Band 83, Heft 11

We studied intracellular magnetite particles produced by several morphological types of magnetotactic bacteria including the spirillar (helical) freshwater species, Magnetospirillum magnetotacticum, and four incompletely characterized marine strains: MV-1, a curved rodshaped bacterium; MC-1 and MC-2, two coccoid (spherical) microorganisms; and MV-4, a spirillum. Particle morphologies, size distributions, and structural features were examined using conventional and high-resolution transmission electron microscopy. The various strains produce crystals with characteristic shapes. All habits can be derived from various combinations of the isometric {111}, {110}, and {100} forms. We compared the size and shape distributions of crystals from magnetotactic bacteria with those of synthetic magnetite grains of similar size and found the biogenic and synthetic distributions to be statistically distinguishable. In particular, the size distributions of the bacterial magnetite crystals are narrower and have a distribution asymmetry that is the opposite of the nonbiogenic sample. The only deviation from ideal structure in the bacterial magnetite seems to be the occurrence of spinel-law twins. Sparse multiple twins were also observed. Because the synthetic magnetite crystals contain twins similar to those in bacteria, in the absence of characteristic chains of crystals, only the size and shape distributions seem to be useful for distinguishing bacterial from nonbiogenic magnetite.

Samples collected in low-temperature (2-50 °C) waters near hydrothermal vents of the Southern Explorer Ridge, in the northeast Pacific Ocean, contained fine (< 500 nm) Feand Mn-oxide and Fe-silicate particles coating bacterial surfaces. Partially to totally mineralized bacteria, along with bacterial exopolymers, were covered with a mixture of poorly ordered Si-rich Fe-oxides (possibly ferrihydrite), Mn-oxides, and Fe-silicates (possibly nontronite). Minerals occur as very fine (2-20 nm) granular material, fine (20-100 nm) needles and sheets, small (200-500 nm) nodules and filaments (i.e., mineralized exopolymers). Under saturation conditions, we infer that bacterial surfaces provided nucleation sites for poorly ordered oxides and silicates. The formation of Fe- and Mn-oxides was likely initiated by the direct binding of soluble Fe and Mn species to reactive sites (like carboxyl, phosphate, and hydroxyl groups) present within the bacterial cell wall and the exopolymers. Fe-silicate formation involved a more complex binding mechanism, whereas metal ions, such as Fe, possibly bridged reactive sites within the cell walls to silicate anions to initiate silicate nucleation.

Magnetite is a common product of bacterial iron reduction and may serve as a potential physical indicator of biological activity in geological settings. Here we report the formation of single-domain magnetite under laboratory conditions by a thermophilic fermentative bacterial strain TOR-39 that was isolated from the deep subsurface. Time-course analyses were performed at 65 °C to study the effect of bacterial activity on solution chemistry and magnetite formation during the growth of TOR-39. Run products were examined by transmission electron microscopy. Magnetite particles formed exclusively outside of bacterial cells and exhibited octahedral shapes having relatively equal length and width (< 15% difference). Tiny magnetite particles (< 12 nm) nucleated between 10 and 11 h of incubation and increased to average lengths of 55.4 ± 26.8 nm after 24 h of incubation. Between 24 h and 22 d of incubation, magnetite particles maintained average lengths of 56.2 ± 24.8 nm. Based on size constraints, greater than 85% of the particles observed fell within the magnetic single domain. Little to no magnetite was detected in abiotic controls at 65 or 95 °C, or in TOR-39 cultures whose activity was suppressed. Unlike mesophilic iron-reducing bacteria (e.g., GS-15), TOR-39 produced crystals having shapes and sizes similar to some particles produced intracellularly by magnetotactic bacteria. Thus the single- domain magnetite produced by thermophiles such as TOR-39 may represent a heretofore unrecognized biological contribution to natural remanent magnetization in sedimentary basins and other geothermal environments.

Iron oxide and hydroxides can be precipitated from solution with both Fe 2+ and Fe 3+ states by a microbial consortium enriched from surface water draining a granitic batholith. The Fe 2+ /Fe 3+ ratio of the microbial precipitate is determined by both the initial environment and subsequent diagenesis. To evaluate the thermal aspects of diagenesis, biological precipitates, either largely Fe 2+ or equally divided between Fe 2+ and Fe 3+ states, were heated at 80 °C for 12 weeks, under various redox conditions and compared to samples maintained under the same conditions at 4 °C . Mössbauer spectroscopy showed the iron oxide and hydroxides precipitated as Fe 2+ to be more stable than that as Fe 3+ . Only under air at 80 °C are the ferrous minerals altered to hematite, while the more labile ferric minerals are altered to Fe(OH) 2 at 4 °C and to hematite at 80 °C . In contrast, chemically precipitated Fe compounds, when incubated with the consortium, only form Fe 3 compounds, mainly fine-grained hematite. When no microbes are present, goethite is formed during diagenesis. Fe speciation in sediments may reflect a combination of microbial mediation that causes the initial precipitation of iron oxides and hydroxides and the subsequent conditions of the diagenetic processes characteristic of that particular depositional environment.

Microbiologic reduction of synthetic and geologic Fe 3+ oxides associated with four Pleistocene- age, Atlantic coastal plain sediments was investigated using a dissimilatory Fe reducing bacterium (Shewanella putrefaciens, strain CN32) in bicarbonate buffer. Experiments investigated whether phosphate and anthraquinone-2, 6-disulfonate, (AQDS, a humic acid analogue) influenced the extent of crystalline Fe 3+ oxide bioreduction and whether crystalline Fe 3+ oxides in geologic materials are more or less reducible than comparable synthetic phases. Anaerobic incubations (10 8 organisms/mL) were performed both with and without PO 4 and AQDS that functions as an electron repository and shuttle. The production of Fe 2+ (solid and aqueous) was followed with time, as was mineralogy by Xray diffraction. The synthetic oxides were reduced in a qualitative trend consistent with their surface area and free energy: hydrous ferric oxide (HFO).goethite.hematite. Bacterial reduction of the crystalline oxides was incomplete in spite of excess electron donor. Biogenic formation of vivianite [Fe 3 (PO 4 ) 2 ·8H 2 O] and siderite (FeCO 3 ) was observed; the conditions of their formation was consistent with their solubility. The geologic Fe 3+ oxides showed a large range in reducibility, approaching 100% in some materials. The natural oxides were equally or more reducible than their synthetic counterparts, in spite of association with non-reducible mineral phases (e.g., kaolinite). The reducibility of the synthetic and geologic oxides was weakly effected by PO 4 , but was accelerated by AQDS. CN32 produced the hydroquinone form of AQDS (AHDS), that, in turn, had thermodynamic power to reduce the Fe 3+ oxides. As a chemical reductant, it could reach physical regions of the oxide not accessible by the organism. Electron microscopy showed that crystallite size was not the primary factor that caused differences in reducibility between natural and synthetic crystalline Fe 3+ oxide phases. Crystalline disorder and microheterogeneities may be more important.

Surface colonization and microbial dissolution of pyrite were studied in the laboratory and by in situ surface colonization experiments conducted at Iron Mountain, California. Laboratory experiments involved organisms obtained from Iron Mountain and cultured in pH < 1.0, 42 °C solutions designed to enrich for chemolithotrophic species present at acidgenerating sites. Planktonic and sessile microorganisms grew in enrichment cultures containing small amounts of yeast extract. The maximum density of attached cells was approximately 8 × 10 6 cells/cm 2 . Attachment was specific for pyrite and occurred nonrandomly; rod-shaped bacteria tended to orient parallel to {100} and {110} pyrite. Attachment resulted in formation of euhedral dissolution pits. Cultures grown without yeast extract contained only planktonic cells and euhedral dissolution pits were not developed on the pyrite surface. All cultured organisms were identified as bacteria by fluorescence in situ hybridization and domain-specific probes. Leptospirillum ferrooxidans comprised 10-40% of planktonic organisms in both enrichments. Thiobacillus ferrooxidans was not identified in either enrichment. Oxidation rates were approximately equivalent in both enrichments (4 × 10 -7 μM Fe/cell·day) over a 28 day period. Pyrite cubes were exposed to natural solutions at Iron Mountain for two months. A subset of samples was exposed only to solutions that had passed through 0.22 μm Teflon filters. Denser colonization (by distinctive elongate bacteria not observed in laboratory cultures) occurred on pyrite in filter-covered vessels. Attachment specificity, orientation, and resulting degradation morphology were similar to that observed in laboratory cultures. Results show that interaction between attached cells and pyrite surface is highly specific and the impact on surface morphology evolution is different from that associated with planktonic microorganisms, despite the similarity in effect (per cell) on total dissolution rates.

The top 20 cm of sediments at active cold seeps in Monterey Bay, coastal California, contain framboidal pyrite that occurs as infillings and pseudomorphs of the chambers of the tests of foraminifera and rarely as irregularly shaped grains. Sulfur isotope compositions obtained with the ion microprobe show depletions in 34 S (δ 34 S = - 41 to -5‰, CDT), and large variations both within and among these pyrite grains. Intergranular differences in δ 34 S values in the same sediment are as large as 35‰, and intragranular zoning reaches 15‰. Zoning is regular in some grains, with systematic isotope changes from core to rim or from one foraminiferal chamber to another, but irregular in others. The regular zoning is consistent with an increase in 34 S through time. Backscattered-electron imaging reveals three types of pyrite: isolated framboids in a porous aggregation (‘‘PF-pyrite’’), agglomerated framboids with cementing interstitial pyrite (‘‘F+I-pyrite’’), and recrystallized pyrite with isolated relicts of framboids (‘‘RF-pyrite’’). In individual grains, RF-pyrite cores grade into F+I-pyrite toward grain rims, and F+I-pyrite grades into PF-pyrite at the grain edges. These textures are consistent with a paragenetic sequence whereby framboids first agglomerate (PF-pyrite), then cement (F+I-pyrite), and finally recrystallize (RF-pyrite). The δ 34 S values of RF-pyrite are generally lower than that of F+I-pyrite; if the paragenetic sequence is correct, then this trend parallels the regular core-rim isotopic zoning observed in some grains. The implied increase in δ 34 S with time is consistent with Rayleigh fractionation of sulfur in a closed system. Bacteria are intimately involved in the production of pyrite from our samples, and heterogeneous colonization by bacteria provides a simple explanation for the sulfur isotope heterogeneity among and within grains: The foraminifera provide open space for colonization and local nutrients for bacterial growth, whereas the cell walls of the bacteria may provide a local nucleation site for sulfides. If so, then initial colonization is reflected in lower δ 34 S values, whereas later bacterial emigration to other foraminifera chambers is indicated by higher δ 34 S values.

Using transmission electron microscopy, we studied the structures and compositions of Fe sulfides within cells of magnetotactic bacteria that were collected from natural habitats. Ferrimagnetic greigite (Fe 3 S 4 ) occurred in all types of sulfide-producing magnetotactic bacteria examined. Mackinawite (tetragonal FeS) and, tentatively, sphalerite-type cubic FeS were also identified. In contrast to earlier reports, we did not find pyrite (FeS 2 ) or pyrrhotite (Fe 1-x S). Mackinawite converted to greigite over time within the bacteria that were deposited on electron microscope grids and stored in air. Orientation relationships between the two minerals indicate that the cubic-close-packed S substructure remains unchanged during the transformation; only the Fe atoms rearrange. Neither mackinawite nor cubic FeS are magnetic, and yet they are aligned in chains such that when converted to magnetic greigite, the probable easy axis of magnetization, [100], is parallel to the chain direction. The resulting chains of greigite are ultimately responsible for the magnetic dipole moment of the cell. Both greigite and mackinawite magnetosomes can contain Cu, depending on the sampling locality. Because bacterial mackinawite and cubic FeS are unstable over time, only greigite crystals are potentially useful as geological biomarkers.

Microbial mats on the surfaces of modern, marine stromatolites at Highborne Cay, Bahamas, were investigated to assess the role of microbial processes in stromatolite formation. The Highborne Cay stromatolitic mats contain Schizothix as the dominant cyanobacterium and show millimeter-scale lamination: Some layers in the mat are soft (unlithified), whereas other layers are crusty (lithified). Lithified layers within the mats correspond to micritic horizons composed of thin (20-50 μm) micritic crusts, which commonly overlie truncated, micritized carbonate sand grains. These features are identical to lithified laminae in the underlying stromatolite; the micritic crusts are similar in thickness to micritic laminae in many ancient stromatolites. Biogeochemical parameters in a representative stromatolitic mat from Highborne Cay were measured to identify the role of bacteria in precipitation and dissolution of CaCO 3 . Depth distributions of O 2 , HS - , and pH were determined with microelectrode measurements in the field. Oxygen profiles were used to calculate photosynthesis and aerobic respiration. Sulfate reduction was determined using 35 SO 4 2- and sulfide oxidation potential was measured in homogenized samples. Our results indicate that cyanobacterial photosynthesis, sulfate reduction, and anaerobic sulfide oxidation in stromatolitic mats at Highborne Cay are responsible for CaCO 3 precipitation, whereas aerobic respiration and aerobic sulfide oxidation cause CaCO 3 dissolution. A close coupling of photosynthesis and aerobic respiration in the uppermost few millimeters of the mats results in no, or very little, net lithification. Sulfur cycling, on the other hand, shows a close correlation with the formation of lithified micritic layers. Photosynthesis, combined with sulfate reduction and sulfide oxidation results in net lithification. Sulfate reduction rates are high in the uppermost lithified layer and, on a diel basis, consume 33-38% of the CO 2 fixed by the cyanobacteria. In addition, this lithified layer contains a significant population of sulfide-oxidizing bacteria and shows a high sulfide oxidation potential. These findings argue that photosynthesis coupled to sulfate reduction and sulfide oxidation is more important than photosynthesis coupled to aerobic respiration in the formation of lithified micritic laminae in Highborne Cay stromatolites.

Uranium sorption experiments were carried out at ~25 °C using natural samples of the lichen Peltigera membranacea. Thalli were incubated in solutions containing 100 ppm U for up to 24 h at pH values from 2 to 10. Equilibrium sorption was not observed at less than ~6 h under any pH condition. U sorption was strongest in the pH range 4-5, with maximum sorption occurring at a pH of 4.5 and an incubation time of 24 h. Maximum U uptake by P. membranacea averaged ~42 000 ppm, or ~4.2 wt% U. This appears to represent the highest concentration of biosorbed U, relative to solution U activity, of any lichen reported to date. Investigation of post-experimental lichen tissues using electron probe microanalysis (EPM) reveals that U uptake is spatially heterogeneous within the lichen body, and that U attains very high local concentrations on scattered areas of the upper cortex. Energy dispersive spectroscopic (EDS) analysis reveals that strong U uptake correlates with P signal intensity, suggesting involvement of biomass-derived phosphate ligands or surface functional groups in the uptake process.

This study investigates the ability of the unicellular green alga Nannochloris atomus to precipitate CaCO 3 , quantifies mineral precipitation rates, estimates sediment production in a N. atomus bloom, and discusses the implications of microbial calcification for carbonate sediment deposition. A series of N. atomus cultures, isolated from Lake Reeve, Australia, were incubated at various pH and calcium concentrations to determine environmental parameters for calcification. Rates of calcification were calculated from initial and postincubation alkalinity, pH, and calcium measurements. Replicate experiments and controls consisting of non-calcifying cultures, uninoculated media, and dead cell cultures were performed using environmental culture parameters determined in series cultures. Average calcification rates from replicate experiments were used to predict daily sediment production rates in a small bloom of N. atomus. N. atomus precipitates 0.138 g/L of calcite in approximately 4 h when incubated at pH 8.5, 14.24 mM calcium concentration, 33 °C, 100 μE/m 2 /s light intensity, and a cell population density of 10 7 cells/mL. Assuming continuous precipitation, this corresponds to a maximum estimated sediment production rate of 1.6 × 10 6 kg of CaCO 3 per 12 h day in a single bloom of 3.2 × 10 9 L. Our results suggest that microbial calcification contributes significantly to the carbonate sediment budget.

Soluble proteins in the scallop (Patinopecten yessoensis) foliated calcite shell layer were characterized using biochemical and molecular biological techniques. SDS PAGE of these molecules revealed three major protein bands, 97 kD, 72 kD, and 49 kD in molecular weight, when stained with Coomassie Brilliant Blue. Periodic Acid Schiff staining and Stains-All staining indicated that these proteins are slightly glycosylated and may have cation-binding potential. N-terminal sequencing of the three proteins revealed that all three share the same amino acid sequence at least for the first 20 residues. A partial amino acid sequence of 436 amino acids of one of these proteins (MSP-1) was deduced by characterization of the complementary DNA encoding the protein. The deduced sequence is composed of a high proporton of Ser (31%), Gly (25%), and Asp (20%), typifying an acidic glycoprotein of mineralized tissues. The protein has a basic domain near the Nterminus and two highly conserved Asp-rich domains interspersed in three Ser and Glyrich regions. In contrast with prevalent expectations, (Asp-Gly)n-, (Asp-Ser)n-, and (Asp- Gly-X-Gly-X-Gly)n-type sequence motifs do not exist in the Asp-rich domains, demanding revision of previous theories of protein-mineral interactions.

Laboratory growth experiments were conducted to investigate the oxygen isotope effects associated with bacterial metabolism of phosphatic compounds commonly available in nature. The observed oxygen isotope fractionations suggest complex patterns of exchange between dissolved inorganic phosphate (P i ) and water, and significant circulation of P i between intracellular and extracellular locations with extensive recycling of the dissolved Pi pool, even at high concentrations of dissolved P i . Results of these experiments also support current models for bacterial utilization of phosphate. These results have important implications for the use of δ 18 O values of dissolved P i to trace sources of P, and bear on integrity of original oxygen isotope compositions of biogenic and sedimentary apatite minerals that have been subjected to processes of recrystallization and diagenesis. SEM images of laboratory synthesized apatite minerals show that similar textural features may be produced by microbially mediated and abiotic reactions, and that spheroidal structures may be produced by processes of dissolution as well as precipitation. The interpretation of certain mineral structures as microbial in origin solely on the basis of morphological and textural features may be misleading.

Phosphorus and nitrogen are essential macronutrients necessary for the survival of virtually all living organisms. In groundwater systems, these nutrients can be quite scarce and can represent limiting elements for growth of subsurface microorganisms. In this study we examined silicate sources of these elements by characterizing the colonization and weathering of feldspars in situ using field microcosms. We found that in carbon-rich anoxic groundwaters where P and N are scarce, feldspars that contain inclusions of P-minerals such as apatite are preferentially colonized over similar feldspars without P. A microcline from S. Dakota, which contains 0.24% P 2 O 5 but < 1 μmol/ g NH 4 + was heavily colonized and deeply weathered. A similar microcline from Ontario, which has no detectable P or NH 4 + , was barren of attached organisms and completely unweathered after one year. Anorthoclase (0.28% P 2 O 5 , ~1 mmol/g NH 4 + ) was very heavily colonized and weathered, whereas plagioclase specimens (<0.01% P, <1 mmmol/g NH 4 + ) were uncolonized and unweathered. In addition, the observed weathering rates are faster than expected based on laboratory rates. We propose that this system is particularly sensitive to the availability of P, and the native subsurface microorganisms have developed biochemical strategies to aggressively scavenge P (or some other essential nutrient such as Fe 3+ ) from resistant feldspars. The result of this interaction is that only minerals containing P will be significantly colonized, and these feldspars will be preferentially destroyed, as the subsurface microbial community scavenges a limiting nutrient.

We report the detection of living colonies of nano-organisms (nanobes) on Triassic and Jurassic sandstones and other substrates. Nanobes have cellular structures that are strikingly similar in morphology to Actinomycetes and fungi (spores, filaments, and fruiting bodies) with the exception that they are up to 10 times smaller in diameter (20 nm to 1.0 μm). Nanobes are noncrystalline structures that are composed of C, O, and N. Ultra thin sections of nanobes show the existence of an outer layer or membrane that may represent a cell wall. This outer layer surrounds an electron dense region interpreted to be the cytoplasm and a less electron dense central region that may represent a nuclear area. Nanobes show a positive reaction to three DNA stains, [4′,6-diamidino-2 phenylindole (DAPI), Acridine Orange, and Feulgen], which strongly suggests that nanobes contain DNA. Nanobes are communicable and grow in aerobic conditions at atmospheric pressure and ambient temperatures. While morphologically distinct, nanobes are in the same size range as the controversial fossil nannobacteria described by others in various rock types and in the Martian meteorite ALH84001.

Mineral dissolution experiments using batch cultures of soil and groundwater bacteria were monitored with solution chemistry and various microscopic techniques to determine the effects of these organisms on weathering reactions. Several strains of bacteria produced organic and inorganic acids and extracellular polymers in culture, increasing the release of cations from biotite (Si, Fe, Al) and plagioclase feldspar (Si, Al) by up to two orders of magnitude compared to abiotic controls. Microbial colonies on mineral grains were examined by cryo-scanning electron microscopy (cryo-SEM), confocal scanning laser microscopy (CSLM), and epifluorescence microscopy. Bacteria colonized all mineral surfaces, often preferentially along cleavage steps and edges of mineral grains. Low-voltage high-resolution cryo-SEM of high-pressure cryofixed and partially freeze-dried colonized minerals showed many bacteria attached by extracellular polymers of unknown composition. These biofilms covered much larger areas of the mineral surfaces than bacterial cells alone. Mineral surfaces where bacteria and extracellular polymers occurred appeared more extensively etched than surrounding uncolonized surfaces. CSLM was used to observe microbial colonization of biotite and to measure pH in microenvironments surrounding living microcolonies using a ratiometric pH-sensitive fluorescent dye set. A strain of bacteria (B0693 from the U.S. Department of Energy Subsurface Microbial Culture Collection) formed large attached microcolonies, both on the outer (001) surface and within interlayer spaces as narrow as 1 mm. Solution pH decreased from near neutral at the mineral surface to 3-4 around microcolonies living within confined spaces of interior colonized cleavage planes. However, no evidence of pH microgradients surrounding exterior microcolonies was noted.

Previous studies have documented dissimilatory growth of bacteria on solid Mn 4+ oxide, but Mn 3+ oxides have not been previously studied; here we have demonstrated for the first time the bacterial reduction of manganite. Strain MR-4 of Shewanella putrefaciens was able to grow on and rapidly reduce insoluble needle-shaped crystals of synthetic manganite (MnOOH), converting them to soluble Mn 2+ in the process. The rate of Mn 3+ reduction was optimal at pH of 7.0 and 26 °C consistent with an enzymatic reaction. In addition the rates of reduction were in proportion to the amount of manganite added, but nearly independent of the cell concentration present (e.g., cell number had only a small effect on the rate of Mn 3+ reduction at early stages of growth) suggesting that surface properties were dictating the rates of metal reduction. This thesis was supported by major differences in reduction rates when Mn oxides of different surface areas were studied. Removal of the carbon source (formate or lactate) or addition of metabolic inhibitors reduced the rate of metal reduction. No Mn31 reduction was observed when the samples were oxygenated, nor when the cells were separated from the Mn 3+ oxide by a dialysis membrane. Environmental scanning electron microscopy (ESEM) images showed close contact of the cells with the needle-shaped crystals during early stages of reduction. In later stages, the closely associated cells were coated with a layer of extracellular polymeric material that obscured the cells when viewed by ESEM. When manganite crystals were dried and gold coated, and viewed by standard scanning electron microscopy (SEM), abundant bacteria could be seen on the surface of the metal oxide in a thin biofilm-like layer. The layer of extracellular polymer is a new finding, and neither the composition, function, nor importance in the manganese reduction process have been elucidated.

Microorganisms can accelerate the rate of Mn 2+ oxidation by up to five orders of magnitude compared to abiotic Mn 2+ oxidation. Mn 2+ oxidation in Pseudomonas putida strain GB-1 involves an enzyme incorporated in the outer membrane that oxidizes Mn 2+ extracellularly. This Mn 2+ -oxidizing factor has to be synthesized inside the cell and transported across the outer membrane. We used a method known as transposon mutagenesis to generate two mutants that are incapable of Mn 2+ oxidation because they are unable to transport the Mn 2+ 1-oxidizing factor across the outer membrane. However, when cells were lysed, Mn 2+ oxidation occurred, verifying that transport and not synthesis of the Mn 2+ -oxidizing factor was affected. Transport of the Mn 2+ 1-oxidizing factor was restored when normal sequences obtained from a genomic library of Pseudomonas putida strain GB-1 were introduced into the mutant strains. By sequencing the DNA of the disrupted genes of these two mutants it was determined that the affected genes are very similar to the xpc gene family of the related species, Pseudomonas putida WCS358 and Pseudomonas aeruginosa. This gene family is known to be involved in the two-step protein secretion process in Gram-negative bacteria. In one of the two mutants, the disruption occurs in the gene that encodes a subunit of a complex that spans the membrane. In the other mutant the disruption occurs in the pilD/xcpA gene, which encodes an enzyme (peptidase) that modifies the subunits that are assembled into this membrane-spanning complex. These results indicate the involvement of a two-step protein secretion pathway in the transport of the Mn 2+ - oxidizing factor of Pseudomonas putida strain GB-1.

Sulphur River in Parker Cave, Kentucky receives sulfurous water (11-21 mg sulfide/L) from the Phantom Waterfall and contains a microbial mat composed of white filaments. We extend a previous morphological survey with a molecular phylogenetic analysis of the bacteria of the microbial mat. This approach employs DNA sequence comparisons of small subunit ribosomal RNA (SSU rRNA) genes obtained from the mat with those from an extensive database of rRNA sequences. Many of SSU rRNA gene clones obtained from the mat are most similar to rRNA sequences from sulfur-oxidizing bacteria (Thiothrix spp., Thiomicrospira denitrificans, and ‘‘Candidatus Thiobacillus baregensis’’). The Sulphur River SSU rRNA gene clones also show specific affiliations with clones from environmental surveys of bacteria from deep-sea hydrothermal vent communities and subsurface microcosms. Affiliations with sequences from bacteria that are known to have the ability to obtain energy for CO 2 fixation from the oxidation of inorganic compounds (chemoautotrophs), in combination with the environmental conditions surrounding the microbial mat, indicate that chemoautotrophic metabolism of bacteria in this mat may contribute to the biomass of Sulphur River. Cave communities, such as the one identified in Sulphur River, provide sites to study such relatively autonomous chemoautotrophic communities that are much more accessible than similar communities associated with deep-sea hydrothermal vents. Subsurface microbiology and the contribution of microbial activity on cave development are also discussed.

The inability to culture the majority of microorganisms from natural habitats has been well documented. In response to this constraint, the development and application of culture- independent methods for studying microbial communities have been a focus of microbial ecology for the past several years. Molecular approaches, in particular, offer great potential to investigate the role of microorganisms in areas as diverse as biogeochemistry, pollutant remediation, and disease causation. The purpose of this review is to give a brief critique of nucleic acid-based techniques currently used in microbial ecology, to serve as a reference for those wishing to employ such techniques, and to examine the applicability of these techniques toward studying the biogeochemical roles of bacteria, especially as related to metal cycling and biomineralization. Although published reports on the application of molecular techniques to biogeochemical and biomineralogical problems are scarce, this review aims to familiarize biogeochemists and biomineralogists with the strengths and weaknesses of these approaches. Advances have been made in describing the diversity of uncultured microbial communities, detecting/identifying specific cells and functional genes in environmental samples, detecting in situ activities of specific microorganisms, and determining some of the factors controlling gene expression in in vitro studies. However, the current methods provide limited quantitative information about natural processes. Acknowledging these limitations can assist the development of methods to answer basic questions concerning the in situ distribution and activities of microorganisms. Studies relating these in situ measurements of microbial activity and distribution to the physical and chemical microenvironment will ultimately permit a better understanding of the importance of microorganisms in mineral formation and dissolution processes.