 1‐month‐old control (WT) and IL14α Tg mice. Representative I/V curves developed from acinar cells isolated from 1‐month‐old control (WT) and IL14α Tg mice are potted. (D) The average current intensities (mean ± SEM) from 9 to 12 individual cells in each condition are shown as bar graph. NA indicates no significant difference between the two groups. (E) Tg‐induced currents (at −80 mV) were evaluated in acinar cells isolated from 5‐month‐old control (WT) and IL14α Tg mice. Representative I/V curves developed in these conditions are potted in part (F). pA/pF represents currents in picoamperes per picofarad. (F) The average current intensities (current/mean ± SEM) from 8–10 individual cells are shown as bar graph. * p < .05 indicates values that are significantly different from untreated WT acinar cells (Student's t test). (G) Tg‐induced currents (at −80 mV) were evaluated in acinar cells isolated from 12‐month‐old control (WT) and IL14α Tg mice. Representative I/V curves developed in these conditions are potted in part (G). (H) The average current intensities (mean ± SEM) from 8–15 individual cells are shown as bar graph. p < .05 indicates values that are significantly different from untreated WT acinar cells (Student's t test). Values are expressed as mean ± SE. In general, the “*”, “**” and “***” indicate p values less than .05, .01 and .001, respectively. (I) Confocal images showing the expression of TRPC1 in salivary glands from age‐matched (55 years old) control and primary Sjogren's syndrome (pSS) samples. Images shown are representation from three individual samples. Part (J) shows hematoxylin and eosin (H&E) staining in salivary glands from control and TRPC1^−/− (6‐month old, images are representation of three individual samples). (K) Bar graphs represent the total number of immune cells marked positively for the specific immune markers used in submandibular glands of control and TRPC1^−/−. “***” Indicates significance p < .001. Data shown are representative of three‐independent experiments with similar results.

FIGURE 5

DAMPs released by human submandibular gland (HSG) cells after treatment with endoplasmic reticulum (ER) stress inducers promote immune cell migration and increase of inflammatory response. (A) Schematic representations of migration assay where 1 × 10⁶ HSG cells were plated in six wells plates in media containing tunicamycin (10 µM), brefeldin A (BFA) (10 µM), SKF96365 (10 µM) treated for 12 h. An insert is placed inside the wells with .5 × 10⁶ primary macrophages cells on the top, and then treatment is added in the lower part. Cells that migrated through the pores to the lower side of the insert membrane were fixed and counted at 40× magnification. (B) Data shown are representative of three‐independent experiments with similar results. Bar graphs depict average ± SD for positive cells, ***p ≤ .001 (Student's t test). (C) Damage‐associated molecular patterns (DAMPs) were analysed, by colorimetric analysis in supernatants of HSG cells after various treatments. Data shown are representative of three‐independent experiments with similar results. Bar graphs depict average ± SD for relative values, ***p ≤ .001 (Student's t test and/or analysis of variance [ANOVA]). (D) Pro‐inflammatory cytokines levels were analysed, by colorimetric analysis in supernatants (HSG cells and migrated macrophages) after treatment. Bar graphs (from three‐independent experiments) depict average ± SD for relative values, ***p ≤ .001 (ANOVA). (E) Confocal images showing the expression of CHOP (marker for ER stress) and tunel staining in salivary glands from age‐matched control and primary Sjogren's syndrome (pSS) samples. Images shown are representation of three individual experiments. (F) Confocal images showing HMGB1 expression in salivary glands from 6‐month‐old control and IL14α Tg mice. Images shown are representation of three individual experiments. (G) Bone marrow‐derived primary macrophages were also treated with supernatants of HSG cells treated with ER stress inducers (tunicamycin or BFA) or Ca²⁺ channel blocker (SKF‐96365) with and without paquinimod for 12 h, and then the media was used as chemoattracts for migration stimulation. Immune cell migration was measured in response to various treatments exposed for 5 h, at which time, migrating cells were quantified. Data shown are representative of three‐independent experiments with similar results. Bar graphs depict average ± SD for positive cells, ***p ≤ .001 (Student's t test).

FIGURE 6

Paquinimod reduces infiltration of inflammatory cells and restore salivary gland function in IL14α transgenic mice. (A) Representative plots of Ca²⁺ imaging performed on Fura‐2AM loaded human submandibular gland (HSG) cells with bath application of paquinimod (300 µm) or pre‐treated with paquinimod (300 µm for 30 min) as indicated in part (B). Bar graphs show quantification of Ca²⁺ entry under control and paquinimod‐treated conditions from 120 to 140 cells. (C) Six‐month‐old IL14αTg mice were treated with paquinimod (n = 8) or saline (control, n = 8) for 9 weeks. Representative hematoxylin and eosin (H&E)‐stained salivary gland sections (n = 3) from saline‐treated IL14α Tg mice (control), or IL14α Tg mice treated with paquinimod for 9 weeks, showed few to no infiltration of immune cells. Scale bars shown are 100 µm (D) Total saliva secretion (n = 8) and rate of saliva in saline or paquinimod‐treated (for 9 weeks) IL14α Tg mice. (E) Quantifications of total saliva and saliva flow rate in each condition are shown as bar graph (average ± SD, *p ≤ .05 (Student's t test). (F) Single cell suspensions from salivary glands analysed by flow cytometry, and cells were stained with monocytes marker (CD11b+ AF594), dendritic cells marker (MHC class II+ AF488), B‐cell marker (CD19+ allophycocyanin [APC]) and neutrophils marker (Ly6G+ eFluor 450), showing a significant decrease in immune infiltrated in paquinimod‐treated IL14αTg mice. Data shown are representative of three‐independent experiments with similar results. Bar graphs (shown below) depict average ± SD for relative values, ***p ≤ .001 (Student's t test).

FIGURE 7

Schematic showing that loss of transient receptor potential canonical‐1 (TRPC1)‐mediated Ca²⁺ entry induces endoplasmic reticulum (ER) stress. Loss of Ca²⁺ signalling induces the release of alarmins needed for immune cell infiltration in salivary tissues. Moreover, paquinimod treatment restores Ca²⁺ entry, prevents alarmin release and reverts the development of primary Sjogren's syndrome (pSS) phenotype.