Microbiome characterization of urbanized lakes impacted by legacy arsenic contamination

Bacteria and other microbes are important contributors to arsenic biotransformation processes, which can alter the bioavailability and toxicity of arsenic within a contaminated environment. Multispecies biofilms, known as periphyton, have been identified as a significant site of arsenic bioaccumulation within shallow freshwater lakes impacted by legacy arsenic contamination. We hypothesized that prolonged arsenic exposure results in the formation of distinct prokaryotic communities within the periphyton and other environmental compartments in arsenic-contaminated lakes compared to uncontaminated lakes. We also predicted that the periphyton prokaryotic communities would be distinct from, but partially overlapping with, those found in the surrounding water column and nearby littoral sediment. To test these hypotheses, we determined the taxonomic composition and modeled the assembly processes that yielded the bacterial communities found within three environmental compartments (periphyton, littoral sediment, and water column) of three lakes that had been differentially impacted by legacy arsenic contamination. We identified unique microbiomes within these environmental compartments and observed a clear shift in microbial community composition within high arsenic-contaminated periphyton. Hyperaccumulation of arsenic (~400 ppm) in the periphyton correlated with non-random (deterministic) selection for prokaryotic taxa that are more related than expected by chance (homogenizing selection). We also identified key prokaryotic genera within the arsenic-contaminated periphyton that suggest prolonged arsenic contamination may shift iron and methane biogeochemical cycles, which may regulate arsenic accumulation and mobilization. Our results imply that legacy arsenic contamination, by altering bacterial community composition and metabolic potential at the base of the food web, may influence biogeochemical and nutrient cycles at a larger scale within a freshwater lake ecosystem.

Methods
Lake sampling:

Samples from three environmental compartments (littoral sediment, lake water, periphyton) were collected in September 2023 from three south King County, WA lakes that represent a gradient of arsenic contamination: Lake Killarney (high), Steel Lake (mid), Trout Lake (low; reference). All lakes were of similar depths and shared other general geomorphological features (Figure 1; Table 1). Samples were collected from three independent sites per lake. Surface water was collected from a paddleboard in sterile 1L glass bottles. Littoral sediment was collected in sterile 50 ml tubes by hand at a water depth of ~0.5m. Periphyton was cultivated on acrylic plates, which were deployed within a plastic frame that was submerged at a depth of ~1m in each lake. The periphyton plates soaked in the lake for ~60 days prior to collection. During sample collection (n = 10 per sample type, per lake), sediment and periphyton were placed in sterile transport vessels and immediately processed in the lab. Subsamples of periphyton were 1) scraped off the acrylic plates using a sterile razor blade, flash frozen with liquid nitrogen and stored at -80°C to preserve for DNA extraction or 2) dried at 60°C for 72 hours to process for inductively coupled plasma mass spectrometry (ICP-MS) analysis. Water samples were 1) filtered using  0.2 𝜇m sterile filters, which captured microbes and were flash frozen and stored at –80°C to preserve for DNA extraction or 2) filtered using 0.45μm pore size surfactant-free cellulose acetate vacuum filter and acidified to 2% HNO3 to preserve for ICP-MS.

Total arsenic and iron concentration:

Samples were prepared for ICP-MS according to previously described methods (Hull et al., 2021). Briefly, the dried sediment and periphyton were digested in a microwave with trace metal grade HNO3 in pressurized digestion vessels (CEM MARS 5) following the total digestion protocol (modified EPA method 3015a). After digestion, sample solutions were diluted to 2% (v/v) HNO3 and filtered using 0.45μm pore size surfactant-free cellulose acetate syringe filters. Water samples did not undergo microwave assisted digestion. Digestion procedure efficacy was evaluated using certified reference material DOLT-5 (dogfish liver), which yielded a recovery of 97.18% (n = 5). Total arsenic concentrations in water, digested sediment, and periphyton samples were determined by ICP-MS (Agilent 7900). The lower limit of detection (LOD) was 0.25 μg/L for arsenic. Calibration was performed using a certified multi-element metals standard (Agilent Multi-Element Calibration Standard 2A) with 7 different dilution points. Instrument performance was assessed by checking two standards after every batch of 10 unknown samples.

16S rRNA gene amplicon sequencing for microbiome analyses:

DNA isolation from field samples:

DNA was extracted using the PowerSoil Pro kit (Qiagen; Hilden, Germany). Negative controls for the extraction process were run in parallel at all stages of extraction and sequencing. DNA was checked for purity and concentration on a Nanodrop (Thermo Scientific). DNA samples were normalized to the lowest concentration per sample type (sediment, periphyton, and water) and sent to MR DNA (Shallowater, TX) for paired-end Illumina sequencing of the 16S rRNA gene. The V3-V4 region of the 16S rRNA gene was amplified using 341F and 785R primer pairs to target high diversity and richness of samples (Thijs et al., 2017). Mock community reference samples (ATCC, n=3) were included as positive controls.