Sclerosing Sialadenitis Is Associated With Salivary Gland Hypofunction and a Unique Gene Expression Profile in Sjögren's Syndrome

Abstract

Purpose: To develop a novel method to quantify the amount of fibrosis in the salivary gland and to investigate the relationship between fibrosis and specific symptoms associated with Sjögren's syndrome (SS) using this method.

Materials and methods: Paraffin-embedded labial salivary gland (LSG) slides from 20 female SS patients and their clinical and LSG pathology data were obtained from the Sjögren's International Collaborative Clinical Alliance. Relative interstitial fibrosis area (RIFA) in Masson's trichrome-stained LSG sections was quantified from digitally scanned slides and used for correlation analysis. Gene expression levels were assessed by microarray analysis. Core promoter accessibility for RIFA-correlated genes was determined using DNase I hypersensitive sites sequencing analysis.

Results: RIFA was significantly correlated with unstimulated whole saliva flow rate in SS patients. Sixteen genes were significantly and positively correlated with RIFA. In a separate analysis, a group of differentially expressed genes was identified by comparing severe and moderate fibrosis groups. This combined set of genes was distinct from differentially expressed genes identified in lung epithelium from idiopathic pulmonary fibrosis patients compared with controls. Single-cell RNA sequencing analysis of salivary glands suggested most of the RIFA-correlated genes are expressed by fibroblasts in the gland and are in a permissive chromatin state.

Conclusion: RIFA quantification is a novel method for assessing interstitial fibrosis and the impact of fibrosis on SS symptoms. Loss of gland function may be associated with salivary gland fibrosis, which is likely to be driven by a unique set of genes that are mainly expressed by fibroblasts.

Trial registration: ClinicalTrials.gov NCT02327884.

Keywords: Sjögren’s syndrome; salivary gland hypofunction; salivary gland interstitial fibrosis; sclerosing sialadenitis; transcriptomic gene expression profile.

Conflict of interest statement

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2021 Yin, Pranzatelli, French, Zhang, Warner, Chiorini and NIDCD/NIDCR Genomics and Computational Biology Core.

Figures

Figure 1

Flow chart of study design. The purpose of this study was to investigate the application of quantification of digitally scanned slides of salivary gland interstitial fibrosis and its correlation with patients’ clinical data and transcriptome changes. We developed a novel quantification method based on measuring RIFA and compared this with the conventional interstitial fibrosis evaluation methodology by microscopic classification (4). Following validation, this method was applied to study the correlation of RIFA with Sjögren’s syndrome (SS) clinical features, such as unstimulated whole saliva flow rate. Genes associated with interstitial fibrosis development could be identified by correlating RIFA with gene expression data. These genes were then compared with the list of differentially expressed genes associated with severe fibrosis using microarray data from the SCS group (severe fibrosis) versus the NSCS group or FLS/SCS group (moderate fibrosis). This transcriptome profile that was associated with sclerosing sialadenitis in SS (RIFA-correlated genes and differentially expressed genes from SCS vs. NSCS or FLS/SCS groups) was compared with differentially expressed genes in IPF. Change in expression in this group of genes was further analyzed for common transcriptional factors binding motifs, cellular source of expression, and core promoter openness.

Figure 2

RIFA can be used to numerically quantify interstitial fibrosis in LSG from Sjögren’s syndrome patients. Masson’s trichrome staining was used to detect interstitial fibrosis in LSGs from Sjögren’s syndrome patients. Interstitial fibrosis areas and whole gland areas were measured using Aperio Scan Scope software. Non-fibrosis areas (outlined in yellow) that were included in the interstitial fibrosis areas (outlined in green) were subtracted to obtain accurate measurement. Area less than 50µm2 (orange arrows as example areas) were not excluded. RIFA was quantified as total interstitial fibrosis area (in mm2) per mm2 total glandular size on the section [(green circled areas – yellow circled areas)/total area].

Figure 3

RIFA assessment corresponds with categorical classification of salivary gland interstitial fibrosis. Left: Representative images to detect interstitial fibrosis in the LSG from four different microscopic groups comprising FLS, NSCS, FLS/SCS and SCS. Right: Comparison of mean (±SEM) RIFA in four different microscopic groups: FLS, NSCS, FLS/SCS and SCS (n = 5/group). Statistical significance was determined with unpaired Student’s t test.

Figure 4

RIFA correlates with unstimulated whole saliva flow rate in Sjögren’s syndrome patients. (A) Correlation of RIFA and USW flow rate in Sjögren’s syndrome (SS) patients was analyzed using linear regression analysis. A significant negative correlation was found between RIFA and UWS flow rate (n = 20). (B) Comparison of mean (± SEM) UWS flow rate in different microscopic groups based on mean (± SEM) of RIFA measurement (FLS, NSCS, FLS/SCS, and SCS).

Figure 5

Differentially expressed genes associated with severe interstitial fibrosis are conserved among RIFA-correlated genes. Venn diagram to identify common differentially expressed genes in SCS group (severe fibrosis) compared with NSCS or FLS/SCS group (moderate fibrosis). Fifteen common upregulated differentially expressed genes from SCS group compared with moderate interstitial fibrosis NSCS group or FLS/SCS group were identified. Of these 15 genes, 8 genes were also significantly correlated with RIFA (highlighted in yellow, refer to Table 1).

Figure 6

RIFA-correlated gene expression is primarily localized to fibroblasts in LSG from Sjögren’s syndrome and non-Sjögren’s patients. (A) Single-cell RNA sequencing analysis was used to quantify expression of RIFA-correlated genes in different cell types within LSGs from Sjögren’s syndrome patients and non-Sjögren’s subjects. Data shown are mean(± SEM) of “percent of total expression” of each gene within each cell type. Data revealed that RIFA-correlated genes were, on average, more highly expressed in fibroblasts than in any other cell type. (B) Assessment of “percent of total expression” of individual genes within each cell type revealed that eight genes were preferentially expressed by fibroblasts, which is reflected in increased average expression of list of genes in fibroblasts. Gene expression levels are represented on a color scale from high (red) to low (blue). The sum of values in a row equals 100% of the expression.

Figure 7

Core promoter openness is significantly higher for RIFA-correlated genes in lung fibroblasts than in whole lung tissue. DNase-seq reads were used to estimate chromatin accessibility of promoters of genes associated with fibrosis, all promoters, and the whole genome (as background) in both lung tissue and lung fibroblasts. Data shown are promoter coverage (mean ± SEM). Student’s t test was used to compare promoter coverage between pairs of groups.

Figure 8

Interconnection of RIFA-correlated genes with epithelial-mesenchymal transition and salivary gland interstitial fibrosis formation in Sjögren’s syndrome. Illustration of the hypothesis of the EMT and interstitial fibrosis formation in Sjögren’s syndrome regulated by the16 RIFA-correlated genes. RIFA-correlated genes are identified in red. Figure is created with BioRender.com.