Salivary proteins offer insights into keratinocyte death during aphthous stomatitis. A case-crossover study

Abstract

Background: The death of oral keratinocytes is a crucial step in the emergence of recurrent aphthous stomatitis (RAS, also known as aphthae or aphthous ulcers). Since there are no experimental models available to research aphthous ulcers, little is understood about this process. We hypothesize that saliva can be a data bank of information that offers insights on epithelial damage.

Methods: In this case-crossover study, we assessed the salivary proteome of patients with RAS (n = 36) in the presence and absence of ulcers using discovery proteomics and bioinformatics. Additionally, we contrasted these patterns with those of healthy individuals (n = 31) who had no prior aphthous ulceration.

Results: Salivary proteome showed that during the ulcerative phase, controlled cell death was downregulated. Due to its ability to distinguish between individuals with and without ulcers, the ATF6B protein raises the possibility that endoplasmic reticulum (ER) stress is responsible for the damage that leads to the death of oral keratinocytes. The high abundance of TRAP1 and ERN1 matches with this biological discovery. The type of death is immunogenic, according to the functional data found in a cell death database.

Conclusion: We identified a cellular process that can lead to the death of oral keratinocytes in the etiopathogenesis process of RAS. Future studies should be conducted to identify what is responsible for the increase in ER stress signaling that would lead to an anti-cell death response.

Keywords: Endoplasmic reticulum stress; Proteomics; Recurrent aphthous stomatitis; Regulated cell death.

Fig. 1

Salivary proteome of recurrent aphthous stomatitis biological processes. (A) Circles represent the number of salivary proteins from the respective pooled samples (sets 1, 2 and 3). The numbers on the right indicate the numbers of proteins exactly found in each set. “All” corresponds to all proteins that were detected in at least one set. (B) The majority of the salivary proteins discovered were proteins in common, as shown by a Venn diagram. (C) Bar graph shows the fraction of saliva to which proteins of each group belong (The Human Salivary Proteome Wiki). (D) FunRich categorized the biological processes associated with salivary proteins. Cumulatively, the data show us that there is an increase in the number of proteins involved in the negative regulation of apoptosis

Fig. 2

ATF6B provides the best discrimination between the groups. (A) Multiple-sample comparison. The table shows the significant comparisons obtained after 1,011 contrasts. In the presence and absence of ulcers, the abundance of ATF6B and TRAP1 are statistically different (ulcerative stage vs. healthy controls, ulcerative stage vs. remission stage). (B) Grouping was generated using the Euclidean distance method. Six proteins grouped together the states of absence of aphthous ulcers. ATF6B stands out since it generates a dendrogram on its own. (C) The predictive power of proteins was tested with a discriminant analysis. In the upper panel, canonical discriminant functions simultaneously enter all independent variables that satisfy tolerance criteria. All original grouped cases were correctly classified (best ATF6B + and TRAP1-, Wilks’ Lambda < 0.05). FA83E and GFRA1 were not used in the analysis. Bottom panel uses stepwise analysis to control variable entry and removal. All original grouped cases were correctly classified (best TRAP1 + and ATF6B-, Wilks’ Lambda < 0.05). HV321 and GFRA1 were not used in the analysis

Fig. 3

ATF6B is over-expressed in the saliva of RAS patients. During the clinical course of recurrent aphthous stomatitis, salivary ATF6B was confirmed. The ulcerative phase (n = 36) and the remission (healing, n = 36) phase are depicted in the left panels. The recurrence of new ulcers (n = 15) and the healthy controls (n = 31) are shown in the right panels. (A) Dot blotting (10 µL/dot) showed that ATF6B was more frequent in ulcerative stage samples (13/15 = 0.86) and less frequently in healthy controls (24/31 = 0.77). As a positive control we used the ATF6b antigen (APREst 83,224, Sigma). NC, negative control. (B) Dot blot quantification. Recurrent aphthous stomatitis groups demonstrate more ATF6B staining compared to the control. The multiple comparison between groups showed significant differences *p<0.05, ** p<0.01. The uncropped scans of the gels are shown in Supplementary Fig. 1, dot blot quantification on Supplementary dataset file 2 and western blot analysis result is shown in Supplementary Fig. 2

Fig. 4

Abundance ranking of proteins involved in cell death and ER stress. The ranking’s possible positions range from 1 to 1,011. ATF6B and ERN1 are involved in parthanatos, a type of programmed cell death, according to the XDeathDB database (https://pcm2019.shinyapps.io/XDeathDB/). Independent of the presence or absence of ulcers, patients with RAS have a similar expression profile. Acute-phase response proteins including HBA, HBB, HBD, and HBG1 continue to be abundant even during remission. This could indicate that subclinical inflammatory phenomena are still present. APP, APOE, MADD, ATF6B, ERN1 and TRAP1 are abundant during the presence of ulcers