INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Sea grasses colonize shallow coastal marine environments (2, 26, 61) and thus are rooted in reduced, anoxic sediments where high rates of sulfate reduction can yield high concentrations of sulfide (5). Sediments colonized by the sea grass Halodule wrightii contain high numbers of readily culturable acetogens and acetate-utilizing sulfate reducers, and these numbers are significantly higher than in unvegetated sediments (38). Colonization of the sea grass rhizosphere by bacteria might give those organisms ready access to plant-derived substrates that could supply the energy needed for nitrogen fixation, thus yielding a beneficial plant-microbe interaction (11, 12). Although most rhizobacteria likely colonize root tips or outermost root cell layers (9), microautoradiographs of sea grass root thin sections hybridized with ³³P-labeled probes revealed the presence of acetogens, clostridia, and sulfate-reducing bacteria in both the rhizoplane and deep cortex cells of sea grass roots (38).

The O₂ produced by leaf photosynthesis is transported to the roots and generates transient O₂ gradients around the roots (1, 33). Thus, anaerobic endorhizobacteria likely experience periods of elevated O₂ tension, suggesting that such bacteria must cope with O₂. Indeed, exposure of sea grass roots to O₂ has minimal affect on root-surface-associated, sulfate-reducing activity (5). This observation is consistent with other findings demonstrating that some sulfate-reducing bacteria are aerotolerant (10, 19, 28).

Most acetogens have been isolated from habitats with stable anoxic conditions (e.g., sediments or sewage sludge) (21, 50). However, acetogens colonize the leaf litter and mineral soil of oxic forest soils (34, 37, 47), tolerate periods of oxygenation in soils (60), and are active in termite guts that have steep O₂ gradients (7, 56). The objectives of this study were to (i) isolate an acetogen from sea grass roots, (ii) determine its survivability under oxic conditions, and (iii) investigate its physiological response and protective mechanisms to elevated O₂ tensions.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Collection of sea grass roots. Sea grasses (H. wrightii) were sampled in clear, plastic cores from a depth of about 1 m below the surface of the overlying brackish water near Big Sabine Point in Santa Rosa Sound, located in northwestern Florida. The cores were released into a sterile glass dish, and the roots were carefully separated from the sediment. The pH of the sediment was 6.7, and the water temperature was 33°C. Healthy roots, white and free of lesions, were excised with a sterile razor blade and washed twice in sterile phosphate-buffered saline (120 mM sodium phosphate and 0.85% NaCl, pH 7.2).

Medium composition and growth conditions. The anoxic, carbonate-buffered, undefined (U) medium contained yeast extract, vitamins, and trace metals but did not contain reducing agents (16). U_salt medium was U medium supplemented with NaCl (20 g liter¹) and MgCl₂ (2 g liter¹). The defined medium was U medium without yeast extract. Media were dispensed under CO₂ into 27-ml culture tubes (7 ml of medium per tube) or 1-liter infusion bottles (500 ml of medium per bottle, used for preparation of cell extracts), which were then sealed and autoclaved; the pH approximated 6.7. Tryptic soy broth (TSB) medium (28 g of Bacto TSB without glucose [Difco Laboratories, Detroit, Mich.] per liter) was made anoxic by boiling and cooling under 100% argon. The pH was adjusted with H₃PO₄ or NaOH to the indicated pH. Anoxic stock solutions of substrates (prepared under argon) were filter sterilized and were added by syringe injection using O₂-free techniques. Unless otherwise indicated, culture tubes and bottles were incubated in a horizontal, static position, cultivation was in anoxic U medium, and the temperature of incubation was 30°C. Escherichia coli K12 (DSM 423) was cultivated in peptone-beef extract medium (5 g liter¹) (Difco Laboratories) at pH 7.0.

Alternative electron acceptors, reduction of C₂H₂, and toxic effects of oxygen. The dissimilation of nitrate or sulfate was evaluated with TSB medium supplemented with 10 mM glucose and 5 mM NaNO₃ or 5 mM Na₂SO₄, respectively. The reduction of Fe(III) was determined by assessing the growth-dependent production of white Fe(II) precipitates in medium formulated for Fe(III)-reducing bacteria containing U growth factors (8). Nitrogenase activity was determined with a modification of the C₂H₂ reduction method (30, 41, 58). Sterile O₂ was added to the headspace of culture tubes to the concentrations (vol/vol) indicated.

Enrichment cultures. Washed roots (5 g) were brought into an O₂-free chamber (Mecaplex, Grenchen, Switzerland) (100% N₂ gas phase; room temperature) and were homogenized with a grinder in 45-ml basal salt solution (16). The root suspension was serially diluted and transferred to culture tubes containing U_salt medium and H₂-CO₂ (80:20 [vol/vol]) in the headspace. Stable acetogenic enrichment cultures were obtained from the highest growth-positive dilution series and were subsequently streaked onto solidified U_salt medium. Isolated colonies were picked with a sterile needle and were transferred to liquid U_salt medium; cultures were subsequently restreaked onto solidified U_salt medium, and isolated colonies were again transferred to liquid U_salt medium. This procedure was repeated two additional times, and isolates were examined microscopically for purity.

Electron microscopy. Cells were cultivated in U_salt medium or on solidified U_salt medium (1% Gelrite; Carl Roth GmbH, Karlsruhe, Germany); both media were supplemented with 10 mM glucose. For negative staining, cells were fixed for 30 min in liquid medium by adding glutaraldehyde to a final concentration of 2% (vol/vol). Cells were harvested by gentle centrifugation (1,000 × g; 15 min), adsorbed to carbon film (59), and stained with aqueous uranyl acetate solution (2% [wt/vol]). For preparation of thin sections, cells were fixed in glutaraldehyde-OsO₄ (39, 57).

Preparation of cell extracts and enzyme assays. Cell extracts were prepared under anoxic conditions (35, 42). Membranous and cytoplasmic fractions were prepared under oxic conditions (27). Oxidoreductase activities were assayed spectrophotometrically at 20°C in Tris-HCl (50 mM, pH 7.5) buffer containing 0.5 mM benzyl viologen (16). Assay tubes contained 20% H₂ (vol/vol, with N₂ in gas phase) for hydrogenase, 20% CO (vol/vol, with N₂ in gas phase) for carbon monoxide dehydrogenase or 4 mM sodium formate for formate dehydrogenase (100% N₂ in gas phase). These enzyme activities are expressed in micromoles of substrate oxidized minute¹.

1

1

1

1

Redox difference spectra. Membranous and cytoplasmic fractions were reduced with sodium dithionite, and redox difference spectra were obtained with a Uvikon 930 (Kontron Instruments, Milan, Italy) double-beam recording spectrophotometer (34).

G+C content. Cells were treated with penicillin G (250 µg ml¹) 3 h before harvesting. DNA extraction included treatments with lysozyme/proteinase K and RNase/proteinase K (43). The G+C content was determined by high-performance liquid chromatography using nonmethylized lambda DNA for calibration (45, 54).

Phylogenetic analysis. A total of 1,317 bases of the 16S rRNA gene of RD-1 were sequenced. DNA extraction, PCR-mediated amplification of the 16S rRNA gene, and purification of the PCR products were performed according to published protocols (20, 49). Cells were lysed by boiling at approximately 100°C. Purified PCR products were sequenced by using an ABI PRISM Ready Reaction Dye Terminator kit (Applied Biosystems, Foster City, Calif.). Sequence reaction mixtures were electrophoresed with an Applied Biosystems model 373A DNA sequencer. Alignment of the sequence data and sequence similarity calculations were performed using the tools of the ARB software package (http://www.mikro.biologie.tu-muenchen.de).

DNA-DNA hybridzation. DNA-DNA hybridization was determined by spectrophotometric reassociation kinetics according to published protocols (13, 17, 24, 31).

Additional analytical methods. Growth was measured as optical density at 660 nm; the optical-path width (inner diameter of culture tubes) was 1.6 cm. Uninoculated medium served as a reference. Protein in cell extracts was determined colorimetrically (6). Substrates and products were determined by high-performance liquid chromatography and gas chromatography (16, 36, 44). Nitrate was measured colorimetrically (14). Sulfate was analyzed by ion chromatography (36). Results are representative of replicate experiments.

Accession numbers. RD-1 has been deposited at the German Collection of Microorganisms and Cell Cultures (Braunschweig, Germany) under accession number DSM 13865. The 16S rRNA gene sequence of RD-1 has been deposited at the EMBL Nucleotide Sequence Database (Cambridge, United Kingdom) under accession number AJ291746.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Acetogenic enrichment cultures and isolates. When root homogenates were serially diluted in U_salt medium containing H₂-CO₂, stable enrichment cultures that produced 1 mol of acetate per 5 mol of H₂ consumed were obtained from the two highest growth-positive dilutions, 10⁴ and 10⁵. Two gram-positive, rod-shaped isolates, RD-1 (obtained from the 10⁵ growth-positive dilution) and RD-3 (obtained from the 10⁴ growth-positive dilution), converted H₂-CO₂ to acetate in stoichiometries indicative of an acetyl coenzyme A (acetyl-CoA) pathway-dependent metabolism: 4 H₂ + 2 CO₂  CH₃COOH + 2 H₂O (21). Colonies of both isolates were shiny, convex, white to beige, and 2 to 3 mm in diameter. Initial screening of growth substrates indicated that RD-1 and RD-3 were the same organism, and RD-1 was selected for further characterization.

Morphology and ultrastructure of RD-1. Cells were 4 by 0.8 to 6 by 0.8 µm (Fig. 1A) and were motile in wet mounts. Thin sections revealed a multilayered cell wall but no outer membrane (Fig. 1A). Cells formed terminal spores (Fig. 1B); free spores were rarely observed in cultures under the light microscope. Up to seven cells of RD-1 were often tethered to one another by connecting filaments (Fig. 1C). The connecting filament consisted of a core and a surrounding sheath (Fig. 1D). A morphological continuity was apparent between the surface of the cell and the sheath (Fig. 1E). However, the sheath was not always present, indicating that it could be dissociated from the core. The negatively stained core usually revealed a darkened center (Fig. 1D), which might have been due to a high affinity of the inner portion of the core for uranyl acetate.

View larger version (132K): [in this window] [in a new window]   FIG. 1.   Transmission electron micrographs of RD-1. (A) Thin section of a dividing cell. (B) Thin section showing formation of a terminal spore. (C to E) Negatively stained specimens revealing the ultrastructural features of connecting filaments. Abbreviations: SL, surface layer; CW, cell wall; CM, cytoplasmic membrane; SP, prespore; CF, connecting filament; Sh, outer sheath; C, core. Bars are in micrometers.

Growth properties of RD-1. Growth was observed after cultures were heated for 15 min at 80°C, confirming the occurrence of spores. Growth could not be maintained in defined medium, indicating that RD-1 required some undefined factors for growth. In TSB medium, growth occurred at pH 6.1 to 8.2; growth did not occur at pH 5.6 and 8.5. The optimal pH was 7.0 to 7.5. RD-1 grew at 25 to 42°C in U_salt medium (pH 7.0) supplemented with fructose; no growth occurred at 20 and 45°C. At pH 7.0, growth rates and cell yields were optimal at 37 to 40°C; under these conditions, the doubling time was 2.9 h.

Substrate range and fermentation stoichiometries of RD-1. The following substrates were utilized in U_salt medium: fructose, glucose, maltose, xylose, pyruvate, lactate, formate, and H₂-CO₂. RD-1 produced acetate from H₂-CO₂, formate, pyruvate, and lactate in stoichiometries indicative of acetogenesis (22) (Table 1 and data not shown). Fructose, glucose, maltose, and xylose were fermented to acetate and ethanol and small amounts of H₂ (Table 1 and data not shown). The presence of elevated salt concentrations affected the ratio of acetate produced to fructose consumed (Table 2). Smaller amounts of acetate but greater amounts of ethanol and H₂ were produced when RD-1 was grown on fructose in U_salt medium than when RD-1 was grown on fructose in U medium. Growth on yeast extract (i.e., U medium without supplemental substrates) yielded acetate and isovalerate. Ethylene glycol was fermented to acetate and ethanol, and propylene glycol was fermented to propionate and propanol and small amounts of acetate. Neither growth nor utilization of substrate was observed with galactose, sucrose, melitose, cellobiose, mannose, lactose, vanillate, syringate, ferulate, methanol, ehanol, propanol, 2,3-butandiol, citrate, succinate, fumarate, butyrate, oxalate, acetate, or CO. RD-1 did not dissimilate nitrate, sulfate, or Fe(III) and did not produce methane. In medium lacking inorganic nitrogen compounds, RD-1 did not reduce C₂H₂, indicating that RD-1 did not produce nitrogenase.

Effect of O₂ on growth of RD-1. Glucose-dependent growth rates and cell yields were not significantly affected by up to 6% (vol/vol) O₂ in the headspace of static, horizontally incubated tubes that were periodically shaken prior to measuring optical densities (Fig. 2A). When culture tubes were continuously shaken (60 rpm), RD-1 grew in the presence of up to 4% (vol/vol) O₂ in the headspace (Fig. 2B).

View larger version (25K): [in this window] [in a new window]   FIG. 2.   Effect of O2 on the glucose-dependent growth of RD-1. Culture tubes were either incubated horizontally and shaken prior to each measurement (A) or continuously shaken (B). Data are the means of three replicates (± standard deviations). The initial concentrations of O2 (vol/vol) in the headspace were , 0%; , 4%; , 5%; , 6%; , 7%; and , 9% (A) or , 0%; , 2%; , 3%; and , 6% (B).

Effect of O₂ on physiological capabilities of RD-1. When RD-1 was cultivated on glucose with 0.3, 2.1, 2.8, 3.8, 4.9, or 5.8% (vol/vol) O₂ in the headspace of static, horizontally incubated tubes, O₂ decreased to 0, 1.5, 2.0, 3.1, 3.9, or 4.7% (vol/vol), respectively, by the end of incubation (2 days). O₂ was not consumed in uninoculated controls. RD-1 consumed O₂ concomitantly with exponential growth and the consumption of glucose (Fig. 3).

View larger version (20K): [in this window] [in a new window]   FIG. 3.   Effect of O2 on the glucose-derived product and growth profiles of RD-1. Data are the means of three replicates (± standard deviations). (A) , O2 in cultures containing supplemental O2; , O2 in cultures without supplemental O2; , optical density in cultures containing supplemental O2; , optical density in cultures without supplemental O2; , H2 in cultures containing supplemental O2; , H2 in cultures without supplemental O2. (B) , glucose in cultures containing supplemental O2; , glucose in cultures without supplemental O2; , acetate in cultures containing supplemental O2; , acetate in cultures without supplemental O2. Acetate values were corrected with control values from cultures lacking glucose. The small amounts of ethanol formed from 2 mM glucose could not be accurately quantitated.

1

View larger version (25K): [in this window] [in a new window]   FIG. 4.   Effect of O2 on the consumption of H2 by RD-1. Culture tubes contained an H2-CO2 gas phase (80:20%, vol/vol). Symbols: , , and , O2 in cultures with supplemental O2; , O2 in cultures without supplemental O2; , , and , H2 in cultures containing 1, 0.5, and 0.25% (vol/vol) O2, respectively; and , H2 in cultures without supplemental O2.

Oxidative stress enzyme activities of RD-1. NADH oxidase, superoxide dismutase, peroxidase, and catalase activities in the cytoplasmic fraction of RD-1 approximated 4.3, 0.4, 0.1, and 0 U mg¹ of protein, respectively. These enzyme activities were not detected in the membrane fraction. NADH oxidase, superoxide dismutase, peroxidase, and catalase activities in the cytoplasmic fraction of E. coli K12 approximated 3.3, 0.5, 0.5, and 230 U mg¹ of protein, respectively.

Oxidoreductase activities and redox difference spectra of membranes of RD-1. Carbon monoxide dehydrogenase, hydrogenase, and formate dehydrogenase activites in cell extracts obtained from fructose-cultivated cells of RD-1 approximated 0.6, 0.9, and 0.2 U mg¹ of protein, respectively. No absorption maxima indicative of cytochromes were detected in the membranous or cytoplasmic fractions of RD-1 (data not shown).

Phylogenetic analysis, G+C content, DNA-DNA hybridization, and protein profiles. The 16S rRNA gene sequence of RD-1 was most closely related to that of bacteria that group in cluster XI of the genus Clostridium (15). The highest sequence similarity value (99.7%) was to C. glycolicum DSM 1288^T. The DNA base composition of RD-1 was 31.6 (±0.8) (n = 3) mol%. The G+C mol% of the DNA of C. glycolicum DSM 1288^T is 29 (32). DNA-DNA hybridization experiments revealed a 91.4% genome sequence similarity between RD-1 and C. glycolicum DSM 1288^T. Cells of RD-1 and of C. glycolicum DSM 1288^T that had been cultivated on fructose yielded nearly identical protein profiles when cell extracts were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis (data not shown).

Metabolic capacities of C. glycolicum DSM 1288^T In U medium, C. glycolicum DSM 1288^T fermented fructose (10 mM) to acetate (21 mM) and ethanol (2.2 mM). In U_salt medium, fructose (10 mM) was fermented to acetate (17.8 mM), ethanol (2.5 mM), and trace levels of H₂. In contrast to RD-1, C. glycolicum DSM 1288^T did not catalyze the H₂- or formate-dependent formation of acetate. Nonetheless, carbon monoxide dehydrogenase and hydrogenase activities were detected in cell extracts of fructose-cultivated cells of C. glycolicum DSM 1288^T (activities approximated 1.1 and 0.6 U mg¹ of protein, respectively). The temperature optimum of C. glycolicum DSM 1288^T was 30°C.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

RD-1 was isolated as an anaerobe with acetogenic capabilities yet was also able to grow in the presence of 4 and 6% (vol/vol) O₂ in the headspace of continuously and periodically shaken culture tubes, respectively. The capacity of RD-1 to consume small amounts of O₂ without an apparent lag phase suggests that this metabolic capacity was constitutive. Many so-called obligately anaerobic bacteria and archaea exhibit some degree of aerotolerance (48). NADH oxidase is used by the aerotolerant anaerobe Brachyspira hyodysenteriae for the consumption of O₂ (51). Facultative bacteria and certain obligate anaerobes contain superoxide dismutase or NADH oxidase (48, 51). Superoxide dismutase, peroxidase, and NADH oxidase activities were detected in the cytoplasmic fraction of RD-1, indicating that RD-1 can disproportionate superoxide and reduce hydrogen peroxide or O₂. The oxidative stress enzyme catalase, present in certain methanogens (40), was not detected in RD-1. Desulfoferrodoxin and rubrerythrin function together as an oxidative stress defense mechanism in the sulfate reducer Desulfovibrio vulgaris (42), and genes for similar proteins have been identified in the acetogen Moorella thermoacetica (A. X. Das and L. G. Ljungdahl, Abstr. 100th Gen. Meet. Am. Soc. Microbiol., abstr. H-117, p. 374, 2000). In comparison to the ability of RD-1 to tolerate relatively high amounts of O₂, the acetogens Acetobacterium woodii, M. thermoacetica, Clostridium magnum, and Sporomusa silvacetica can only tolerate 0.5, 1, 2, and 2% (vol/vol) O₂, respectively, in the headspace of culture tubes when cultivated on glucose (A. Karnholz, K. Küsel, and H. L. Drake, Abstr. 100th Gen. Meet. Am. Soc. Microbiol., abstr. I-91, p. 401, 2000; unpublished data). Methanogenic (40) and acetogenic (H. Boga and A. Brune, Abstr. Ann. Meet. Verein. Allgem. Angewand. Mikrobiol. BioSpectrum, abstr. 15.P.11.33, p. 143, 2000) termite gut isolates can grow and consume O₂ in agar-O₂-gradient tubes. Thus, certain acetogens, like other so-called obligate or strict anaerobes, have the capacity to cope with oxidative stress, and it should not be considered a paradox that acetogens occur in habitats subject to fluxes of O₂.

Under anoxic conditions, RD-1 utilized both ethanol fermentation and the acetyl-CoA pathway as terminal, electron-accepting processes when grown on glucose (Fig. 5). In contrast, growth on pyruvate did not yield ethanol as a product. Thus, the simultaneous engagement of acetogenesis and fermentation was substrate dependent. When RD-1 was grown on glucose in the presence of O₂, ethanol and, to a lesser extent, lactate and H₂ became the main reduced end products, indicating that reductant flow was diverted away from the acetyl-CoA pathway when redox conditions became more oxic (Fig. 5). Ethanol and lactate fermentation are also important energy-conserving processes for the acetogen Ruminococcus productus U1, and the simultaneous engagement of multiple catabolic processes might contribute to the in situ competitiveness of acetogens (22).

View larger version (20K): [in this window] [in a new window]   FIG. 5.   Hypothetical scheme illustrating how oxic conditions cause a shift in the flow of reductant ([e]) by RD-1. The acetate formed under oxic conditions is derived from the oxidation of glucose (upper portion of scheme) rather than reductive synthesis from CO2 via the acetyl-CoA pathway (depicted as the acetate that is produced as a reduced end product under anoxic conditions (lower left branch of figure).

Many of the enzymes of the acetyl-CoA pathway are very sensitive to O₂ (21), and H₂-dependent acetogenesis by RD-1 was more sensitive to O₂ than was fermentation. O₂ is likewise inhibitory to the H₂-dependent acetogenic capabilities of certain termite gut isolates (Boga et al., Abstr. Ann. Meet. Verein. Allgem. Angewand. Mikrobiol. BioSpectrum, abstr. 15.P.1133, p. 143, 2000). The ionic strength of the medium also caused a shift in the metabolism of RD-1 towards the synthesis of ethanol. However, lactate was only formed from sugars when RD-1 was subjected to oxidative stress.

Molecular characterization of RD-1 indicated that it was a strain of C. glycolicum. Strains of C. glycolicum have been isolated from soil, mud, bovine intestine, human feces, and human wounds (23, 25). C. glycolicum DSM 1288^T is described as a saccharolytic fermenter that produces acetate, ethanol, CO₂, and H₂ (29). In contrast to RD-1, C. glycolicum DSM 1288^T had negligible H₂- and formate-utilizing capacities. RD-1 and C. glycolicum DSM 1288^T also had dissimilar temperature optima. Although acetogenic capacities were not readily apparent with C. glycolicum DSM 1288^T, the organism nonetheless contained carbon monoxide dehydrogenase. Carbon monoxide dehydrogenase activity is also present in the nonacetogenic butyrate fermenter Clostridium pasteurianum (18).

A phylogenetically uncharacterized H₂-utilizing acetogen, Clostridium sp. strain 22 (ATCC 29797), might be a strain of C. glycolicum (46; information from the American Type Culture Collection [Manassas, Va.] Bacteriology Program). Strain 22 was isolated on H₂-CO₂ from sewage sludge, is a gram-positive rod that forms subterminal spores, and has a pH optimum of 8.0 (46). Since RD-1 formed terminal spores and had a pH optimum of 7.0 to 7.5, RD-1 and strain 22 appear to be dissimilar. C. glycolicum DSM 1288^T might have lost its ability to reductively synthesize acetate from H₂-CO₂. Most acetogens grow poorly on H₂-CO₂, in part because of the thermodynamic constraints of the carbonyl branch of the acetyl-CoA pathway (21), and certain acetogens lose their ability to efficiently utilize H₂ after prolonged cultivation in the laboratory on H₂-CO₂ (39).

Phylogenetically, acetogens are not tightly clustered but are widely dispersed throughout the low-G+C-content, gram-positive bacteria (55). To date, no general 16S rRNA gene probe is available to detect acetogens in natural samples. The 16S rRNA gene probes used to evaluate the occurrence of acetogens in sea grass roots (38) are specific for acetogenic species of Acetobacterium (probe AW), the acetogen Eubacterium limosum (probe AW), and Clostridium species that group in cluster I of their genus (probe Clost I), respectively (15). About 60% of the epidermal cells, 24% of the outermost cortex cell layers, and 2% of the deeper cortex cell layers of H. wrightii roots were colonized with acetogens that hybridized with probe AW (38). However, based on 16S rRNA gene sequence similarities, RD-1 grouped in cluster XI of the genus Clostridium (15) and would not be detected by the 16S rRNA gene probes AW and Clost I. Further studies will be required to elucidate the number and types of acetogens associated with sea grass rhizospheres and to understand how these microorganisms affect the viability of sea grasses.