DNA Adductomics for the Biological Effect Assessment of Contaminant Exposure in Marine Sediments

Abstract

Exposure to chemical pollution can induce genetic and epigenetic alterations, developmental changes, and reproductive disorders, leading to population declines in polluted environments. These effects are triggered by chemical modifications of DNA nucleobases (DNA adducts) and epigenetic dysregulation. However, linking DNA adducts to the pollution load in situ remains challenging, and the lack of evidence-based DNA adductome response to pollution hampers the development and application of DNA adducts as biomarkers for environmental health assessment. Here, we provide the first evidence for pollution effects on the DNA modifications in wild populations of Baltic sentinel species, the amphipod Monoporeia affinis. A workflow based on high-resolution mass spectrometry to screen and characterize genomic DNA modifications was developed, and its applicability was demonstrated by profiling DNA modifications in the amphipods collected in areas with varying pollution loads. Then, the correlations between adducts and the contaminants level (polycyclic aromatic hydrocarbons (PAHs), trace metals, and pollution indices) in the sediments at the collection sites were evaluated. A total of 119 putative adducts were detected, and some (5-me-dC, N⁶-me-dA, 8-oxo-dG, and dI) were structurally characterized. The DNA adductome profiles, including epigenetic modifications, differed between the animals collected in areas with high and low contaminant levels. Furthermore, the correlations between the adducts and PAHs were similar across the congeners, indicating possible additive effects. Also, high-mass adducts had significantly more positive correlations with PAHs than low-mass adducts. By contrast, correlations between the DNA adducts and trace metals were stronger and more variable than for PAHs, indicating metal-specific effects. These associations between DNA adducts and environmental contaminants provide a new venue for characterizing genome-wide exposure effects in wild populations and apply DNA modifications in the effect-based assessment of chemical pollution.

Keywords: DNA adducts; amphipods as sentinel species; biological effect monitoring; biomarkers; environmental contaminants; high-resolution mass spectrometry.

Figure 1

Sampling stations in the Bothnian Sea (top inset) and the Northern Baltic Proper (bottom inset) along the Swedish Baltic Sea coast. The stations with high contaminant levels (PAHs and metals) in the sediment are shown as stars, and those with contaminant concentrations below critical levels are shown as circles. The sites were classified as Contaminated and Reference, respectively; see Note S1 for details on the stations, contaminant concentrations, and the evaluation of the pollution status.

Figure 2

Workflow of sample preparation (1), HRMS analysis and data processing (2), and statistical evaluation (3) performing permutational multivariate analysis of variance (PERMANOVA), similarity percentage analysis (SIMPER), orthogonal partial least squares discriminant analysis (OPLS-DA), and receiver operating characteristic (ROC) analysis.

Figure 3

Example of a DNA adductome map representing modified nucleobases detected in genomic DNA using the untargeted nLossFinder approach followed by TraceFinder processing. The X-axis shows the HPLC retention times of detected 2′-deoxyribonucleoside adducts, and the Y-axis shows the m/z values of their precursor ions (200–600 range). Circle size represents the relative abundance of the adducts (i.e., the peak area of the adduct normalized to dG). The DNA originated from an individual inhabiting highly contaminated sediments (station Bråviken 6). Structurally identified adducts confirmed by standards are marked in color. 2′-Deoxyribonucleoside adducts with a low mass range have an elution time between 1 and 15 min, with a major cluster eluting in 6–10 min intervals, where the mobile phase mainly consists of water, implying that low-mass adducts are more hydrophilic. Adducts detected at a retention time of 15–25 min are more hydrophobic as they elute toward the end of the gradient program, where the mobile phase mainly consists of methanol.

Figure 4

Heatmap of Pearson correlations between the relative abundance of the adducts and contaminant concentrations for the low-mass (A) and high-mass adducts (B). The color shade indicates the strength (weak/strong), and the color indicates the direction of the correlation (positive/negative).

Figure 5

Score plot of OPLS-DA. The data points represent the samples collected in the stations classified as contaminated (red) or reference (green).

Figure 6

Volcano plot showing the adducts that were significantly upregulated (red) and downregulated (blue) in the contaminated sites compared to the reference sites. The data include the 37 putative adducts used for the statistical analysis following the SIMPER screening. The X-axis represents the log₂ fold change (FC) with a threshold of 2; the direction of the comparison is Contaminated/Reference. The Y-axis is −log 10(p) with a p-value threshold of 0.1.