Metabolomics Reveals Elevated Macromolecular Degradation in Periodontal Disease

Introduction

Periodontal disease is a chronic bacterial infection defined by persistent inflammation, connective tissue breakdown, and bone destruction. It is characterized by pocket formation and gingival recession. Although improvement in dental care has resulted in a marked increase of periodontally healthy individuals in the industrial world, global problems still persist. Worldwide, severe periodontitis is found in 5 to 20% of the adult population (Petersen et al., 2005). Periodontal disease is associated with several systemic diseases, including cardiovascular disease, diabetes, and infant prematurity (Behle and Papapanou, 2006; Demmer and Desvarieux, 2006; Tonetti et al., 2007).

In the early phase of the disease (gingivitis), inflammation is confined to the gingivae, but in periodontitis it extends to deeper tissues, leading to swelling and bleeding. In the late phase of periodontal disease, the supporting collagen has deteriorated and the alveolar bone is resorbing. Moreover, the gingival epithelium has migrated along the tooth surface to form a pocket that is ideal for the proliferation of micro-organisms (Cochran, 2008; Kornman, 2008; Van Dyke, 2008). A great deal of effort has been devoted to assessing the composition of oral fluids in the diseased state. Analysis of the gingival crevicular fluid (GCF) from this region has identified contributing inflammatory mediators and tissue-destructive molecules, including metalloproteinases (Embery and Waddington, 1994; Akalin et al., 2007; Pradeep et al., 2007). Moreover, metabolic signatures associated with host-bacterial interactions have been identified in the GCF (Barnes et al., 2009). The surrounding saliva plays an important role in maintaining the integrity of teeth. Saliva contains a complex mixture of secretions from salivary glands, GCF, mucosal fluid, epithelial cells, and microbial products. Human saliva lubricates the oral tissues and facilitates oral functions such as speaking, eating, and swallowing. In addition, it protects teeth and oral mucosal surfaces from bacterial infections by remineralization, removing food debris, and neutralizing pH (Lenander-Lumikari and Loimaranta, 2000; Stookey, 2008).

We recently performed untargeted global metabolomic profiling of GCF samples from healthy and periodontitis sites that proved a rich source of metabolites (Barnes et al., 2009, 2010). Many metabolites associated with inflammation, oxidative stress, tissue degradation, and bacterial metabolism were found to be significantly induced by the disease. Here we have expanded the analysis to investigate the global metabolomic impact of periodontitis on the primary fluid in the oral cavity, saliva. The metabolic integration of these 2 oral fluids provides a comprehensive biochemical basis for an understanding of the pathogenesis of periodontitis.

Materials, Participants, & Methods

Results

To understand the oral environment and the biochemical basis underlying periodontitis, we took an untargeted metabolomic profiling approach to assess the chemical milieu of saliva. Of the 390 metabolites detected, 250 matched known structures in our chemical reference library. We used Welch’s paired t test to analyze the differences between the healthy and periodontal disease population. Seventy-two metabolites (40 named and 32 without identity) in the saliva showed significantly altered levels (p < 0.05) and were increased in the periodontal population. The metabolites matched with known chemical structures and their statistical comparisons among the healthy and periodontal sites by t test are listed in Appendix B. A summary of the biochemical pathways and compound classes altered by the periodontal diseases is presented in Table 1.

OPEN IN VIEWER

Table 1.

Summary of the Biochemical Super Pathways, Numbers of Compounds Altered, and Direction in the Individuals with Periodontal Disease Compared with Healthy Individuals

Super Pathway	# Changed Biochemicals	Altered in Periodontitis
Peptides	5	up
Amino acids	5	up
Carbohydrates	3	up
Lipids	22	up
Nucleotides	4	up
Co-factor	1	up
Unnamed	32	up

The detailed list of the metabolites and the statistical values by Welch’s t tests are given in Appendix A. All significantly altered metabolites and intermediates associated with each of the super pathways listed in the Table are included. Up: the pathway or metabolites showed higher levels in the diseased samples.

Activation of Macromolecular Degradation

A general trend indicating increased biochemical activation was observed in the saliva of the periodontal population. In particular, biochemicals associated with the degradation of macromolecules, including protein, lipids, and polysaccharides, were altered (Fig.). Close examination of these metabolites showed that both mono- and oligosaccharides were elevated in the periodontal population, suggesting that α-amylase activity, the main glycosidase in saliva, was increased. Further support for the activation of digestive enzymes in the periodontal population was observed in the elevated levels of 15 dipeptides. This is consistent with previous reports that several peptidases, including elastases, amino peptidases, and metalloproteases, have been shown to be elevated in periodontitis (Miller et al., 2006; Rai et al., 2008). Interestingly, only 1 (cysteine) of the 20 amino acids involved in protein biosynthesis was elevated, indicating that dipeptides may be utilized by cells, in particular microbes, rather than further degraded to the free amino acid form. One of the most striking results was the increased levels of lysolipids, monoacylglycerol, and fatty acids, products of glycerophospholipid and triacylglycerol degradation. The increase could possibly be attributed to up-regulation of lipase activity in the periodontal population. Although a lingual lipase has been identified in saliva, the low pH optimum has been thought to render it fairly inactive in the mouth, until it reaches a more favorable pH for catalytic activity in the stomach (Hamosh and Burns, 1977). It may be possible that the altered lipase activity was a result of increased expression, pH, and activated inflammatory response. Several of the increased fatty acids in the periodontal population, including dihomo-linolenate, arachidonate, docosapentaenoate, and docosahexaenoate, are precursors to inflammation molecules such as eicosanoids. This biochemical signature is indicative of an increased inflammatory environment consistent with chronic bacterial infection. Further support for increased macromolecular degradation was the finding that several metabolites involved in nucleotide metabolism were elevated in the periodontal population, consistent with the degradation of DNA or RNA associated with clearance of a more active bacterial population (data not shown).

OPEN IN VIEWER

Figure.

Illustration of the 3 predominant macromolecular degradation pathways and their associated enzyme classes in saliva. The letter R in the depicted phospholipid refers to headgroups such as choline, ethanolamine and inositol. Listed are altered biochemicals from each biochemical class and the associated p-value resulting from the comparison of the periodontal (P) with healthy (H) populations by Welch’s t test. Statistically significant changes are in bold (p > 0.05). Trending values are in normal font (0.05 ≤ p < 0.10). 163 x 135 mm (150 x 150 DPI).

Inflammatory Response and Host-Bacteria Interaction

Several biochemical signatures supported the notion that a more active bacterial environment was observed in the periodontal population. Aromatic amino acid metabolites, including p-cresol sulfate and phenol sulfate, associated with bacterial biochemistry were increased in the periodontal population (Table 2). Moreover, carnitine, which can be used as a sole source of carbon and nitrogen by bacteria, was elevated. 3-dehydrocarnitine, a bacterial degradation product of carnitine, was also elevated, further supporting the utilization of carnitine in oral bacterial metabolism.

OPEN IN VIEWER

Table 2.

Summary of Altered Amino Acid and Carnitine Metabolites When Contrasting the Periodontal (P) and the Healthy (H) Populations

Metabolite	Fold-change P vs. H	p-value
Amino acid metabolites
Phenylalanine	1.15	0.0994
Phenylacetate	1.29	0.0704
p-cresol sulfate	1.84	0.0165
Tyrosine	1.14	0.0571
Phenol sulfate	1.42	0.0454
Indoleacetate	1.66	0.0607
Carnitine metabolites
Carnitine	1.36	0.0166
3-dehydrocarnitine	1.36	0.0249
Acetylcarnitine	1.32	0.073

Many of these metabolites are the product of host-microbial metabolism. Statistically significant changes are in bold (p < 0.05). Trending values are in normal font (0.05 ≤ p < 0.10).

Discussion

In this study, we successfully applied biochemical profiling to assess the differential global metabolism of saliva in a periodontal disease population compared with healthy individuals. Many metabolic changes were found to be associated with periodontal disease. All changes were increased, pointing toward increased global metabolic activity. The increased levels of fatty acids, dipeptides, and monosaccharides were indicative of increased lipase, protease, and glycosidase activities in the periodontal disease state. A direct result of the macromolecular degradation is the increased availability of readily accessible metabolites for energy production by the oral microflora. Thus, a more fertile environment for bacterial expansion was present in the periodontal population. A mechanistic understanding of how bacterial infection has affected the host to alter salivary metabolism to the apparent benefit of the bacteria is limited. Studies have shown that many host and bacterial proteases are elevated in a periodontal population consistent with the elevated levels of dipeptides (Miller et al., 2006; Rai et al., 2008). Large amounts of the secretory cysteine proteases Arg- and Lys-gingipain are produced in Porphyromonas gingivalis to facilitate the extraction of peptides from surrounding tissues and contribute to the maturation of cell-surface proteins such as fimbrilin (Fitzpatrick et al., 2009). It is interesting to note that P. gingivalis and many other bacteria are unable to utilize saccharides as carbon and energy sources and instead rely on oligopeptides for biomass generation. The increased levels of dipeptides would provide a richer energy pool for the expansion of these organisms (Takahashi and Sato, 2001, 2002). The absence of elevated levels of free amino acids is consistent with the uptake of dipeptides for energy use by bacteria.

Numerous proteomic studies have shown that amylase is increased in the periodontal population consistent with elevated levels of mono- and oligosaccharides observed in this study (Wu et al., 2009; Gonçalves Lda et al., 2010; Sanchez et al., 2011). Salivary amylase is a multifunctional enzyme that hydrolyzes starch to glucose and readily adheres to teeth. Many streptococci contain amylase-binding proteins that facilitate colonization of oral surfaces (Scannapieco et al., 1989; Scannapieco, 1994). The bound amylase retains enzymatic activity and feeds the bound bacteria energy-rich monosaccharides. In this context, the elevated levels of glucose and oligosaccharides suggested that this may be a more active process, with amylase activity exceeding the bacterial demand for saccharides in the periodontal population.

The strongly altered lipid profile was consistent with elevated lipase activity. Studies on salivary lipids have shown that neutral lipids dominate, with a small population of glycerophospholipids (Larsson et al., 1996). Increased levels of lysolipids, products of glycerophospholipid degradation, and the elevation of complex fatty acids suggest that a significant part of increased lipolysis is the result of endogenous breakdown of predominantly glycerophospholipids originating from the oral cavity. The action of polymorphonuclear leukocytes on biofilm and associated bacteria results in the release of enzymes into the surrounding tissue, as well as the synthesis and secretion of pro-inflammatory lipids such as arachidonic acid metabolites. This is consistent with the increased levels of inflammatory precursors observed in the periodontal population.

A great deal of effort has been devoted to the identification of biomarkers that are associated with periodontal disease (Embery and Waddington, 1994; Miller et al., 2006; Christodoulides et al., 2007; Pussinen et al., 2007). Prediction of risk and accurate diagnosis of current disease activity may facilitate effective prevention and treatment. GCF and saliva have been extensively examined in attempts to assess the oral disease status. Several different approaches have confirmed that identifying a single marker is unlikely; rather, a combination of biomarkers would be an effective clinical test. Our recent metabolomic contrast of GCF from healthy individuals and those with periodontal disease confirmed this approach (Barnes et al., 2009, 2010). In considering that saliva is a readily collected material, in contrast to GCF, and metabolically rich, the current study provides an extensive pool of potential biomarkers for evaluating periodontal disease in a point-of-care environment. In particular, the dipeptides leucylisoleucine, phenylphenol, and serylisoleucine as well as the fatty acids arachidonate, arachidate, and dihomo-linolate are attractive candidate markers. Efforts to validate and refine the biomarker candidate list are currently in process.