Thiol methylation potential in anoxic, low-pH wetland sediments and its relationship with dimethylsulfide production and organic carbon cycling

Abstract

Dimethylsulfide (CH₃SCH₃) is formed in anoxic freshwater sediments by biological methylation of methanethiol (CH₃SH). We measured thiol methylation potential in low-pH, Sphagnum peat sediments from Alaska and Alabama by adding ethanethiol (CH₃CH₂SH) to peat slurries and quantifying the rate of ethylmethylsulfide (CH₃CH₂SCH₃) formation. Thiol methylation potential ranged from 12 to 154 nM h⁻¹ and was significantly related to dimethylsulfide accumulation rates (P=0.0007; r²=0.48). Addition of methanol or syringic acid stimulated thiol methylation potential and dimethylsulfide accumulation rate, suggesting that these compounds could be methyl donors. Addition of acetate or its metabolic precursors (glucose or Sphagnum plant material) inhibited thiol methylation potential, but not carbon dioxide or methane production. Inhibition of methanogenesis with either 2-bromoethanesulfonic acid or KNO₃ consistently inhibited thiol methylation potential and dimethylsulfide accumulation. These results suggest that methanogens play a role in thiol methylation and therefore dimethylsulfide formation.

Methanogenesis, Acetate, Sphagnum, Methanethiol, Methylation, Sediment

1 Introduction

Dimethylsulfide (DMS) is a climatically active sulfur gas that dominates flux of biogenic sulfur to the atmosphere [1]. In natural aquatic systems DMS is produced by several distinct mechanisms [2]. One of these is the degradation of certain organosulfur compounds, most notably, the algal-derived compound dimethylsulfoniopropionate which leads to formation of DMS in surface ocean waters and Spartina alterniflora marshes [3]. Another, less well understood mechanism of DMS formation is through methylation of thiols in anaerobic sediments [4] and in a variety of oxic freshwater habitats [5]. DMS formation by methylation occurs via transfer of a methyl group to the acceptor, methanethiol (MeSH), which can originate either from biomethylation of hydrogen sulfide (H₂S) or by degradation of organosulfur precursors (e.g. methionine) [6,,7]. Bacteria capable of transferring methyl groups from methoxylated aromatic compounds to H₂S and MeSH, and forming DMS, have been isolated from anoxic freshwater sediments [8,9].

Kiene and Hines [10] found that potential thiol methylation rates in Sphagnum peats greatly exceeded the observed rates of DMS production, suggesting that thiol methylation could be a significant source of sedimentary pools of MeSH and DMS in anaerobic freshwater sediments. The same study also found that addition of the potential methyl donor, syringate (4-hydroxy,3,5-dimethoxybenzoate), or the methyl acceptor, MeSH, resulted in greater DMS production. In a survey of anaerobic freshwater sediments with circumneutral pH, Lomans et al. [11] found that in situ concentrations of H₂S correlated strongly with in situ concentrations of both MeSH (r=0.91) and DMS (r=0.81), suggesting a link between availability of H₂S and production of MeSH and DMS.

Production and consumption of MeSH and DMS were closely coupled in circumneutral-pH freshwater sediments leading to low standing pools of these compounds [11]. Sulfate-reducing bacteria appear to be the primary DMS consumers in freshwater sediments with high concentrations of sulfate, whereas methanogens appear to dominate DMS consumption in low-sulfate sediments [12,13]. Kiene and Hines [10], however, did not observe significant DMS consumption in slurried sediments from a low-pH, high-latitude Sphagnum bog, yet DMS was produced at rates of about 40 nM day⁻¹. The rapid accumulation of DMS in these sediments is somewhat puzzling since they typically have low concentrations of sulfate and other nutrient anions. Based on work with neutral-pH sediments, one would expect DMS consumption in these low-sulfate sediments to be dominated by methanogens [11]. However, suppression of methanogenesis with the inhibitor 2-bromoethanesulfonic acid (BES) led to slower rates of DMS accumulation in low-pH peat sediments [10,14], suggesting that methanogens were not responsible for DMS consumption but may instead be involved with DMS formation. In fact very little is known about the mechanisms, or the organisms responsible for DMS production in anoxic freshwater sediments, especially in low-pH peat systems that contribute significantly to terrestrial DMS fluxes to the atmosphere [15].

The present study investigated several factors governing DMS production in low-pH Sphagnum peat sediments from both high and low latitudes. Slurried sediments were incubated and tested for the capacity to methylate thiols and ultimately to produce DMS. Chemical amendments and temperature manipulations were used to provide insight into the mechanism controlling thiol methylation potential (TMP) and DMS production in these sediments.

2 Materials and methods

2.1 Study sites

The Sphagnum peats used for this study were collected from one low-latitude site, located in southern Alabama, USA and two high-latitude sites located in Alaska, USA. The low-latitude site, Boggy Branch Creek, is a small tributary to the Tensaw River located in an undeveloped area of the Alabama Forever Wild Lands near Spanish Fort, AL (30°45′N, 88°15′W). While this area is not a true raised peat bog, it shares many of the same sediment characteristics including low pH (3.8–4.4), low concentrations of nutrient anions, high sediment organic content (67–70% based on ignition at 500°C for >3 h), and an understory dominated by Sphagnum moss amid a mixed oak–birch forest.

The high-latitude sites were Turnagain Bog, located in western Anchorage, AK near the Cook Inlet (60°10′N, 149°11′W) and Talkeetna Bog located near Cantwell, AK (61°N, 148°W). Turnagain Bog is a true high-latitude bog with an understory of Sphagnum moss [16]. The sediments are 48–60% organic material (based on ignition of sediment at 500°C for >3 h) with pH 4.5–5.3. Sphagnum mosses were typically the only species found on Turnagain Bog hummocks, whereas hollows had a mixture of Sphagnum sp. moss and other shrubs such as dwarf birch (Betula nana), Labrador tea (Ledum spp.), and horsetail (Equisetum spp.). Talkeetna Bog is a waterlogged expanse of Sphagnum moss with some growth of ericaceous shrubs. Sediments had a low pH (4.1) and were primarily organic material.

2.2 Sediment collection and preparation of slurries

Sediments were collected by hand using a knife to cut a small core out of the substrate. Typically sediments of 5–10-cm depth were used in slurry incubations, ensuring that anoxic zones were sampled. Slurries were prepared by homogenizing approximately 1 l of water and 1 l of sediment collected from each site studied. Collected sediments were placed in a blender and homogenized at low speeds for <30 s. The homogenized sediment slurries were then sieved (1-mm pore size) and poured through a funnel into a series of 14-ml and 72-ml glass serum vials. The smaller vials each received approximately 10 ml of slurry and were subsequently used for assays of total CO₂ production and TMP (see below). The larger vials each received approximately 60 ml of slurry and were used for measurements of sulfur gases and anions (see below). Each vial of slurry was capped with a gray, Teflon-faced butyl stopper and crimp-sealed. To ensure anoxic conditions, the headspace of each sealed vial was swept with N₂ gas for 3 min. Each batch of slurry yielded approximately 12–24 replicate vials of each size that were used for the experiments described below. We found no differences in patterns of sulfur gas production or acetate accumulation in slurries prepared as described above with those prepared in a glove bag under nitrogen. While slurrying of sediments could be disruptive to certain microbial processes, earlier work with Sphagnum peats found no appreciable differences in TMP, sulfur gas production rates, or acetate accumulation rates between slurried and non-slurried peats [17,18]. This suggests that processes occurring in slurries are reasonably representative of those occurring in whole peat.

2.3 Experimental procedures

Slurry vials were generally preincubated for ∼24 h in the dark at the ambient field temperatures before making any subsequent experimental treatments. Sub-sets of the preincubated vials were used for different treatments, which included chemical amendments and temperature manipulations. Chemical amendments were typically made just before the first measurement, but BES (2 mM) was always added just after crimp-sealing the vials (i.e. before the preincubation period) as this proved to be more effective at inhibiting methanogenesis. All experiments also included a series of control vials that remained untreated (‘No Treatment’). Duplicate vials were used for all experimental treatments. After treatments were made, all vials were incubated in the dark at ambient field temperatures with the exception of those for which temperature was an experimental variable. Sub-samples from the experimental vials were collected over time courses that lasted several days. Time course measurements included determinations of DMS, total CO₂, CH₄, acetate, and inorganic anion concentrations. TMP was assessed at discrete times during the incubations with short-term (1 h) assays. Details on the methods and procedures used for these determinations are described below.

2.4 Sediment amendments

The influence of fresh, recently produced, Sphagnum plant material on DMS accumulation was tested in one experiment by comparing slurried sediments from 5–10 cm depth (i.e. near the surface where living Sphagnum was present) with those from deeper (>15 cm depth) where sediment consisted only of degraded peat.

The effects of specific organic amendments on TMP and anaerobic metabolic activities were tested in several different experiments. Glucose, a representative fermentable carbon substrate, was added to a final concentration of 500 or 670 μM in two separate experiments. Acetate, a major metabolic end product in anaerobic peats [16], was added to a final concentration of 2 mM. Syringic acid, a methoxylated aromatic compound found in lignin, and a known source of methyl groups for some anaerobes [4], was added (20 μM) in one experiment to test whether this compound could serve as a methyl donor for the methylation of sulfides. In separate experiments, methanol was added over a range of concentrations (10–2000 μM) and its effects on methylation activity and DMS production determined.

Incubation temperature was manipulated in one experiment to change metabolic rates while avoiding the potential non-specific effects associated with chemical additions. The experiment was performed using sediments from Boggy Branch Creek, AL (24°C, pH 3.8) that were incubated at temperatures ranging from 11 to 30°C, with heat-killed control vials incubated at 80°C.

Availability of molecular hydrogen (H₂) is critical to terminal metabolism and can influence the relative importance of metabolic processes such as methanogenesis and homoacetogenesis [19]. Therefore, in one experiment, H₂ gas was added to the headspace of anaerobic peat slurries from Boggy Branch Creek to stimulate terminal metabolic processes and determine the relationship between these processes and TMP. Approximately 3 ml (120 μmol) of pure H₂ gas was added to incubation vials. Consumption of added H₂ resulted in loss of headspace pressure. Therefore, vials amended with H₂ were monitored daily and H₂ was added to maintain 3 ml overpressure (approximately 1.4 atm) in the vials.

In nearly all experiments, the role of methanogenesis in thiol methylation and other sediment metabolic processes was investigated by treating one set of slurries with BES to inhibit methanogenesis [20,21]. In addition, the effects of the alternative electron acceptor, nitrate (500 μM as KNO₃), on TMP, DMS formation and methanogenesis was tested in one experiment. The effects of sulfate were not tested because sulfate concentrations in Turnagain Bog are extremely low (<10 μM) [16] and bog peats showed very low potential for sulfate reduction (R. Kiene, unpublished data).

2.5 Measurement of sulfur gases

DMS concentrations were measured by gas chromatography using a cryotrapping system as described previously [10]. Briefly, sediment slurries were sub-sampled by withdrawing 0.5–1.0 ml through the stopper with a Tuberculin syringe. The sub-samples were transferred to a sealed 14-ml vial and vortexed for 5 s to equilibrate the dissolved gases with the headspace. The headspace of the 14-ml vial was then swept for 3 min with a stream of He (100 ml min⁻¹) into a Teflon loop submersed in liquid N₂. Trapped sulfur gases were volatilized by replacing the liquid N₂ with hot water and swept into the gas chromatograph by means of a Rheodyne six-port Teflon valve, which redirected carrier flow through the sample loop. Separation of the sulfur gases was achieved with a Carbopak B 1.5% XE-60, 1% H₃PO₄ (60/80 mesh) column using He (60 ml min⁻¹) as the carrier gas. Sulfur gas analysis was carried out with a Shimadzu gas chromatograph (GC) model 14A or 9A equipped with a flame photometric detector and sulfur-selective filter. The oven temperature was 90°C and injector temperature was 175°C. Retention times for DMS were typically 1.2–1.5 min. DMS analyses were calibrated with permeation standards [10]. Experimentally determined sweep efficiencies were consistently 77±5% for DMS in all sediment types and corrections were made for this efficiency. Limits of detection were approximately 1 nM DMS for both GC models.

2.6 Thiol methylation assay

The potential for sediments to methylate thiols was tested by amending sediment slurries with 8–10 μM ethanethiol (CH₃CH₂SH; ESH) as an artificial methyl receptor and measuring the production of ethylmethylsulfide (CH₃CH₂SCH₃; EMS) by gas chromatography. ESH is an analog of naturally occurring MeSH, the putative precursor of DMS in Sphagnum peats [10]. The added concentrations of ESH were ∼100–1000 times higher than the natural MeSH concentrations. Therefore, under the assay conditions the added ESH should have been the dominant methyl receptor for endogenous methyl donors capable of methylating thiols. Addition of 8–10 μM ESH gave nearly identical methylation rates as the addition of 30 μM ESH (P=0.91, two-tailed t-test) indicating that methylation activity was saturated using 8–10 μM ESH. Thus, the measured methylation potential is the maximum potential rate with natural methyl donors [10,17]. Although added ESH concentrations decreased exponentially due to adsorption to sediments, the short-term incubations (<1 h) ensured that ESH concentrations remained saturating to methylation activity during the incubations.

To perform the assay, 14-ml slurry vials were amended with 8–10 μM ESH (final added concentrations), shaken, and allowed to incubate for 30–40 min. A 1-ml sub-sample of the slurry was removed with a Tuberculin syringe and placed in another sealed 14-ml vial containing 0.2 ml 5 N NaOH and vortexed. The NaOH effectively ended the incubation by killing microorganisms in the sub-sample and it also removed most of the ESH from the volatile phase. Removal of the micromolar levels of ESH greatly improved quantification of the nanomolar levels of EMS. EMS was measured by gas chromatography by the headspace sweeping method as described for DMS above. EMS eluted at ∼3 min and was well separated from endogenous DMS. Detection limits for EMS were typically 1 nM. TMP was calculated as a rate of EMS formation and expressed as nM h⁻¹.

Standard curves of EMS were generated by analyzing different volumes of freshly prepared stock solutions added to sealed empty 14-ml vials. Stock solutions were prepared gravimetrically and analyzed using the cryotrapping method as described above. Sweep efficiencies for EMS were found to be 69±5% and corrections were made for this efficiency.

2.7 Analysis of anions

Anion samples were prepared for analysis by removing a 1-ml sub-sample from anoxic incubation bottles with a Tuberculin syringe fitted with an 18-gauge needle. The sub-samples were placed in 2-ml microcentrifuge tubes and centrifuged at 12 600 rpm for 5 min. The supernatant liquid was removed with a clean Tuberculin syringe and filtered through a 0.2-μm IonChrom Acrodisc filter unit. Samples were stored at −18°C in 2-ml cryule vials until analysis (storage time was typically <48 h).

Anions (acetate, nitrate, and sulfate) were measured with a Dionex DX-120 ion chromatograph equipped with an IONPAC AS14 analytical column. The eluent was a solution of 3.5 mM Na₂CO₃/1.0 mM NaHCO₃ made with >18 MΩ purified water. Eluent flow rate was 1.0 ml min⁻¹. Standard curves were generated using SPEX Standards or Dionex Seven-Anion Standards diluted to known concentrations in 18 MΩ purified water. For acetate, standard solutions were prepared gravimetrically from sodium acetate in 18 MΩ purified water. Detection limits were 5 μM for acetate and about 1 μM for sulfate and nitrate.

2.8 Measurement of methane and CO₂

Methane was analyzed by gas chromatography with a Shimadzu GC-14A equipped with a flame ionization detector. A Porapak Q column was used to separate gases. The carrier was high-grade He (100 ml min⁻¹) and methane typically eluted at 1.0–1.1 min. Vials were shaken to equilibrate CH₄ with the headspace. Sub-samples of 100–250 μl were taken directly from the headspace of incubation vials using a gas-tight Hamilton syringe. Total CH₄ produced was calculated by multiplying concentration in the headspace by volume of headspace at the time of sampling ((μmol CH₄/ml headspace)×ml headspace). This amount was divided by the slurry volume to calculate the equivalent moles of CH₄ per liter of slurry. This facilitated comparisons with other measured dissolved compounds. Standard curves were generated by injecting different volumes of a Scotty Analyzed standard gas mixture (98.6 ppm CH₄) from a lecture bottle into the GC.

CO₂ was analyzed with a Shimadzu GC-8 equipped with a thermal conductivity detector and a Porapak N column. Helium was used as the carrier gas (100 ml min⁻¹) and the oven temperature was 40°C. Measurements of CO₂ were performed on series of 14-ml vials prepared and incubated alongside the main experimental vials. At specified times in the experiment, 0.1 ml of 10% HCl was added to these vials to end the incubation and drive all of the CO₂ into the headspace. Vials were vortexed and stored until analysis. CO₂ was measured by withdrawing 100–250 μl headspace and injecting it directly into the GC. Standards curves were generated using a combination of Scotty Analyzed gas from a lecture bottle and stock solutions prepared gravimetrically using sodium bicarbonate and 10% HCl.

2.9 Statistical tests

Measurements of DMS and acetate concentrations over time allowed rates of consumption and accumulation to be calculated using least-squares linear regression. When appropriate, these rates were compared using analysis of covariance [22] with α set at 0.05. Effects of chemical additions on TMP were evaluated with one-way analysis of variance (ANOVA) at each time point. The effect of individual treatments was tested using post-hoc analysis (LSD) with α set at 0.05. Regression between TMP and net DMS accumulation in unamended peat slurries was tested using linear least-squares regression. Fit of the regression is expressed in terms of fit (r²) and significance (P) with α set at 0.05. Samples in which net DMS consumption was evident were excluded from the regression. The raw data did not follow a normal distribution, therefore, log-transformed thiol methylation and DMS accumulation rates were used in the regression.

Sigmaplot 2000 was used to test a non-linear regression between added methanol and TMP (Fig. 5) using the model y=y₀+[(V_max×S)/(K_s+S)], a modified version of the Michaelis–Menten equation. For the purposes of this regression, y₀ is TMP in unamended slurries (nM h⁻¹), V_max is the calculated maximum TMP (nM h⁻¹), S is MeOH added (μM), and K_s is the apparent half-saturation constant (μM).

5

TMP in peat slurries amended with different concentrations of methanol (MeOH). Peat for this experiment was collected from Talkeetna Bog, AK on 08-06-99.

2.10 Chemicals

All chemicals were obtained commercially from either Sigma-Aldrich or Fisher Scientific, and were of the highest purity available.

3 Results

3.1 Relationship between DMS accumulation and TMP

Thiol methylation activity was detectable in all sediments tested. Rates of methylation potential ranged from 12 to 154 nM h⁻¹ with higher rates generally occurring late in the incubations. Natural DMS accumulation rates were much lower than potential thiol methylation rates ranging from <1.0 to 22.2 nM h⁻¹. There was a significant relationship between TMP and rates of DMS accumulation (r²=0.48, P=0.0007) (Fig. 1). This regression includes data from both the high-latitude site (Turnagain Bog, AK) and low-latitude site (Boggy Branch Creek, AL). Only peat slurries with no evidence of DMS consumption were considered for the regression since consumption of the DMS pool would presumably mask any relationship between DMS production and its accumulation in the sediments.

1

Regression between DMS accumulation rates and potential thiol methylation rates in all sediments tested. Regression includes unamended slurries from three separate experiments and is for periods when no DMS consumption was apparent. The equation for the regression is log DMS=−1.53(±0.44)+1.047(±0.26) log TMP (r²=0.48, P=0.0007, n=20).

3.2 Methylation potential in different depth horizons

In sediments from Boggy Branch Creek, TMP was much higher in the deeper horizon (>15 cm depth) than in the shallower horizon (5–10 cm depth) (Fig. 2A). The deep sediments also accumulated DMS more rapidly than shallow sediments early in the incubation before the development of net DMS consumption (Fig. 2A). In contrast, sediments collected from the shallow 5–10-cm depth horizon accumulated DMS very slowly and showed no evidence of consumption during the 4.5-day experiment (Fig. 2A). Sediments from the upper horizon, which were richer in fresh Sphagnum material, accumulated far greater amounts of acetate than peat collected from deeper than 15 cm (Fig. 2B). However, deep sediments accumulated nearly twice as much CO₂ as the shallower sediments (data not shown), and had much higher rates of methanogenesis (Fig. 2C), suggesting that terminal metabolism was greater in the deeper sediments.

2

Time course of (A) TMP (bars) and DMS concentration (circles), (B) acetate concentrations, and (C) methane concentrations in sediments collected at 5–10 cm (open symbols) and >15 cm depth (closed symbols) from Boggy Branch Creek, AL (3-26-01).

3.3 Effects of carbon substrate additions

When anaerobic sediment slurries from the low-latitude site at Boggy Branch Creek, AL (24°C, pH 3.8) were amended with 670 μM glucose, TMP was consistently lower than in unamended controls, but the effect was significant only at 191 h (P=0.011, one-way ANOVA) (Fig. 3A). Glucose-amended samples produced greater amounts of acetate, reaching 1750 μM compared to only 600 μM in the control (Fig. 3B). CO₂ production rates were not stimulated by glucose additions (one-way ANOVA, data not shown); however, glucose-amended vials had significantly greater rates of methane production for the first 91 h of incubation (Table 1). DMS production was not measured in this experiment.

3

Effects of H₂, 670 μM glucose, or 22 μM syringic acid on (A) TMP and (B) acetate concentrations over time in slurried peat collected from Boggy Branch Creek, AL (7-13-01). No acetate data were available for the syringic acid treatment.

1

Effect of glucose and hydrogen on rates of methanogenesis in sediments from Boggy Branch Creek, AL (07-13-01)

Treatment	Methane production rate (nM h−1±S.D.) over time interval:
	46–65 h	66–91 h	119–148 h	164–184 h
None	8.34±3.75	43.3±0.60	124±32.0	159±12.2
+500 μM glucose	31.4±1.79*	75.5±2.27*	104±17.6	212±27.8
+Hydrogen	25.2±0.23*	51.7±1.66*	132±15.0	180±12.9

Methanogenic rates were calculated as the increase in measured CH₄ concentration over the specified time interval.

*Observed rate is significantly greater than in No Treatment samples (α=0.05).

As found with the low-latitude peat sediments, TMP in high-latitude sediments from Turnagain Bog, AK (15°C, pH 4.9) was inhibited by additions of glucose (P=0.007, two-way ANOVA; Fig. 4A), and DMS production was inhibited by glucose as well. The addition of 500 μM glucose to these sediments increased acetate concentrations by >300 μM during the first 20 h of incubation. CO₂ concentrations also increased nearly 600 μM over the first 40 h, however, by the end of the incubation, CO₂ concentrations were not significantly different (P=0.45, two-tailed t-test) in samples amended with glucose relative to unamended controls (Fig. 4C). Glucose additions also failed to stimulate methanogenesis in this experiment with Turnagain Bog sediments (Fig. 4C). In a separate experiment, addition of acetate (2 mM) significantly inhibited TMP in Turnagain Bog sediments (Table 2), suggesting that high acetate accumulations in glucose treatments might have caused the lower TMPs (Fig. 4A).

4

Effects of glucose addition (500 μM) on the time course of (A) TMP (bars) and DMS concentration (circles), (B) acetate concentrations, and (C) CO₂ (circles) and methane concentration (triangles) in peat slurries from Turnagain Bog, AK (7-26-99). Open symbols or bars represent the glucose treatment, whereas closed symbols or bars represent unamended controls.

2

Effects of acetate, methanol or BES (2 mM each) on TMP in peat slurries from Turnagain Bog, AK

	Treatment	TMP (nM h−1±S.D.)
Experiment #1	None	1098±126
	Acetate	490±60
	Methanol	3477±340
Experiment #2	None	720±11
	BES	288±19

The assays were conducted at 22°C.

Addition of methylated organic compounds (syringic acid or methanol) effectively stimulated TMP in Sphagnum-dominated sediments. For example, 22 μM syringic acid added to sediments from Boggy Branch Creek, AL led to significantly greater TMP relative to unamended samples (two-way ANOVA, Fig. 3A). In an experiment carried out using Sphagnum bog sediments from Talkeetna, AK (15°C, pH 4.1), a range of added methanol concentrations (10–500 μM) stimulated methylation potential with the 500 μM methanol treatment resulting in thiol methylation rates more than five times greater than those from unamended vials (Fig. 5). The concentration dependence fit a modified Michaelis–Menten curve (r²=0.799, P=0.0003, non-linear regression), implying a maximum methylation rate (V_max) of 183 nM h⁻¹ and background methylation potential of ∼40 nM h⁻¹. In separate experiments, methanol additions strongly stimulated TMP in Turnagain Bog sediments (Table 2), and also stimulated DMS production (Fig. 6). Like that observed for methylation potential in the Talkeetna peat, the response of DMS production in the Turnagain Bog peat to methanol additions appeared to follow a saturation function.

6

A: Time course of DMS production in peat slurries from Turnagain Bog treated with different concentrations of methanol. B: Dependence of DMS accumulation rate on added methanol concentration.

3.4 Effects of BES

Addition of 2 mM BES, an inhibitor of methanogenesis, to sediments from Boggy Branch Creek, AL (14°C, pH 3.6) inhibited CH₄ production by >80% (data not shown) and TMP by 75–85% (Fig. 7A). Accordingly, vials amended with 2 mM BES also had slower rates of DMS accumulation early in the incubation (Fig. 7B). Late in the incubation net DMS consumption developed in unamended vials, whereas DMS concentrations remained constant in BES-amended vials suggesting no net DMS production or consumption.

7

Effects of 2 mM BES on the time course of (A) TMP and (B) DMS concentrations in peat slurries from Boggy Branch Creek, AL (02-25-99).

The inhibitory effects of BES on TMP were also observed in peat slurries from Turnagain Bog, AK (15°C, pH 4.9) (Table 2). In several incubations with Turnagain Bog sediments, 2 mM BES additions consistently inhibited DMS production by about 40–50%, while at the same time inhibiting methanogenesis 50–100% (data not shown). On the other hand, acetate concentrations were not affected by the addition of BES (P=0.26, two-tailed t-test) (data not shown).

3.5 Effects of nitrate addition

Endogenous nitrate concentrations in peat slurries were at or below the limit of detection. However, when 1.5 mM KNO₃ was added to experimental slurries from Boggy Branch Creek (19°C, pH 4.0), nitrate consumption was observed immediately (Fig. 8A). Like BES, KNO₃ effectively inhibited TMP (Fig. 8A) and methanogenesis (Fig. 8C). After NO₃⁻ was completely consumed, TMP recovered to approach the values in control bottles. Methanogenesis appeared to recover at this time as well (Fig. 8C). DMS accumulated more slowly in KNO₃-amended vials and net DMS consumption did not develop after more than 160 h of incubation (Fig. 8B). Both unamended controls and KNO₃-treated samples accumulated acetate early in the incubation to a concentration of nearly 550 μM. However, after 43 h of incubation, unamended vials had steady-state concentrations of acetate whereas net consumption developed in KNO₃-amended vials (Fig. 8B).

8

Effects of nitrate addition on (A) TMP (bars) and nitrate concentrations (circles), (B) DMS (circles) and acetate (triangles) concentrations, and (C) methane concentrations, in sediment slurries from Boggy Branch Creek, AL (03-26-01). Open symbols or bars represent the nitrate treatment, whereas the closed symbols or bars represent unamended controls.

3.6 Effects of hydrogen addition

Addition of H₂ to anaerobic slurries from Boggy Branch Creek stimulated acetate production (Fig. 3B), suggesting that homoacetogens were present in these sediments and potentially H₂-limited. Methanogenic rates were also greater in H₂-amended vials (Table 1), but the effect was not significant after 91 h incubation time. Despite the significant increases in acetogenesis and marginal increases in methanogenesis, H₂ additions did not affect TMP (Fig. 3A).

3.7 Influence of incubation temperature

Sediments from Boggy Branch Creek were collected in summer (in situ temperature=24°C) and slurries were incubated at different temperatures in an attempt to manipulate microbial activities without chemical additions. TMP rates were 8.9 nM h⁻¹ at 11°C and increased to 44 nM h⁻¹ at 22°C but methylation potential was not significantly stimulated above 22°C (Fig. 9A). No significant differences in CO₂ production rates were detected between vials incubated at different temperatures below 30°C, although they were highest at 22–30°C (Fig. 9B). Methane production rates, however, increased by two orders of magnitude over the range of temperatures tested from 6.7 nM h⁻¹ at 11°C to 708 nM h⁻¹ at 30°C (Fig. 9C). In the 80°C heat-killed vials, TMP, methane production and CO₂ production were either not detectable or near the limit of detection.

9

Effects of incubation temperature on (A) TMP, (B) carbon dioxide accumulation rate, and (C) methane accumulation rate in peat slurries from Boggy Branch Creek, AL (07-13-01). The natural temperature for this peat sediment was 24°C.

4 Discussion

The results of this study provide evidence that thiol methylation is an important microbial process in the formation of DMS in anoxic, low-pH Sphagnum peats from both high and low latitudes. The relationship between potential methylation rates and DMS accumulation (Fig. 1) suggests that methylation rates can influence DMS accumulation in sediments that have little or no DMS consumption activity. On average, potential thiol methylation rates were nearly an order of magnitude greater than DMS accumulation rates. This result is not surprising since the methylation assay was carried out under saturating concentrations of methyl acceptor (ESH). The results support findings by both Lomans et al. [11] and Kiene and Hines [10] that DMS production can be stimulated by addition of thiols and that methylation is limited in part by methyl acceptors (e.g. MeSH in natural sediments).

The addition of naturally occurring methyl donor compounds such as syringic acid (Fig. 3A) and MeOH (Fig. 5, Table 2) stimulated potential methylation rates, and sediment DMS production (Fig. 6; see also [10]), suggesting that these types of compounds might be sources of methyl groups transferred to MeSH in sediments. Methylation potential in Talkeetna Bog sediments responded to MeOH additions according to Michaelis–Menten saturation kinetics. The data fit a modified Michaelis–Menten equation with a term (y₀) added to account for TMP in unamended sediment slurries. This relationship would be expected since there was presumably some background pool of methyl-donating compounds in the sediments besides the added MeOH. The fact that MeOH stimulated both TMP (Fig. 5; Table 2) and DMS production (Fig. 6), albeit in different experiments, provides further evidence of the linkage between these microbial processes. Methoxyaromatic acids are a product of lignin degradation [23] whereas MeOH formation can occur through degradation of either pectin or lignin (see [19]). Large amounts of pectin or lignin may be delivered to peatlands during plant senescence or autumn leaf drop and may result in elevated concentrations of methyl-donating compounds during brief periods. The relatively rapid response of sediment microorganisms to high concentrations of MeOH (Figs. 5 and 6) suggests that they are capable of increasing metabolic rates to take advantage of these inputs.

TMP did not appear to be directly related to total anaerobic metabolism. Additions of glucose inhibited TMP despite stimulating acetate production (Figs. 3 and 4) and methane production (Table 1). Furthermore, enhancement of nitrate reduction, which stimulated overall anaerobic metabolism and acetate consumption, actually inhibited TMP, rather than stimulating it (Fig. 8). Therefore it is unlikely that TMP is related to general anaerobic metabolism, but rather, it appears to be due to a discrete group of microorganisms carrying out a specific reaction (see discussion further below). On the other hand, it appears that acetate does have a strong influence on methylation potential, and ultimately DMS formation. Direct addition of acetate (Table 2) or experimental treatments that increased acetate concentrations (glucose addition, inclusion of fresh Sphagnum biomass (Figs. 2–4) all decreased methylation potentials and DMS production. Kiene and Hines [10], working with peat from Sallies Fen in New Hampshire, also found that addition of acetate (1.5 mM) inhibited DMS production. Those authors speculated that high concentrations of acetate might inhibit terminal anaerobic metabolism, particularly acetogenesis, thereby inhibiting methylation reactions leading to DMS formation. However, high concentrations of acetate did not appear to inhibit CO₂ or CH₄ production (Fig. 4, Table 1) indicating that overall metabolic rates did not change significantly. Further study will be needed to ascertain why acetate affects methylation and DMS production.

Stimulation of TMP by addition of syringic acid (Fig. 3A) suggests a possible role of homoacetogenic bacteria in DMS formation [4,8]. However, direct evidence linking homoacetogenic activity to thiol methylation in situ is still lacking. For example, increased acetate production in H₂-amended incubation vials (Fig. 3B) is indicative of stimulated homoacetogenesis, but was not accompanied by stimulated thiol methylation activity (Fig. 3A). Schink [24] pointed out that homoacetogens growing by CO₂ reduction have a higher requirement for H₂ than those growing on other substrates. Therefore, adding H₂ to the incubation vials may have stimulated CO₂ reduction by homoacetogens without affecting C₁ utilization by acetogens (and thereby thiol methylation).

The well-known inhibitor of methanogenesis, BES, decreased both TMP (Fig. 7A) and DMS accumulation (Fig. 7B) [10,14], suggesting that methanogens may play a role in methylation processes and DMS formation in low-pH wetlands. Transitory accumulation of DMS in the presence of high concentrations (0.2 to >30 mM) of MeSH by methanogens grown in culture has been observed [25,30]. However, a linkage between methanogenesis and DMS production has not been reported before in natural sediments. On the contrary, methanogens are thought to be the main consumers of DMS in low-sulfate sediments [12,13,26]. Evidence for a role of methanogenesis in thiol methylation is somewhat complicated by the results of the temperature manipulation experiment. While methanogenic activity increased dramatically with increasing temperature (Fig. 9C), TMP did not increase at incubation temperatures above 22°C (Fig. 9A). It is possible that only a sub-set of the methanogenic population was responsible for thiol methylation activity and that these did not respond to temperature in the same way as the total population. It also remains a possibility that suppression of methanogenesis by BES indirectly inhibits methylation, perhaps by affecting hydrogen metabolism and shifting acetogenic pathways.

Inhibition of H₂/CO₂ methanogenesis by BES could allow H₂ to accumulate and stimulate H₂/CO₂ acetogenesis. H₂/CO₂ methanogenesis dominates methane production in Turnagain Bog [16], raising the possibility that BES might have increased H₂ concentrations. Higher H₂ could increase acetogenesis, as we observed (Fig. 3B), but H₂ additions had no effects on methylation potential (Fig. 3A), nor on DMS production (Kiene, unpublished data). Furthermore, BES had no significant effects on acetate accumulation in Turnagain Bog peat [16]. Therefore, increased H₂ availability does not fully explain why additions of BES result in lower TMP. The fact that inhibition of methanogenesis by 1.5 mM KNO₃ also was accompanied by a similar decrease in TMP (Fig. 8A) and DMS accumulation (Fig. 8B) makes it unlikely that any non-specific effects associated with BES are responsible for these results. Taken as a whole, our results support the conclusion that methanogens were involved in thiol methylation.

A key intermediate step in methanogenesis is the methylation of 2-mercaptoethanesulfonic acid (HS-CoM) [27] by a methyltransferase enzyme [28]. Working with cell extracts of Methanobacterium thermoautotrophicum, Gunsalus et al. [29] found that the methanogenic pathway functioned using several analogs of HS-CoM. This suggests that there is a degree of non-specificity to the methyltransferase enzyme. Since the enzyme is configured to methylate a thiol moiety on HS-CoM, it is possible that methylation of MeSH could occur as an ‘accidental reaction’ during methanogenesis (see also [25,30]). This type of reaction would interrupt the methanogenic pathway and therefore represents an inefficiency in methanogenesis in the presence of thiols. However, the total effect of this process on methanogenesis would likely be small since methane accumulation rates were typically several orders of magnitude larger than DMS accumulation rates. Nonetheless, this mechanism might be important for DMS production in wetland sediments where methanogenesis is a major terminal metabolic process. It remains to be determined whether different sub-populations within the methanogen community are responsible for producing and consuming DMS. Only a limited number of methanogen lineages are capable of utilizing DMS as a substrate [26,30], and the enzymatic methyl transfer reactions involved in DMS degradation appear to be specific for DMS, and distinct from those used in methanol metabolism [28]. DMS consumption was only occasionally observed in acidic peats during this study, but when it was, experiments with BES suggested it was linked with methanogenesis (Stets and Kiene, unpublished observations). Resolving the issue of methanogenic participation in thiol methylation activity in low-pH peat sediments will require much more detailed studies of the biochemical pathways of methanogenesis. However, this study does provide strong evidence that thiol methylation contributes to DMS formation in low-pH peat sediments. Furthermore, it confirms earlier results that both methyl acceptors (e.g. MeSH) and methyl-donating compounds (syringic acid and methanol) can influence DMS production in anaerobic sediments [10,11]. This study also provides evidence that methylation activity is not related to general anaerobic metabolic activity, but rather to action of a discrete group (or groups) of microorganisms carrying out a very specific reaction.

Acknowledgements

Funding for this research was provided by a grant from the National Science Foundation (DEB-96-32421). Laura Linn and Krystyne Duddleston provided helpful assistance with various aspects of this study. Jonathan Pennock provided helpful comments on an earlier draft. E.G.S. would like to thank the Dauphin Island Sea Lab and the Department of Marine Sciences for financial support during this study.