The dental calculus metabolome in modern and historic samples

Abstract

Introduction: Dental calculus is a mineralized microbial dental plaque biofilm that forms throughout life by precipitation of salivary calcium salts. Successive cycles of dental plaque growth and calcification make it an unusually well-preserved, long-term record of host-microbial interaction in the archaeological record. Recent studies have confirmed the survival of authentic ancient DNA and proteins within historic and prehistoric dental calculus, making it a promising substrate for investigating oral microbiome evolution via direct measurement and comparison of modern and ancient specimens.

Objective: We present the first comprehensive characterization of the human dental calculus metabolome using a multi-platform approach.

Methods: Ultra performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS) quantified 285 metabolites in modern and historic (200 years old) dental calculus, including metabolites of drug and dietary origin. A subset of historic samples was additionally analyzed by high-resolution gas chromatography-MS (GC-MS) and UPLC-MS/MS for further characterization of metabolites and lipids. Metabolite profiles of modern and historic calculus were compared to identify patterns of persistence and loss.

Results: Dipeptides, free amino acids, free nucleotides, and carbohydrates substantially decrease in abundance and ubiquity in archaeological samples, with some exceptions. Lipids generally persist, and saturated and mono-unsaturated medium and long chain fatty acids appear to be well-preserved, while metabolic derivatives related to oxidation and chemical degradation are found at higher levels in archaeological dental calculus than fresh samples.

Conclusions: The results of this study indicate that certain metabolite classes have higher potential for recovery over long time scales and may serve as appropriate targets for oral microbiome evolutionary studies.

Keywords: Archaeology; Dental plaque; GC–MS; Metabolomics; Oral microbiome; UPLC–MS/MS.

Fig. 1

Heat map summary of metabolites observed in modern and historic dental. Metabolites were quantified by area under the curve and normalized to mass of sample extracted. a UPLC–MS/MS-detected metabolites. Samples were hierarchically clustered, and log2 transformed values are presented above, grouped by super-pathway. b Metabolites detected by GC–MS and LC–MS in historic calculus. Non-filled cells containing a dot indicate the compound was not detected

Fig. 2

Compound preservation is correlated with aqueous solubility. a Percent of high ubiquity metabolites in modern calculus that were also recovered in at least one historic calculus sample. Peptides exhibit the poorest representation in historic dental calculus, with only 9% of the peptides observed to be present in all five modern calculus samples also detected in any historic sample. By contrast, Lipids and Energy (TCA cycle) super-pathways exhibit high representation in historic calculus, with > 96% of compounds found in all five modern samples also recovered from historic dental calculus. Xenobiotics, which largely comprise dietary and pharmaceutical compounds, are not shown. b The log2 fold-change (modern/historic) of metabolite abundance versus the 1-octanol versus water partition coefficient (logP), estimated with the ALOGPS tool. In the cases where metabolites were only detected in modern calculus, the log2 values (not fold-change) were plotted relative to logP. The fitted linear model showed a significant effect (p < 0.0001) of logP on metabolite fold-change with and adjusted R² of 0.22. The outlier from this trend was mead acid (20:3n9)

Fig. 3

Differences exist in mean proportions of metabolites detected in at least one historic and one modern dental calculus sample. a Principal components analysis distinctly separates modern and historic calculus samples. b Metabolites with significant differences (q ≤ 0.05, effect size of ≥ 1.0) in mean proportions between historic and modern calculus

Fig. 4

Differences exist in proportions of super-pathways, sub-pathways, and metabolites represented in at least one historic and one modern dental calculus sample. Significantly different (q ≤ 0.05, effect size of ≥ 1.0) proportions of a Super-pathways, b Sub-pathways, and c metabolites. Individual metabolites between historic and modern samples

Fig. 5

Partial least squares discriminant analysis of metabolites detected at least one modern and one historic calculus sample. a Calculus samples cluster based on time period rather than biological category (sex, age, caries status and periodontal disease status) when including all metabolites detected in at least one modern and one historic sample. b Historic calculus samples cluster based on biological category (sex, age, caries status and periodontal disease status) when including all metabolites detected in at least one historic sample. Ellipses indicate 95% confidence intervals