Dissolved organic matter thiol concentrations determine methylmercury bioavailability across the terrestrial-marine aquatic continuum

The most critical step for methylmercury (MeHg) bioaccumulation in aquatic food webs is phytoplankton uptake of dissolved MeHg. Dissolved organic matter (DOM) has been known to influence MeHg uptake, but the mechanisms have remained unclear. Here we show that the concentration of DOM-associated thiol functional groups (DOM-RSH) varies substantially across contrasting aquatic systems and dictates MeHg speciation and bioavailability to phytoplankton. Across our 20 study sites, DOM-RSH concentrations decrease 40-fold from terrestrial to marine environments whereas dissolved organic carbon (DOC), the typical proxy for MeHg binding sites in DOM, only has a 5-fold decrease. MeHg accumulation into phytoplankton is shown to be directly linked to the concentration of specific MeHg binding sites (DOM-RSH), rather than DOC. Therefore, MeHg bioavailability increases systematically across the terrestrial-marine aquatic continuum as the DOM-RSH concentration decreases. Our results strongly suggest that measuring DOM-RSH concentrations will improve empirical models in phytoplankton uptake studies and will form a refined basis for modeling MeHg incorporation in aquatic food webs under various environmental conditions.


Site
Cl  S5.Relative concentrations (% of total S) of eight pseudo-components of S species deconvoluted from S XANES spectra by Gaussian distributions.Energies for each component are given in Figure S9.The total organic reduced sulfur pool is represented by the sum of RSSR, RSR and RSH functionalities, where R denotes an organic C backbone structure.

Site Name BC BI CI CM CP CTR GOM HC LIS M MP NB NBD OR OS PAW PI SI SR ST
Where f RS is the fraction of the MeHg that is MeHg(DOM-RS), U RS is the uptake rate constant of MeHg(DOM-RS), U Cl is the uptake rate constant of MeHgCl, and U is the overall MeHg uptake rate constant (amol µm -3 hr -1 nM -1 ).[MeHg(cell) n+1 ] t is the MeHg concentration accumulated in cells at time t (amol µm -3 ), [MeHg(aq)] t=0 is the total dissolved MeHg concentration at time 0 (nM), t is the time (h) and [MeHg(aq)]' t=0 is the total dissolved MeHg concentration at time 0 (amol µm -3 ).S4).Source data are provided as a Source Data file.(one peak centered at -0.5 eV), RSSR (two peaks at 0.05 and 1.65 eV), RSH+RSR (one peak at 1.1 eV), sulfoxide (one peak at 3.3 eV), sulfone (one peak at 6.9 eV), sulfonate (one peak at 8.4 eV) and estersulfate (one peak at 10.0 eV).The thin solid black line is background, the thick black line full spectrum and the red line is the full model fit.The method for deconvolution is described in Yekta et al. 11 .Source data are provided as a Source Data file.
the site with the determined DOM-RSH/DOC content for the corresponding extracted DOM sample (Table 1).The Cl -concentration (mM) at each site was calculated from the determined salinity (Table 1) as: [Cl -(mM)] = [salinity(psu)]/(0.00180665× 35.45) 17 .The pH at each site was estimated by the measured salinity (Table 1) as: pH = 0.02195 × [salinity(psu)] + 7.49.This relationship was made assuming conservative mixing between the lowest salinity sites with a pH of 7.5 and the highest salinity sites with a pH of 8.2.Therefore, the calculated pH varied from 7.5 to 8.2 for the lowest to highest salinity sites.

Model 3 -Chemical speciation of MeHg in cellular uptake experiment
Model 3a was used to describe cellular uptake of MeHg via passive diffusion of MeHgCl without interaction with cell surface ligands.The model included reactions 2, 5, 7, 9 and 12 (i.e.same as model 2).The total MeHg concentration was 10 pM, the DOM-RSH was variable in the ranges 17-170 and 2.2-22 nM for the two sites SI and OR, respectively.The Cl -concentration was 500 mM and pH was 8.2.
Model 3b was used to describe cellular uptake of via MeHg binding to cell surface thiol groups (cell-RSH) followed by cellular internalization.In addition to the reactions of Model 3a, also reactions 10 and 11 were included.In addition to the concentration input values for Model 3a, a total cell-surface thiol content of 100 nM was estimated using a measured density of ~1 fmol/cell for T. Pseudonana 18 a total cell number of 2.4×10 7 cells and an experimental assay volume of 200 ml.

Figure S2 .Figure S3 .
Figure S2.MeHg ligand exchange kinetics.The 204 Hg/ 200 Hg isotope ratio of the MeHg(Nacpen) complex (R MeHg(Nacpen) ) divided by the 204 Hg/ 200 Hg isotope ratio of total MeHg (R total MeHg ) over time measured in the kinetic experiment for marsh (SI and BI), river (PAW), estuarine (NB) and shelf (SB) sites.The individual lines represent the different concentrations of competing ligand (20000, or 200 nM Nacpen) tested in the experiment.At equilibrium R MeHg(Nacpen) / R total MeHg = 1.0 and a deviation from 1.0 at steady-state conditions indicates the presence of "non-exchangable" MeHg pools.The Inserted data table shows calculated rate constants for the forward (k f ) and backward (k r ) ligand exchange reactions using equation (6), and the turnover calculated as: turnover = (k f + k r ) -1 .Source data are provided as a Source Data file.

Figure S4 .Figure S5 . 6
Figure S4.MeHg volume concentration factors.Comparison between the volume concentrations factors measured by Luengen et al. 10 and those determined in this study.The model fit includes both studies.Source data are provided as a Source Data file.

Figure S6 .
Figure S6.The in situ speciation of MeHg at the study sites.Percent fraction of MeHg(DOM-RS) complexes of total MeHg in water versus total DOM-RSH concentration for all study sites (TableS4).Source data are provided as a Source Data file.

Figure S7 .
Figure S7.Overview of the raw fluorescence excitation-emission matrices (EEMs) of samples in this study.(a): average fluorescence across all samples.(b): standard deviation of fluorescence across all samples.The fluorescence in each EEM was normalized to its sum prior to the calculation of the standard deviation.Superimposed crosses and lines in the left-upper portion of the figures show the position of the three fluorescence indices used in this study (HIX: humification index (next to y-axis), FrI: freshness index (center), FluI: fluorescence index (right).Source data are provided as a Source Data file.

Figure S8 .
Figure S8.Overview of the four-component Parallel Factor Analysis (PARAFAC) model.(a)-(d): Component loadings as excitation-emission matrix.(e)-(h): Component excitation (dashed lines) and emission (solid lines) loadings as line plots.The results of the splithalf-validation are superimposed in different colors.Source data are provided as a Source Data file.

Figure S9 .
Figure S9.Sulfur K-edge XANES spectra of the extracted DOM.Spectra are deconvoluted into eight pseudo-components (dashed black lines) by use of Gaussian distributions.From left to right: FeS 2 +S 0 (one peak centered at -0.5 eV), RSSR (two peaks at 0.05 and 1.65 eV), RSH+RSR (one peak at 1.1 eV), sulfoxide (one peak at 3.3 eV), sulfone (one peak at 6.9 eV), sulfonate (one peak at 8.4 eV) and estersulfate (one peak at 10.0 eV).The thin solid black line is background, the thick black line full spectrum and the red line is the full model fit.The method for deconvolution is described in Yekta et al.11 .Source data are provided as a Source Data file.

Table S1 .
Concentrations of dissolved organic sulfur, total Hg and MeHg at the different sampling sites.

Table S2 .
Comparison of MeHg stability constants with specific low molecular mass (LMM) thiol containing compounds, non-thiol containing compounds, and dissolved organic matter (DOM).

Table S3 .
Chemical reactions and stability constants used in the WinSGW speciation software.

Table S4 .
Calculated chemical speciation of MeHg (expressed as % of total MeHg concentration) for each species at each study site, and the Cl -concentration and pH used in the modeling (SI Text Model 1, Table