Periodontal bacterial supernatants modify differentiation, migration and inflammatory cytokine expression in human periodontal ligament stem cells

Abstract

Periodontal ligament stem cells (PDLSC) play an important role in periodontal tissue homeostasis/turnover and could be applied in cell-based periodontal regenerative therapy. Bacterial supernatants secreted from diverse periodontal bacteria induce the production of cytokines that contribute to local periodontal tissue destruction. However, little is known about the impact of whole bacterial toxins on the biological behavior of PDLSC. Therefore this study investigated whether proliferation, migration, inflammatory cytokines expression and transcriptional profile would be affected by exposure to endotoxins from bacterial species found in the subgingival plaque. PDLSC were cultured with the following bacterial supernatants: S. mutans, S. anginosus, P. intermedia, F. nucleatum, P. gingivalis and T. denticola. These supernatants were prepared in dilutions of 1:1000, 1:500, 1:300 and 1:50. Using quantitative RT-PCR, gene expression of selected inflammatory cytokines (IL-6, IL-8 and IL-1β) and cell-surface receptors (TLR2, TLR4) showed upregulation of ≈2.0- to 3.0-fold, when exposed to P. intermedia, F. nucleatum, P. gingivalis and T. denticola. However, supernatants did not affect proliferation (MTT) and migration (wound scratch assays) of PDLSC. Next generation RNA sequencing confirmed modified lineage commitment of PDLSC by stimulating chondrogenesis, adipogenesis and inhibition of osteogenesis under P. gingivalis supernatant treatment compared to control. Taken together, this study shows stem cell immunomodulatory response to different periodontal bacteria supernatant and suggests that stem cell transcriptional capacity, migration/proliferation and osteogenesis may differ in the presence of those pathogens. These results bring into question stem cell contribution to periodontal tissue regeneration and onset of inflammation.

Fig 1. Bacterial supernatants modified expression of immune-related molecules in PDLSC.

Gene expression levels determined by RT-PCR analysis showed upregulation of IL-1β (A) IL-8 (B), IL-6 (C), TLR2 (D) and TLR4 (E) in PDLSCs. Representative blots of at least three independent experiments are shown. Statistically significant differences presented as fold change relative to the untreated negative control (control, open columns) * p < 0.05. Mean ± S.D.

Fig 2. P. gingivalis supernatant impairs the osteogenic potential of PDLSC.

(A) IL-1 β, IL-6 and IL-8 production was measured using ELISA after PDLSCs were treated with P. gingivalis (P.g.) supernatant’s dilutions (1:500 and 1:50), or untreated cells as control simultaneously for 24 h. IL-1 β, IL-6 and IL-8 is expressed in pg/ml ( ± standard deviation). (B) Western blot analysis showed the protein expression of TLR2, TLR4 and GAPDH was used as the internal control. (C) Confluent PDLSC were stained with Alizarin Red after 21 days of cultivation in osteogenic medium containing P. gingivalis supernatant’s dilutions (1:500 and 1:50) and Alizarin Red was then extracted and measured for light absorbance at 405 nm. *p < 0.05 as determined by unpaired two-tailed t tests. Data from three biologically independent replicates.

Fig 3. Viability and proliferation of PDLSC exposure to P. gingivalis and T. denticola.

Upon supernatant treatment (untreated, 1:1000, 1:500, 1:300 and 1:50), PLDSC viability and proliferation percentages did not alter throughout periods of time (24 h, 48 h and 72 h) when compared with untreated controls (open columns). ns: non-significant. Data are shown as the mean ± S.D of 3 samples (3 wells each) from one representative experiment.

Fig 4. PDLSC migration was increased with supernatants of P. gingivalis and T. denticola by scratch wound healing assay.

Representative images are shown from 3 independent experiments and light gray area define the areas lacking cells (Scale bar 120 μm). Images were analyzed using ImageJ software to calculate wound area. Data is expressed as the mean values of percentage wound closure relative to the corresponding 0 h time point and represent the mean percentage closure ± SEM (n = 3): *p < 0.01 vs. time-matched treated control for each time-point.

Fig 5. Transcriptomic analysis elucidates PDLSC response to bacterial supernatant-induced bioactivity.

(A) Bacterial supernatant treated groups (1:500 and 1:50 dilutions) and untreated group (Ctrl) were marked. Heatmap showed significant gene profile for RNAseq data. Normalized log2 scale shown. (B) Volcano analysis was performed for significantly changed genes in the dilution of 1:50 supernatant treated group (red dots represented upregulated genes and blue dots represented downregulated genes, Fold change > 2.0, p < 0.05). (C) Significant GO terms of associated biological processes from differentially regulated genes from a ranking by enrichment scores (p < 0.05). Adipogenesis and chondrogenesis processes identified by red columns. Enrichment scores were calculated as −log₁₀ (P-value), (D) Comparison between RNAseq and qPCR gene expression showed similar values.