Changes in Transcriptome Profile During Experimental Gingivitis Induction and Resolution

Objective: We conducted a 28 day experimental gingivitis model using stents to explore the changes in gingival gene expression profiles. Methods: At baseline (Day-0) 15 healthy subjects were given two posterior sextant acrylic stents to be worn during tooth brushing to promote plaque overgrowth during the 28 day gingivitis induction period. Patients were followed weekly and at Day-28, stents were discontinued, subjects received prophylaxis and oral hygiene was re-established. Interproximal tissue biopsies were collected at Day-0, Day-28 and at resolution Day-35 for mRNA extraction and transcriptome analyses by Affymetrix array (U133 Plus 2.0, ~54500 probe sets). Exploratory analyses were performed (p<0.05) on 40 arrays to examine time-dependent changes as well as gene-symptom associations. Results: We found a significant change in gene expression of 4108 genes comparing Day-0 to Day-28, which was greater than the chance alone expected finding of 2731 probe sets at Ą=0.05. At Day 28, 88 genes were up-regulated by >1.5-fold and 27 by>2.0 fold. Several gene pathways demonstrated significant changes during induction that were reversed by treatment by Day-35. These include pathways associated with inflammation, oxidative stress, T-cell activation and wound repair that involved most cellular compartments including epithelial cells, vascular cells, fibroblasts and immunocytes. Pathway analyses consistently indicate the temporal modulation of chemokines especially CCL4, 5 and CxCL2, 3, 8, 10, 11, as well as increases in Interleukin-1-alpha and beta. Neutrophil recruitment and activation appeared to dominate over TLR-4/LPS activation processes. The TLR-7 receptor was the major toll-like receptor that was upregulated. The neutrophil neurogenic inflammatory protein prokinecticin-2 was upregulated 4.6 fold at p=0.006. Conclusion: In experimental gingivitis the gene expression profile suggests a dominant chemokine and neutrophil-mediated response that may include bacterial nucleic acid activation of neurogenic pathways, rather than LPS-TLR-4/2 processes. Supported by Procter and Gamble and RR-00046.