Proteomic profiling of host-biofilm interactions in an oral infection model resembling the periodontal pocket

Abstract

Periodontal infections cause inflammatory destruction of the tooth supporting tissues. We recently developed a dynamic, in vitro periodontal organotypic tissue model in a perfusion bioreactor system, in co-culture with an 11-species subgingival biofilm, which may recapitulate early events during the establishment of periodontal infections. This study aimed to characterize the global proteome regulations in this host-biofilm interaction model. Semi-quantitative shotgun proteomics were applied for protein identification and quantification in the co-culture supernatants (human and bacterial) and the biofilm lysates (bacterial). A total of 896 and 3363 proteins were identified as secreted in the supernatant and expressed in the biofilm lysate, respectively. Enriched gene ontology analysis revealed that the regulated secreted human tissue proteins were related to processes of cytoskeletal rearrangement, stress responses, apoptosis, and antigen presentation, all of which are commensurate with deregulated host responses. Most secreted bacterial biofilm proteins derived from their cytoplasmic domain. In the presence of the tissue, the levels of Fusobacterium nucleatum, Actinomyces oris and Campylobacter rectus proteins were significantly regulated. The functions of the up-regulated intracellular (biofilm lysate) proteins were associated with cytokinesis. In conclusion, the proteomic overview of regulated pathways in this host-biofilm interaction model provides insights to the early events of periodontal pathogenesis.

Figure 1. Venn diagram of identified protein numbers in culture supernatants.

Protein numbers for each category is depicted. Details of human and bacterial proteins in each group are listed in Supplementary Information Table S1 and Table S2, respectively.

Figure 2. Heat map of differentially regulated human proteins in culture supernatants.

Hierarchical clustering of significantly altered proteins (LogFC P<0.05) were performed by the average linkage method using the Heml software. The colours in the map display the mean value for individual proteins (represented by a single row) within each experimental group (represented by a single column). Expression values are shown as a colour scale, with higher values represented by red and lower represented by blue. The colour scale bar gradient is shown at the right corner of the figure.

Figure 3. Analysis of the most significantly regulated network: relationships of the regulated human proteins.

Networks of protein interactions in regulated human proteins (including Calreticulin, AP-2 alpha subunits, AHNAK, LTBP2, SPRR1A) using MetaCore. The networks between the regulated proteins were calculated based on the “analyse network” algorithm value, then the network maps of their putative protein interactions and related proteins were predicated accordingly from MetaCore database. The regulated human proteins from the list (Table S3) are denoted in smaller circles. Smaller red and blue circles indicated up-regulated and down-regulated proteins, respectively. Green arrows indicate activation of the corresponding proteins; red arrows, inhibition; and gray arrows, unspecified. The cyan highlighted edges in the legend represent the canonical pathways, as recorded by MetaCore. For detailed network symbol legend, see supplementary Figure S4 for full annotations of nodes.

Figure 4. Analysis of the second most significantly regulated network: relationships of the regulated human proteins.

Networks of protein interactions in regulated proteins (including Plectin 1, hnRNP K, Envoplakin, TGM1, SPRR1B) using MetaCore. The networks between regulated proteins were calculated based on the “analyze network” algorithm value, then the network maps of their putative protein interactions and related proteins were predicated accordingly from the MetaCore database. The regulated human proteins from our list (Table S3) are denoted in smaller circles. Smaller red and blue circles indicated up-regulated and down-regulated proteins, respectively. Green arrows indicate activation of the corresponding proteins; red arrows, inhibition; and gray arrows, unspecified. For detailed network symbol legend, see Supplementary Figure S4 for full annotations of nodes.

Figure 5. Analysis of the third most significantly regulated network: relationships of the regulated human proteins.

Networks of protein interactions in regulated human proteins (including IDE, Kappa chain (Ig light chain), Cofilin, MNDA, MHC class I proteins) using MetaCore. The networks between regulated proteins were calculated based on the “analyze network” algorithm value, then the network maps of their putative protein interactions and related proteins were predicated accordingly from the MetaCore database. The regulated human proteins from our list (Table S3) are denoted in smaller circles. Smaller red and blue circles indicated up-regulated and down-regulated proteins, respectively. Green arrows indicate activation of the corresponding proteins; red arrows, inhibition; and gray arrows, unspecified. For detailed network symbol legend, see Supplementary Figure S4 for full annotations of nodes.

Figure 6. Annotation of regulated secreted bacterial protein functions by Gene Ontology (GO) terms enrichment.

The GO terms from all regulated bacterial functions were categorized into biological process, molecular function, and cellular component as displayed in pie charts use tissue + biofilm group, compared to the biofilm group alone. The numbers of GO terms for each of the three categories are shown, whereas the proportion of each specific subcategory is also provided. Subcategories with GO terms less than 2% are classified as “other”.

Figure 7. Venn diagram of identified protein numbers in biofilm lysates.

Protein numbers for each category is depicted. Details of proteins in each group are listed in Supplementary Information Table S5.

Figure 8. Annotation of regulated bacterial protein functions in biofilm lysates by enrichment of Gene Ontology (GO) terms.

The GO terms from all regulated bacterial functions were categorized into biological process, molecular function, and cellular component as displayed in pie charts use tissue + biofilm group, compared to the biofilm group alone. The numbers of GO terms for each of the three categories are shown, whereas the proportion of each specific subcategory is also provided. Subcategories with GO terms less than 2% are classified as “other”.