Dynamics of Human Palatal Wound Healing and the Associated Microbiome

Abstract

Wound healing in the oral mucosa is superior to that in the skin, with faster wound closure accompanied by reduced inflammation, less angiogenesis, and minimal scar formation. A well-characterized oral wound model is critical to investigating the mechanisms of oral wound closure and the efficacy of various clinical interventions. Currently, there are a few human oral wound models, although none of them are well characterized. In the present study, we describe and characterize a human hard palate wound healing model. A 3.5-mm circular and two 1 × 5-mm rectangular full-thickness wounds were made in the first and second molar region, 5 mm from the gingival margin, on the hard palate of human subjects. The circular wound was used to monitor wound closure and collect swabs for a microbiome analysis via 16s rRNA sequencing. The rectangular wounds were biopsied and the tissue was used to evaluate the gene expression of wound healing-related mediators by real-time polymerase chain reaction. Saliva was also collected to examine the protein levels of similar molecules by enzyme-linked immunosorbent assays. Circular wounds were nearly closed on day 7 after wounding. Significant changes in the gene expression of inflammatory cytokines, growth factors, antimicrobial peptides, and extracellular matrix-related molecules were identified in day 1 and day 3 wound tissue and compared with unwounded tissue on day 0. Changes in the protein levels of various mediators were limited in the saliva. In addition, alpha diversity, beta diversity, and differential microbiome analysis demonstrated significant changes in bacterial colonization of the wound surface over time compared with unwounded mucosa. In summary, we comprehensively characterize a human hard palate wound-healing model that details the dynamic changes of wound closure, levels of wound healing-related mediators in the wound and saliva, and the oral wound microbiome.

Keywords: hard palate; microbiomes; oral wound model; regeneration; wound mediator.

Figure 1.

Timeline of healing in a circular excisional hard palate wound model. (A) Self-reported pain perception scale scores 2 h after wounding/biopsy on D0, D1, D3, and the morning of D5. (B) Assessment of wound size following injury, reported as an absolute area in mm² (left) and a proportion of wound size relative to the original size of the wound (right). (C) Representative photographs of the excisional circular wound model performed on the hard palate. Filter paper placed over the wound was used to account for variability in image capture. (D) Close-up images of 3.5-mm wounds. Scale bar: 3.5 mm. Rectangular wounds were also made along the gingival margin for D1 and D3 biopsies. n = 22 for each time point, *p < 0.05.

Figure 2.

Relative expression of wound healing-related genes on days 1 and 3 postwounding. Real-time polymerase chain reaction was performed on total RNA extracted from wounded hard palate tissue. n = 22 for each time point. *p < 0.05, #p < 0.01.

Figure 3.

Salivary changes in protein levels of wound healing–related mediators and immunoglobulin A. Saliva samples were collected on D0 (before wounding), D1, D3, and D5 postwounding. Levels of EGF, VEGFA, IL-10, and IgA were assessed using enzyme-linked immunosorbent assay. n = 22 for each time point. p < 0.05, #p < 0.01.

Figure 4.

Alpha and beta diversities of bacteria at the genus level on wounded and unwounded mucosa. (A) Alpha diversity of the wounded (W) and unwounded (U) sites and between different days postwounding on the same site. (B) Comparison of beta diversity between different days postwounding on the W or U sites. (C) Comparison of beta diversity between the W and U sites on different days postwounding. n = 13 for each time point. *p < 0.05, #p < 0.01.

Figure 5.

Differential abundance of bacterial taxa. (A) The bar plots of the top 10 abundant bacterial taxa at the genus level on the wounded (W) and unwounded (U) sites. (B) Percentages of each of the top 10 abundant bacterial taxa. n = 13 for each time point. *q < 0.05.