INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Massive communities of Thioploca species occur as dense mats in the top sediment underlying the oxygen minimum zone of the continental shelf off the coast of Chile and Peru (17). Extending down to 5 to 10 cm into the sediment, the total biomass (including sheaths) of these colorless sulfur bacteria may be as high as 800 g (wet weight) m² (32), potentially covering several thousands of square kilometers along a 3,000-km stretch of coast.

Thioploca chileae and Thioploca araucae are the two dominant species in the mat, measuring 12 to 22 and 28 to 42 µm in diameter, respectively (32, 36). Both species produce 2- to 7-cm-long trichomes (filaments), each of which consists of a uniseriate row of many vacuolated cells. Morphologically and phylogenetically, they are similar to vacuolated Beggiatoa species (24, 37), and it has been suggested that their physiology might be similar as well. A chief difference between the genera is, however, that Thioploca produces characteristic bundles of usually 10 to 20 trichomes, surrounded by 10- to 15-cm-long sheaths up to 1.5 mm in diameter. Individual trichomes can glide independently within the sheaths and extend up to 3 cm into the water phase above the sediment (16). In general, Thioploca species and Beggiatoa species appear to occupy different niches, the former living in vertical and horizontal sheaths down to 10 cm in sediments that contain relatively little sulfide. In contrast, Beggiatoa species live in the top layer of sediments that have relatively high sulfide concentrations. Since their discovery by V. A. Gallardo (8, 9), it has been assumed that the Thioploca mats play a crucial role in balancing the sulfur cycle of their marine habitat by reoxidizing all, or at least a substantial portion, of the sulfide produced in the sediment. The sulfide results from high rates of bacterial sulfate reduction, up to 2.4 g of sulfur m² day¹ (6), driven by extremely high primary productivity (up to 9.6 g of carbon m² day¹) over the continental shelf (7). Recently, Thioploca spp. have been identified off the coast of Namibia (10), where similar oceanographic conditions exist, i.e., upwelling, high primary productivity, and oxygen-depleted bottom water.

Given that the high remineralization rates of organic compounds result in the often-observed depletion of oxygen in the bottom water overlying the sea floor (7, 9, 32), the question arose as to which electron acceptor might be used for the reoxidation of all the sulfide produced in these sediments. When it was discovered that the vacuolated Thioploca species living in the mat were capable of accumulating up to 500 mM nitrate from the overlying water (containing ~25 µM [7]), it was hypothesized that Thioploca species would be able to use the nitrate as a terminal electron acceptor for sulfide oxidation (7). It was assumed that nitrate would be reduced to dinitrogen, although no experimental data was available to support this (7). The question was, therefore, still open as to whether dinitrogen gas or ammonium would be the product. This was particularly interesting in view of the finding by McHatton et al. (21) that vacuolated Beggiatoa species are also capable of accumulating and reducing nitrate and in view of conflicting observations by others with respect to the final product of nitrate reduction by the nonvacuolated Beggiatoa alba, i.e., ammonium or dinitrogen gas (35, 39).

So far, it has not been possible to cultivate Thioploca species in pure culture. The same is true for the vacuolated (nitrate accumulating) Beggiatoa species. Hence, little is known about their (eco)physiology, specifically, their sulfide and sulfur oxidation rates, abilities to respirate oxygen and/or nitrate, growth rates, or capabilities to grow autotrophically, heterotrophically, or mixotrophically. Clearly, this knowledge is essential for understanding the role of Thioploca species in their habitat.

McHatton et al. (21) studied partially purified cultures of naturally occurring populations of large vacuolated Beggiatoa species and showed that these organisms contain substantial activities of membrane-bound nitrate reductase, indicating that they may indeed be capable of using nitrate as the terminal electron acceptor. Significant activities of ribulose-1,5-bisphosphate carboxylase were also detected, evidencing that the vacuolated marine Beggiatoa species are capable of autotrophic growth. Using rinsed samples of Thioploca material from a mat, Ferdelman et al. (6) were able to demonstrate CO₂-fixing capacity in these preparations, indicating that Thioploca has an autotrophic potential.

Since Thioploca species live at high densities in the mats of the Chilean marine sediments and could be seen with the naked eye, we decided that it was possible to obtain samples of these organisms sufficiently pure to allow the performance of physiological experiments. We developed a simple method by which we handpicked individual sheaths with trichome bundles with forceps from the top 2 cm of sediments incubated under an N₂ atmosphere. After they had been collected and washed, cells were used for various experiments. By using radiolabeled and unlabeled substrates, a study was made of carbon, nitrogen, and sulfur metabolism. The observed activities were compared with data obtained from field measurements. The results indicate that Thioploca species are (metabolically) highly active under anoxic conditions and that they can play a significant role in the total oxidation of sulfide in the mat under anoxic conditions in the presence of nitrate. They appear to be facultative chemolithoautotrophs with a mixotrophic potential, meaning they can use sulfide or sulfur as an energy source for growth and CO₂ fixation and can use acetate under these conditions as an additional carbon source. Evidence presented in this study points to ammonium as the end product of nitrate reduction, although conversion to dinitrogen gas cannot be ruled out. Oxygen at approximately 10% air saturation did not inhibit the observed CO₂ fixation.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Samples were collected in January and February 1997 on the continental shelf within the Bay of Concepción, central Chile, onboard the research vessel Kay Kay, and the laboratory work was performed at the Marine Biological Station of Dichato, both of the University of Concepción. Sampling was performed at 34 m of water depth at station 7 (32), at 36°36'5"S, 73°00'6"W. At this station, at the time of sampling, the percentage of organisms in the upper 2 cm of the sediment found to be T. araucae was 39 to 67%. The ratio of the biovolumes of T. araucae and T. chileae is approximately 70:30, and therefore, the majority of the community consisted, in biovolume, of T. araucae.

Collection. Sediment samples were obtained by a Rumohr corer with Plexiglas cores (inside diameter, 9.5 cm), stored at 4°C, and processed within 8 days of sampling. The top 1 to 2 cm was removed and placed on ice in an N₂-filled glove bag (Sigma-Aldrich, Zwndrecht, The Netherlands). Thioploca sheaths with bundles of trichomes (1 to 2 cm in length) were collected with forceps and transferred to synthetic medium (containing [per liter] 25 g of NaCl, 6 g of MgSO₄ · 7H₂O, 1 g of CaCl₂, 0.5 g of K₂HPO₄, 0.1 g of KH₂PO₄, and 0.5 g of NaHCO₃ [pH 7]) or to synthetic medium without NaHCO₃, supplemented with 0.05% (wt/vol) thioglycolate and 1 mg of catalase per liter (Sigma-Aldrich). The latter medium (hereafter referred to as medium, unless stated otherwise) was found to give the best results. Media were sparged with N₂ for 30 min for anoxic conditions and subsequently stored at 4°C. Shortly before use, media were sparged again with N₂ for 10 min, unless stated otherwise, while kept cold. Under an N₂ atmosphere in a glove bag, the sheaths with trichome bundles were washed twice by transfer into fresh medium and were incubated in airtight 6-ml vials equipped with rubber septa (Exetainer; Labco, High Wycombe, United Kingdom). The medium in the vial was changed twice by decanting under an N₂ atmosphere, with minimal disturbance of the sheaths and trichome bundles. The incubation volume was adapted to the conditions of the experiment performed. Under anoxic conditions, substrates were injected through the rubber septum. Residual sediment, left after washing Thioploca trichome bundles, was used as a control. In these controls, the amount of sediment used was 5 to 10 times higher than the estimated contamination of sediment attached to washed sheaths with trichome bundles. Another control consisted of disrupted Thioploca sheaths with trichome bundles (equal to the amount used in the experiment) which had been disintegrated mechanically by a Potter-Elvehjem homogenizer (Fisher Scientific, Zoetermeer, The Netherlands). This control was necessary because observations under the fluorescence microscope had shown that the sheaths were covered with epibiontic bacteria, including sulfate-reducing filamentous bacteria of the genus Desulfonema (36, 40). After the Potter-Elvehjem treatment, it was observed under the fluorescent microscope that Thioploca trichomes were mechanically disrupted, while the majority of the epibiontic bacteria remained intact. Bundles consisting of sheaths with bundles of trichomes will hereafter in this paper be referred to as "trichome bundles," while sheaths with Potter-Elvehjem-treated bundles will hereafter be referred to as "disrupted trichome bundles."

Incubation. For each experiment, approximately 100 Thioploca trichome bundles were collected in a final volume of 3.5 ml under an N₂ atmosphere in gas-tight vials, unless stated otherwise. Vials were incubated with substrates in a water bath at approximately 12°C. At specific time intervals, samples were taken with a syringe previously flushed with dinitrogen and analyzed for ammonium, nitrite, sulfide, thiosulfate, and sulfate.

Analytical procedures. Nitrite and ammonium in the supernatant were determined colorimetrically (as described in references 14 and 5, respectively). Intracellular nitrate concentrations were measured with a miniaturized version of the standard colorimetric method of Grasshoff et al. (13). Nitrate was measured in 100-µl extracts of rinsed and dried Thioploca trichomes. Trichomes 5 to 40 mm in length were dissected under the microscope with the help of forceps and needles. Length and width of these trichomes were measured and, after washing and drying, the filaments were resuspended in 50 µl of distilled water to measure the nitrate concentration. Biovolume was calculated from trichome length and width. An average nitrate concentration (n = 27) of 160 ± 150 mM was found. Protein was determined by the microbiuret method of Goa (11). The observed protein concentrations were in agreement with calculations for expected protein content. Where no protein measurements were available (in experiments where labeled compounds were used), a protein content of 315 ± 95 µg was assumed for 100 Thioploca sheaths with trichomes (which is based on an average of 34 protein measurements of Thioploca samples). Thiosulfate was derivatized with monobromobimane (4) and analyzed by reversed-phase high-performance liquid chromatography (28). Sulfide was determined either colorimetrically according to the method of Cline (2) or by the method described for thiosulfate determination. Standards for sulfide and thiosulfate were prepared in degassed sulfate-free medium (MgCl₂ instead of MgSO₄ and without thioglycolate and catalase). Sulfate was determined by nonsuppressed ion chromatography as described by Ferdelman et al. (6). Since high concentrations of chloride interfere with sulfate analysis by ion chromatography, chloride was removed from the samples by adding 40 mg of Ag⁺-loaded cation exchange (Ag 50W-X8; Bio-Rad) per 150-µl sample and incubating for 2 h at room temperature. After centrifugation and filtering, the sample was analyzed. Standards were treated in the same way.

¹⁵N experiments and mass spectrometry. Under anoxic conditions, a concentrated solution of Na¹⁵NO₃ was added to gas-tight vials, each with 120 Thioploca bundles in 4.5 ml of medium. No direct protein measurements could be performed and, therefore, a total protein content of 378 ± 114 µg was assumed on the basis of the average protein content of 100 Thioploca bundles (see above). The headspace was changed with He, and at certain time intervals samples were taken for analysis. Total nitrite and ammonium concentrations were measured immediately after centrifugation. To determine the concentration of ¹⁵NO₃, ¹⁵NO₂, and ¹⁵NH₄⁺, samples were removed with a syringe, centrifuged, sterilized by passing the supernatant through a 0.2-µm-pore-size filter (Dynagard; Microgon Inc., Laguna Hills, Calif.), acidified to pH 4 to 5, and stored at 20°C until the time of analysis. At the end of the experiment, 1 to 2% (final concentration) formaldehyde was added to the vial and pressure was equilibrated with He. Vials were stored at 4°C until the headspace could be analyzed for ¹⁵N₂. Analysis and mass spectrometry were performed at the Institute of Biological Sciences, University of Aarhus, Aarhus, Denmark. For determining concentrations and isotopic compositions of ¹⁵NO₃ and ¹⁵NO₂, samples were neutralized and incubated with denitrifying bacteria to convert these compounds to N₂ for mass spectrometry analysis (29). The labeling pattern of the obtained N₂ gives an indication of the ratios of labeled and unlabeled nitrate and nitrite present in the samples. However, the denitrifiers used can reduce both nitrate and nitrite. Therefore, the recovered N₂ is the product of the reduction of nitrate as well as nitrite present in the medium. Standards with known concentrations (120 µM) of ¹⁵NO₃ and ¹⁵NO₂ were included in the assay and confirmed that the conversion efficiency was consistent (standard deviation = 4%) and that residual nitrate and nitrite concentrations were insignificant.

NaH¹⁴CO₃, [2-¹⁴C]acetate, and [³H]acetate incorporation experiments. Several vials were incubated with approximately 30 Thioploca bundles in 1.3 ml of medium, in the dark. On the basis of the average protein content of 100 Thioploca bundles (see above), the total protein for 30 bundles was assumed to be 94.5 ± 28.5 µg. One hundred micromolar NaNO₃, 25 µl of CO₂ gas (headspace was approximately 5 ml), and, for all experiments, 1 mM HCO₃ were added to the medium. Under anaerobiosis, 0.005 nmol of [³H]acetate (~500 µCi) or 0.034 µmol of a [¹⁴C]acetate solution (~5 µCi) was added for the labeled-acetate experiments. For the labeled-bicarbonate experiments, 0.139 µmol (~10 µCi) or 1.85 µmol of a neutralized NaH¹⁴CO₃ solution (~100 µCi for microautoradiography experiments) was added. Experiments with [¹⁴C]bicarbonate were performed in the absence as well as in the presence of approximately 70 µM sulfide. At certain time intervals, a vial was opened and the supernatant was analyzed for NO₂ and NH₄⁺. The pellet of Thioploca bundles or debris from disrupted bundles was washed four times (by vigorous mixing and subsequent centrifugation) in medium containing 10 mM acetate (when incubated with labeled acetate) or containing 10% trichloroacetic acid (when incubated with NaH¹⁴CO₃) and then added to 2.4 ml of H₂O and 7.5 ml of scintillation liquid (EcoLite [+]; ICN Biomedicals). This suspension was subsequently analyzed in a scintillation counter (Packard liquid scintillation analyzer model 1600 TR). When necessary, CO₂ was trapped by suspending a small cup filled with 100 µl of 2 M NaOH in the gas-tight vial. After the vial was opened, this solution was neutralized and added to scintillation liquid and counted. Experiments with [¹⁴C]acetate took 3 h, experiments with NaH¹⁴CO₃ required 4 h of incubation (to test the influence of oxygen, trichomes were incubated for 22 h), and trichomes used for microautoradiography were incubated for 4 h or 20 to 22 h.

Microautoradiography. Microautoradiography was performed on the experiments described above. Following incubation with [³H]acetate or with NaH¹⁴CO₃, bundles were washed six times with medium containing 1 mM acetate or 1 mM HCO₃, respectively. Individual sheaths were then sorted onto 25-mm hydroxyapatite Millipore filters (Millipore Corp., Bedford, Mass.) or left in 2 ml of medium containing 2% formaldehyde. Filters were subsequently washed in filter-sterilized (first through 0.45-µm-pore-size then through 0.2-µm-pore-size Gelman filters [Millipore]) medium with 1 mM phosphate buffer. After drying, the filters were stored at 4°C. At the end of the cruise, the filters and samples, stored in 2% formaldehyde, were analyzed. Some filters were stained with 2% (wt/vol) erythrosin-B (Sigma-Aldrich) and were subsequently destained by placing them face up on deionized H₂O-saturated pieces of gauze. Filters were air dried, attached to microscope slides, and optically cleared by fuming acetone (27). Cleared filters were prepared for microautoradiography by dipping in Kodak NTB-2 nuclear track emulsion. After exposure (1 to 3 weeks), autoradiographs were developed (Kodak D-19 developer), fixed, rinsed, and air dried prior to microscopic examination with a Nikon Labophot 2 phase-contrast microscope at ×200 to ×400 magnification. Photographs were recorded on either Ilford Pan-F fine-grain black-and-white or Kodachrome 200 color slide 35 mm film.

Calculations. (i) ¹⁵N experiments: the ratio between added [¹⁵N]nitrate and intravacuolar [¹⁴N]nitrate. Addition of 100 µM [¹⁵N]nitrate in 4.5 ml of medium yields 0.45 µmol of [¹⁵N]nitrate. One hundred twenty Thioploca bundles have a protein content of approximately 0.38 ± 0.11 mg (see analytical procedures). Assuming that 50% of the dry weight is protein, and 24% of the wet weight is dry weight and knowing that 90% of the cell is vacuole (as measured in this study and by Maier et al. [20]), then the total wet weight of 120 bundles is 0.38 × 2 × [100/24] × [100/10] = 31.7 ± 9.17 mg, of which 28.5 ± 8.25 mg is vacuolar liquid. Assuming that 1 mg is equal to 1 µl of liquid in the vacuole, then the volume of all vacuoles in the bundles used for the experiments will be approximately 28.5 ± 8.25 µl. If all the added [¹⁵N]nitrate is transported into the vacuoles, then this would lead to a concentration of the label of 15.8 ± 4.57 mM (0.45 µmol in 28.5 µl). Since the vacuoles are filled with an average of 160 mM (see Materials and Methods) unlabeled nitrate, the labeled nitrate will be diluted to 9.9% ± 2.85%. If Thioploca trichome bundles were damaged and all internal nitrate were released, then approximately 5.6 µmol (160 mM in 28.5 µl) would be released into 4.5 ml of medium. This would lead to an increase in nitrate concentration of 1.2 mM, i.e., a 12-fold increase, which would be visible during measurement of the ¹⁵N:¹⁴N ratio of the external nitrate pool.

(ii) Sulfide oxidation: the ratio between observed sulfide reoxidation rates and specific activity of Thioploca. Sulfate reduction rates measured in sediments at station 7 at the time of sampling were approximately 30 mmol m² day¹ (26 to 37 mmol m² day¹ [34]). If all sulfide produced from this reduction was subsequently oxidized by the Thioploca mats, then the mats should be able to oxidize sulfide at the same rate, which is equal to 20.8 µmol m² min¹. Schulz et al. (32) estimated the wet biomass of trichomes without sheaths to be 50 to 120 g m². Assuming an average of 85 g (wet weight) m² and knowing that 90% of the biovolume is taken up by the central vacuole, the active cytoplasm weighs approximately 8.5 g (wet weight) m². This active cytoplasm is then responsible for the sulfide oxidation rate as stated above, which would give a specific rate of 20.8/8.5 = 2.4 µmol min¹ g of wet weight¹. Assuming that 24% of the wet weight is dry weight and that 50% of the dry weight is protein, the sulfide oxidation rate in vivo should be 2.4 × (100/24) × 2 = 20.4 nmol min¹ mg of protein¹. In analogy, Ferdelman et al. (6) found an in vivo sulfate reduction rate of approximately 17.5 mmol m² day¹ for station 7. This reduction rate corresponds to a sulfide oxidation rate by Thioploca of 12 µmol m² min¹. Making the same assumptions as above, this oxidation rate is equal to 11.8 nmol min¹ mg of protein¹.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Cultivation and survival. Thioploca bundles, consisting of 10 to 20 trichomes in a sheath, were collected from the top 2 cm of the sediment with forceps and cleaned by several transfers through medium. In developing the method, there were three parameters to be considered. Firstly, motility of Thioploca trichomes under a microscope was a measure of viability (19, 31). Secondly, it was observed during the initial experiments that high nitrite concentrations (50 to 100 µM), in addition to high nitrate concentrations, were obtained within 1 h of anoxic incubation of the bundles, suggesting lysis of the cells. Thirdly, in previous studies of Thioploca and large Beggiatoa species, it was observed that these organisms are highly sensitive to oxygen (16, 22) and that catalase is required in the growth medium of Beggiatoa species (1). These considerations led to several improvements of the final cleaning procedure. To avoid contact with oxygen, all steps of the method (collection, washing, and incubation) were performed under a dinitrogen atmosphere. Sparging of the medium during incubation for anoxic conditions was avoided, because this mechanically affected the trichomes. To keep the growth medium anoxic, thioglycolate was added as a reducing agent and catalase was also included in the synthetic medium. The endogenous ammonium production rate did not increase after thioglycolate was included in the medium, suggesting that this compound was not used as a carbon source. Survival experiments, as monitored under the microscope (31), showed that the medium highly improved survival and that trichomes did not show a decrease in motility over 2 days of incubation. From similar survival experiments, it was further concluded that Thioploca cells could get damaged when transferred through a liquid-gas interface. To avoid this damage, washing was performed twice by draining off approximately 80% of the medium, such that the trichomes were still in the liquid, and then fresh medium was added.

N metabolism. Inside Thioploca cells, intravacuolar nitrate concentrations measured up to 500 mM, with an average of 160 ± 150 mM (n = 27). Thioploca trichome bundles, incubated in medium without addition of NO₃, produced NO₂ and NH₄⁺. The average NH₄⁺ production (with internal sulfur available but no added external electron donor) of eight independent measurements was 1.0 ± 0.3 nmol min¹ mg of protein¹, whereas NH₄⁺ production by the controls (disrupted trichome bundles or NO₃-supplemented sediment) was 0.07 ± 0.03 nmol min¹ mg of protein¹ and 0.03 ± 0.01 nmol min¹ mg of protein¹, respectively. Nitrite production was negligible (<0.1 nmol min¹ mg of protein¹) in most experiments but was sometimes observed at a maximum production rate of 1 nmol min¹ mg of protein¹.

¹⁵N-labeling experiments. To determine whether Thioploca reduces NO₃ to NH₄⁺ or to N₂, experiments were performed by using ¹⁵NO₃. After addition of ¹⁵NO₃, total NO₂ and total NH₄⁺, as well as ¹⁵NO₃/¹⁵NO₂ and ¹⁵NH₄⁺, were monitored over time (the ¹⁵N-labeling method does not differentiate between labeled NO₃ and NO₂; see Materials and Methods). At the end of the experiment, total ¹⁵N₂ and the ratio between unlabeled and (singly or doubly) labeled N₂ were determined. Figure 1A shows that 95% of the externally available NO₃ or NO₂ originated from the supplied ¹⁵NO₃. During the course of the experiment, the specific labeling of the extracellular nitrate pool remained 95%, indicating that the trichomes were not damaged and did not release ¹⁴NO₃ (see calculations in Materials and Methods). Figure 1A shows that nearly all of the nitrate was taken up linearly during the course of the experiment in approximately 3.5 h. As calculated in Materials and Methods, if all of the ¹⁵NO₃ were taken up by Thioploca trichomes, the label would be diluted inside the vacuoles to 9.9% ± 2.85% (see calculations in Materials and Methods). The NH₄⁺ produced during the experiment (1.8 ± 0.4 nmol min¹ mg of protein¹) was 44% (at 85 min) and 48% (at 145 and 215 min) labeled (Fig. 1B). This difference in specific labeling between the externally available nitrate pool and the NH₄⁺ produced indicates that the internal nitrate of Thioploca trichomes contributes substantially to the total NH₄⁺ production. However, the amount of label is not diluted as much as would be expected if all the labeled nitrate were first taken up in the vacuole and subsequently reduced. Therefore, it seems that in the cytoplasm the [¹⁵N]nitrate is readily reduced before it reaches the vacuole. Figure 1C shows that at the end of the experiment N₂ had also been produced, but the amount was only 15% (nanomoles of nitrogen per nanomole of nitrogen) of the total amount of nitrogen compounds produced. The specific labeling of the N₂ was substantially higher than that of NH₄⁺, suggesting that epibionts might be responsible for this production, although the amount of unlabeled N₂ is a minimum estimate (see Materials and Methods). This implies that, although N₂ appeared to not be a major product of the washed Thioploca sample, the present data cannot completely rule out that Thioploca can reduce nitrate to N₂, i.e., denitrify, in addition to the observed full reduction of nitrate to ammonium.

View larger version (30K): [in this window] [in a new window]   FIG. 1.   Distribution of label during ammonium and dinitrogen production by Thioploca trichome bundles after incubation in medium with [15N]nitrate under a helium headspace. (A) 14N- and 15N-labeled nitrate and nitrite in the growth medium; (B) 14N- and 15N-labeled ammonium in the growth medium; (C) total amount of 14N and 15N derived from dinitrogen species in the headspace (14,14N2, 14,15N2, 15,14N2, 15,15N2). Symbols: , total ammonium measured colorimetrically; , nitrite measured colorimetrically.

Sulfur metabolism. To measure sulfide oxidation rates, Thioploca trichomes were incubated in medium. After addition of approximately 50 µM sulfide to the vials, sulfide, nitrite, and ammonium concentrations were observed over time. An ammonium production rate by the Thioploca suspension of approximately 1.9 nmol min¹ mg of protein¹ was observed, whereas the controls (sediment samples and samples treated in a Potter-Elvehjem homogenizer) showed activities of 0.02 and 0.04 nmol of NH₄⁺ min¹ mg of protein¹, respectively (Table 1). The sulfide consumption rate decreased with decreasing sulfide concentrations, but the average maximum rate was approximately 4.2 nmol min¹ mg of protein¹, while the controls showed a 10- to 200-fold lower consumption rate. Since the control, which contained sediment, represented an overestimate of the amount of sediment attached to the trichomes, the contribution of the sediment to the total activity in live Thioploca experiments was negligible. Trichomes, which had been collected and incubated 2 days before the experiment was performed, showed continued motility under the microscope, an ammonium production rate of 3.2 nmol min¹ mg of protein¹, and a sulfide consumption rate of 5.5 nmol min¹ mg of protein¹, after addition of sulfide to the incubation medium. The cells appear to reduce their metabolic activity when no external substrate is present (NH₄⁺ production rate is approximately 1 nmol min¹ mg of protein¹) but are able to respond quickly when substrate is encountered again (NH₄⁺ production increases to 3.2 nmol min¹ mg of protein¹). An internal nitrate concentration of 160 mM would be sufficient for approximately 200 h (given our estimate that 90% of the cell is vacuole and that 1 mg of vacuolar liquid is equal to 1 µl), with an NH₄⁺ production rate of ± 1 nmol min¹ mg of protein¹. This experiment indicates that trichomes are still active and motile after 2 days and that the internal NO₃ is sufficient for at least 2 days of normal metabolism without external supply of fresh substrate. In an experiment where two different concentrations of sulfide were added to Thioploca suspensions (Fig. 2), a small accumulation of thiosulfate, which was higher when the sulfide concentration was higher, was observed. This thiosulfate accumulation suggested that this compound may be a by-product or an intermediate in sulfide oxidation, and therefore, Thioploca might be able to oxidize thiosulfate to sulfate. To investigate this possibility, approximately 100 µM thiosulfate was added to Thioploca trichome bundles incubated in sulfate-free medium with MgCl₂ instead of MgSO₄ and without thioglycolate and catalase, to avoid interference with analytical measurements. In these experiments (results not shown) only a slight thiosulfate consumption was observed.

View larger version (19K): [in this window] [in a new window]   FIG. 2.   Thiosulfate production by Thioploca trichome bundles incubated anoxically in medium with two different initial sulfide concentrations. Open symbols, 100 µM initial sulfide concentration; closed symbols, 400 µM initial sulfide concentration. Symbols: triangle, ammonium; circle, thiosulfate; diamond, sulfide.

3

2

1

1

2

1

1

1

1

Carbon metabolism. In order to gain some insight into the carbon source used by Thioploca for its cell material, experiments were performed with radioactively labeled acetate and bicarbonate additions (Table 2) in combination with microautoradiography.

(i) NaH¹⁴CO₃. Addition of NaH¹⁴CO₃ to a Thioploca suspension resulted in a linear incorporation rate of 0.4 to 0.8 nmol min¹ mg of protein¹ and an NH₄⁺ production rate of approximately 1.3 to 1.7 nmol min¹ mg of protein¹. Addition of sulfide (ca. 70 µM) did not have a significant effect on the incorporation rate. The presence of low concentrations of oxygen (ca. 10% air saturation) also did not have a significant effect on the rates of NaH¹⁴CO₃ fixation. Control experiments with disrupted trichome bundles showed a ¹⁴C fixation rate of only 0.01 nmol min¹ mg of protein¹, which is 1 to 3% of the rates observed in intact Thioploca suspensions. Intact bundles obtained from these experiments were used for microautoradiography. Uptake of ¹⁴C appeared to be largely dominated by Thioploca trichome bundles, since there was no significant uptake of label in epibiontic microbial cells associated with the sheaths. Microautoradiography and control experiments with disrupted trichome bundles both show that the measured uptake rate of ¹⁴C by the Thioploca suspension primarily represents the activity of the trichome bundles and not of epibionts. Differences in intensity of labeling among trichomes were observed, but no differences were observed that could be due to possible differences in the physiology of the two major species in the sample (T. araucae and T. chileae). Among individual trichomes, labeling was homogeneously distributed along their entire length, and labeling was concentrated along the transverse walls (Fig. 3), suggesting the presence of a vacuole.

View larger version (29K): [in this window] [in a new window]   FIG. 3.   High-magnification microautoradiogram of a single Thioploca trichome incubated with NaH14CO3.

(ii) [¹⁴C]- and [³H]acetate. After addition of [¹⁴C]acetate to Thioploca trichome bundles, an acetate uptake rate of approximately 0.4 nmol min¹ mg of protein¹ and an NH₄⁺ production rate of 4 nmol min¹ mg of protein¹ were observed. Unaccounted loss of label was less than 10%. ¹⁴CO₂ production was not significant (less than 2% of acetate incorporation), indicating that under these conditions (i.e., in the presence of internal sulfur) acetate was not used as a significant energy source. Control experiments with disrupted trichome bundles showed an incorporation rate of less than 0.01 nmol min¹ mg of protein¹, which was less than 2% of the activity of intact Thioploca trichome bundles.

View larger version (65K): [in this window] [in a new window]   FIG. 4.   Low-magnification microautoradiogram of a Thioploca sheath with trichomes incubated with [3H]acetate, showing 3H uptake by trichomes and associated bacteria. The heavily labeled trichomes are out of focus to show uptake by bacteria situated on the sheath. This image has been selected for its high concentration of epibionts and is not representative of the overall results from the microautoradiography.

View larger version (114K): [in this window] [in a new window]   FIG. 5.   High-magnification microautoradiogram of Thioploca trichome incubated with [3H]acetate.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Filamentous sulfur bacteria of the genus Thioploca occur along the continental shelf off the coast of Chile and Peru. High sulfate reduction rates in Thioploca mats have been reported (6). Thioploca species are able to store internally high concentrations of sulfur globules and nitrate. It is assumed that these vacuolated Thioploca species use their internally stored nitrate as a terminal electron acceptor for sulfide and sulfur oxidation (7). The product of nitrate reduction, however, was still unknown. Also, Thioploca trichome bundles have been shown to take up both CO₂ and acetate, but quantitative data were lacking (19). Therefore, this study was undertaken to investigate carbon, nitrogen, and sulfur metabolism in Thioploca species.

A method was developed to collect and clean individual sheaths with bundles of trichome bundles from the top 2 cm of the sediment. After being collected and washed under an N₂ atmosphere, trichomes were still motile and could be used for physiological experiments. In the early stages of method development, high cellular or extracellular nitrite and nitrate concentrations were observed, possibly as a result of lysis of the cells. However, after adjustments (anoxic conditions, medium supplemented with thioglycolate and catalase, low temperature, and avoiding transfer through the gas-liquid interface) these nitrite and nitrate accumulations were no longer observed. Thioploca trichome bundles incubated for 2 days still showed activity comparable to activities measured immediately after incubation (Table 1), indicating that trichome bundles were able to survive and remain physiologically intact in the synthetic medium.

Nitrogen metabolism. During experiments performed with intact Thioploca trichome bundles, without addition of external substrate, an ammonium production rate of approximately 1 nmol min¹ mg of protein¹ was observed. Since in these experiments the only available substrates were internally stored sulfur and nitrate in Thioploca trichome bundles, it is highly unlikely that epibiontic bacteria were responsible for this NH₄⁺ production. Experiments using [¹⁵N]nitrate resulted in uptake of all the labeled nitrate in approximately 3.5 h. The specific label of the external NO₃ pool remained 95% and was, therefore, not diluted during the course of the experiment, indicating that trichome bundles were not damaged and leaking NO₃. Analysis showed an increase in NH₄⁺ production (1.8 ± 0.4 nmol min¹ mg of protein¹) immediately after addition of labeled NO₃. The specific label of NH₄⁺ produced was maximally 48%. This indicates that NH₄⁺ is produced from a different NO₃ pool than the external pool, since the external pool was more heavily labeled. The only other source of NO₃ is the internal NO₃ of Thioploca trichome bundles, which is not available to epibiontic bacteria. This indicates that Thioploca species reduce NO₃ to NH₄⁺. Another argument is that there was no electron donor for NO₃ reduction available in these experiments, except the internally stored sulfur.

Sulfur metabolism. After addition of sulfide to Thioploca trichome bundles in a particular experiment, a sulfide oxidation rate of approximately 4.2 nmol min¹ mg of protein¹ was observed in the absence of external nitrate. The NH₄⁺ production was 1.9 nmol min¹ mg of protein¹, resulting in a ratio of 2.2 between sulfide oxidized and NH₄⁺ produced. If the sulfide were oxidized to elemental sulfur and NO₃ were reduced to NH₄⁺, then the expected ratio of sulfide to ammonium would be 4. If sulfide were oxidized to sulfate then a ratio of 1 would be expected. The observed ratio suggests that the sulfide is oxidized to both sulfur and sulfate, since there was no significant accumulation of other (intermediate) sulfur species (i.e., sulfite and thiosulfate). Analogous to observations in marine Beggiatoa (23), it is likely that the immediate product of sulfide oxidation is elemental sulfur, which is stored in Thioploca as globules. The elemental sulfur is then oxidized to SO₄² in a second, independent step, as suggested by Fossing et al. (7). In experiments without addition of sulfide, sulfate production was observed at a rate of 2 to 3 nmol min¹ mg of protein¹, which must have originated from internal elemental sulfur. In the presence of sulfide, the SO₄² production rate did not increase significantly, suggesting that sulfide is oxidized to sulfur and that further oxidation of sulfur to SO₄² occurs independently of the presence of sulfide. In these two experiments, the ratio of SO₄² to NH₄⁺ produced was approximately 1.5 in the absence and approximately 1.7 in the presence of sulfide. If NO₃ is reduced to NH₄⁺ and sulfur is oxidized to SO₄², then a ratio of 1.3 is expected. This is in agreement with the observed ratio in the absence of sulfide, indicating, again, that Thioploca trichome bundles reduce most NO₃ to NH₄⁺ under the conditions tested. It was also observed that addition of different concentrations of sulfide (100 µM and 400 µM) did not result in a significant increase in NH₄⁺ production (Fig. 2). This reconfirms that oxidation of sulfide, and subsequently sulfur, occurs independently. The observed ratios indicate that net sulfur accumulation will occur when external sulfide is present. Addition of sulfide led to a small accumulation of thiosulfate (S₂O₃²) in the medium, suggesting that S₂O₃² may be an intermediate in sulfur oxidation to sulfate. However, addition of S₂O₃² to trichome bundles showed only a very low consumption of S₂O₃². Starvation of the trichome bundles for 45 h in the presence of NO₃ did not enhance this consumption rate. Accumulation of S₂O₃² during starvation of disrupted trichome bundles indicates that Thioploca cells may not be responsible for the observed accumulation in previous experiments. At present, due to variations in the measurements, it cannot be determined whether or not Thioploca produces S₂O₃² as an intermediate.

2

1

1

1

2

2

1

1

1

1

1

1

1

2

Carbon metabolism. Addition of [¹⁴C]bicarbonate resulted in an incorporation rate of 0.4 to 0.8 nmol min¹ mg of protein¹. The presence of sulfide did not increase the incorporation rate significantly. The measured SO₄² production rate (generated from internal sulfur) was 2 to 3 nmol min¹ mg of protein¹, which is equivalent to an average of 1.3 nmol min¹ mg of dry weight¹, assuming that 50% of dry weight is protein. From these data we can predict the CO₂ fixation rate, assuming that 12.5% of the electrons produced go to CO₂ fixation (assuming a yield of 8 g (dry weight) · mol of sulfide¹ [23, 38]). The oxidation of sulfur to SO₄² produces six electron equivalents. Given the fact that CO₂ reduction to biomass (dry weight) requires four electron equivalents, the predicted rate of CO₂ fixation would be 0.125 × (6/4) × 1.3 = 0.24 nmol min¹ mg¹ (dry weight). This rate is equivalent to 0.49 nmol min¹ mg of protein¹ (assuming that 50% of the dry weight is protein), which is the rate observed, suggesting that Thioploca species can grow autotrophically by using internally stored sulfur and NO₃ for energy generation. Results obtained with microautoradiography confirm earlier qualitative experiments by Maier and Gallardo (19) and indicate that the CO₂ fixation measured can be attributed to Thioploca trichome bundles and not to epibiontic bacteria. Ferdelman et al. (6) measured a CO₂ fixation rate in cleaned Thioploca suspensions of 2,400 ± 700 nmol day¹ g¹ (wet weight). Assuming that the wet weight of trichomes is 10% of the wet weight of sheaths and trichomes (32), that 10% of the wet weight of trichomes is cytoplasm, that 24% of the wet weight of the cytoplasm is dry weight, and that 50% of the dry weight is protein (see calculations in Materials and Methods), then the fixation rate was estimated to be 1.4 ± 0.4 nmol min¹ mg of protein¹. This rate is approximately three times as high as the rate observed in our study.

1

1

3

1

1

1

1

2

1

2