nai

Project Progress

IMAGES WERE EMBEDDED

Iron and sulfur cycling: microbial communities and biosignatures

This collaborative project involves many members of the UC Berkeley-led NAI team as well as members of other teams and institutions internationally. The objectives are focused around the question of how microbiology and iron and sulfur cycling are coupled, and the potential of this coupling for generation of biosignatures of possible value to Mars exploration. Other objectives include development of an understanding of how organisms adapt to environmental extremes that are associated with iron-sulfur systems, and their ecology. Iron and sulfur were chosen due to the abundance of these elements in multiple oxidation states at, or close to, the Martian surface.

Iron cycling in aqueous circumneutral pH environments

BioMARS-funded research has focused on studies of microbial Fe redox cycling in a variety of circumneutral-pH environments, including a groundwater Fe seep in Tuscaloosa, AL (Fig. 1A), unsaturated Triassic-age weathered basalt materials from Box Canyon, ID (Fig. 1B), creeks in VA, at Chocolate Pots, Yellowstone National Park), in Delaware Inland Bays and the hydrothermal vents at Kilo-Moana (20°3’S, 176°8’W), located on the East Lau Spreading Centre (ELSC), in the Lau Basin, SW Pacific Ocean.

{{ 1 }}

{{ 2 }}

An initial objective for our team was analysis of the potential of iron-rich volcanic rocks to support microbial life at springs that are potential analogous to inferred sapping channel discharge sites on Mars. As described elsewhere, subsequent work by out team (led by Bill Dietrich) has changed current thinking about the genesis of these channels. However, these features retain interest from the point of view of sites of redox reactions and water at the Martian surface. Incubation of materials from the groundwater seep and weathered basalt sites under anoxic conditions with organic carbon or hydrogen as an electron donor by members of the Roden group revealed the potential for rapid Fe(III) oxide reduction. In addition, culture-based enumerations (MPN) revealed significant numbers of organic carbon and hydrogen-oxidizing dissimilatory Fe(III)-reducing microorganisms (FeRMs) (_{1E4 to 1E6 per mL). These results, together with presence of abundant culturable Fe(II)-oxidizing microorganisms (FeOMs) in both systems (} 1E2 to 1E6), suggests the potential for microbial Fe redox cycling in these different habitats. 16S rRNA gene clone library data from the Fe seep and Box Canyon revealed the presence of a variety of heterotrophic and autotrophic phylotypes. The seep libraries were dominated by representatives of the β-proteobacteria, and relatives to certain FeRMs were found with sequences from the Aeromonadaceae. However, no sequences from well-known FeRM taxa such as the Geobacteraceae or Shewanella were present in the libraries. Nevertheless, FeRMs present in Fe(III) oxide-reducing enrichment cultures obtained from the Fe seep revealed sequence similarity to Geobacter bremensis (95% similarity) and Geobacter pelophilus (98%). A further sequence was obtained with high similarity to Dechlorosoma suillum (99%). The Box Canyon Fe(III)-reducing enrichment culture contains organisms like Sphingomonas panni (98%), Intrasporangium calvum (98%), Zimmermanella faecalis (99%) or Gracilibacter thermotolerans (96%), that have not previously been associated with dissimilatory Fe(III) reduction. Investigation of the Fe(III)-reducing capacity of this community of organisms is still under progress. Fe(II)-oxidizing isolates from both environments revealed the presence Comamonas testosteroni (100%). The Fe seep isolate grew in Fe(II)-O2 opposing gradients, exhibiting an ~ 48-h lag phase and direct correlation between cell production and Fe(II) oxidation (Fe(III) oxide formation) between 48 and 144h. No growth occurred in control systems lacking Fe(II), ruling out the possibility of growth on agar impurities or on trace amounts of hydrogen introduced into the culture tubes during preparation in the anaerobic chamber. Together our results suggest that microscale coupling of microbial Fe oxidation and reduction occurs in both habitats and provide a model for how microbially-catalyzed Fe redox cycling could take place in a wide variety of surface and subsurface. An interesting feature of the weathered basalts from Box Canyon is that they contain substantial quantities of magnetic mineral phases (Fig. 3A), which are comprised primarily of nanometer-sized maghemite crystals in close association with smectitic clay particles (Fig. 3B,C). These phases are analogous to the pedogenic (as compared to lithogenic) magnetic components of soils. A recent exciting result from this system is that dissimilatory Fe(III)-reducing microorganisms present in the weathered basalt are capable of converting these reddish-brown maghemite crystals into black magnetite nanoparticles (Fig. 3D), some of which became aligned in chains on smectite surfaces (Fig. 3E,F). Although we cannot state with certainty that the maghemite phases that were converted to magnetite are of biological origin, previous studies have attributed the presence of maghemite (and magnetite) in soils to the reaction of microbially-produced Fe(II) with amorphous Fe(III) oxide phases such as ferrihydrite. Detailed studies of maghemite in Chinese loess and paleosols have also concluded that all nanoscale ferromagnetic materials are formed via Fe(III) reduction by FeRMs. Thus it appears that these weathered basalts contain extant mineralogical biosignatures of microbial Fe redox processes.

{{ 3 }}

The team led by Eric Roden has also evaluated the composition and function of an anaerobic, nitrate-dependent Fe(II)-oxidizing enrichment culture obtained from Bernhard Schink and colleagues at the University of Konstanz in Germany (Straub et al., 1996, Appl. Environ. Microbiol., 62:1458-1460). The composition and function of the lithoautotrophic Fe(II)-oxidizing, nitrate-reducing enrichment culture described by was studied by a combination of molecular and cultivation techniques. A 16S rRNA gene clone library (60 clones) obtained with universal bacteria primers revealed that the culture is dominated (45 clones) by one phylotype with 94% similarity to Gallionella ferrugina L07897. Nine clones revealed a 94% similarity to Comamonas badia AB021354, and 6 clones a 97% similarity to Parvibaculum lavamentivorans AY38798. Using standard isolation procedures, we were able to isolate a Comamonas strain under nitrate-reducing conditions with acetate as the electron donor, and the Parvibaculum strain on aerobic tryptic soy broth media. The Comamonas and Parvibaculum isolates showed the same 16S rRNA genes sequences found in the clone library. Neither of these two strains alone or in combination were able to oxidize Fe(II) with nitrate. A microorganism capable of oxidizing Fe(II) with nitrate (presumably Gallionella) was isolated using roll tube methods. This isolate showed low Fe(II)-oxidizing activity in the first transfer to liquid culture (only 10-20% of the added 10 mM Fe(II) was oxidized after 8 weeks), and a stable Fe(II)-oxidizing isolate was not obtained. However, initial attempts to cultivate the Fe(II)-oxidizing organism under microaerophillic conditions in Fe(II)-O2 opposing gradient systems show positive results. Experiments examining the patterns of growth of the three organisms during nitrate-dependent Fe(II) oxidation by the enrichment culture are under way.

The Luther group led studies of additional environments where Fe(II) oxidation occurs. In two areas sulfide is not present (creeks in VA and Chocolate Pots, Yellowstone National Park) and in two other locations sulfide is present [local Delaware Inland Bays and the hydrothermal vents at Kilo-Moana (20°3’S, 176°8’W), located on the East Lau Spreading Centre (ELSC), in the Lau Basin, SW Pacific Ocean]. During this last year, two publications have been accepted on the in situ determination of the rates of Fe(II) oxidation (see further details below).

Iron cycling in a hypersaline lake environment
The NAI team recently initiated studies of the geochemistry of processes associated with iron-rich springs that discharge into a hypersaline lake. The saline, Fe-rich groundwater-fed sediments of Lake Tyrell (see Fig. 4 caption) are analogous to what might have formed on Mars during times of standing water, e.g. conditions that were originally (incorrectly) thought to have been present at Gusev Crater, and that recent evidence suggests could have been present at other locations (e.g. intercrater plains north of Holden Crater) on ancient Mars. Although the basis for potential microbial life in such environments is of course unknown, one possibility is upward transport of Fe(II)-rich fluids derived from acid-promoted mobilization (“bleaching”) of primary basalt minerals.

{{ 4 }}

Our team’s research focuses on Lake Tyrell, Victoria, Australia, a site previously identified for Mars analog studies. Of particular interest are layered iron oxyhydroxide deposits that form at the springs, and lake sediment samples in which there is evidence for coupling primary production in mats and biofilms to iron and sulfate reduction, forming nanocrystalline sulfide minerals. The goal of our studies is to gain insight into the biogeochemical organization of Fe-based microbial systems and identification of possible biosignatures associated with these reactions.

During fieldwork in 2005, 2006, and 2007 samples of concretions were collected. In addition, in 2006 we collected 4 sediment cores from the acidic southern part of the lake, and one core from the more neutral site (pH 7) in the center of Lake Tyrrell for chemical analysis and microbial enrichment culture experiments. From preliminary geochemistry data we observed a nearly constant pH around 4 which facilitates Fe(II) oxidation and also Fe(III) reduction processes. At depth of ~10 cm Fe(III) oxide precipitates start to accumulate and may be the starting point of the overall visible iron concretions on the sediment surface (see Fig. 4). Smaller concretions were also found in the sediment cores. Microscopic analysis indicate a low but easily detectable number of microbial cells and MPN analysis for aerobic (TSB) microbes support this observation with MPN of 3.5×104 cells/mL.

To date, we have conducted extensive mineralogical analyses of the iron concretions that outcrop at the southern end of the lake (Ph.D. research of Claudia Jones). The research makes use of state of the art textural and mineralogical characterization methods. We have also documented the geochemistry of the springs and developed some insights into the mechanism of formation of concretions. Analyses of layering in the concretions and possible associated organic films are ongoing.

With microbial culturing techniques the Roden group has established a stable haloacidophilic Fe(II)-oxidizing enrichment culture at 10% NaCl from pH 3-4 sediments and the first obtained 16S rRNA gene sequences revealed two different types of organisms similar to Thiobacilluas prosparus (95%) and Marinobacter sp. (98%). Both of these types of organisms are known to be capable of catalyzing Fe(II) oxidation. We also established haloacidophilc Fe(III)-reducing cultures at 20% NaCl and pH 3 with biomass from an Lake Tyrrell isolate (Salinisphaera sp. 99%). Salinisphaere species are very common in saline systems and we found this species at different sites (pH 3 and 7) at Lake Tyrrell. These results suggest that a coupling of microbial Fe oxidation and reduction takes place in the acidic and iron rich sites of Lake Tyrrell and may provide a model for how microbially-catalyzed Fe redox cycling under hypersaline conditions could take place in subsurface Martian environments where reduced fluids/solids contact oxygen-bearing water or water vapor.

In situ and culture-based studies of iron oxidation

The NAI team is also investigating both natural populations and pure cultures of FeOMs in an attempt to better understand how their physiology and ecology influences the mineralogy and geochemistry that are hallmarks of these organisms in the environment. As mentioned above, one focus of work by the Emerson and Luther groups has been to gain a better understanding the contribution that neutrophilic, oxygen-dependent FeOMs make to iron oxidation kinetics both in situ and in the laboratory. Field work led by Jeremy Rentz, a postdoc in the Emerson lab, showed that FeOMs contribute significantly to overall iron oxidation rates in the field, and that the rates may be population dependent. This work also showed that under closely monitored field conditions strictly chemical rates of iron oxidation were negligible, however rates of auto-oxidation on preformed Fe-oxides approximated those of the biological rates. This suggests that, in effect, the microbes are competing with themselves for Fe(II): they initiate Fe(II) oxidation and once microbially precipitated Fe-oxides have formed, these become good catalysts for abiological oxidation. Another study led by Greg Druschel, previously a postdoc in Dr. Luther’s lab, documents with field results from Contrary Creek in Virginia that FeOMs live in an ecological niche where O2 concentrations are less than 50 M and that Fe(II) concentration increases with depth. This is the first work that accurately describes the in situ environmental conditions where FeOMs reside in nature; this study also used pure cultures of FeOMs to establish the kinetics of Fe oxidation as they are controlled by O2 concentration. These results showed that above 50μM O₂, the bacteria could no longer compete with abiological oxidation. The results of both these sets of studies are in the process of being published with one paper in press at Environmental Science and Technology and another paper submitted to Geochimica et Cosmochimica Acta. In additional work at the Fe rich hot springs at Yellowstone National Park which have cyanobactrial mats, the Luther group has shown that it is possible to obtain in situ reaction kinetics as the mat goes from dark to sunlight (Fig. 5). This system has flowing water over the mat and is similar to areas on Mars suspected of having (or had) flowing water moving down hills into ravines. Analysis of the rate of Fe(II) loss with time shows that the reaction is zero order (Fig. 6; rate = k0) and not dependent on the concentration of either Fe(II) or O2. The zero order rate constant, k0, is a function of the cyanobacteria’s response to the intensity of sunlight and hence oxygen production; i.e., a catalytic effect by an organism. Thus it is now possible to use in situ solid state microelectrodes to distinguish different Fe(II) biosignatures in field experiments with robotic devices that could use light/dark filter experiments as we can now distinguish Fe(II) oxidation by FeOB from cyanobacteria. This detailed kinetic analysis has been submitted to the journal, Electroanalysis.

Figure 5. Fe(II) concentration change over time and on change from dark to light at 0.5 mm depth into the mat. A gold-amalgam electrode was used to obtain the data.

{{ 5 }}

Figure 6. (a) Zero order plot showing microbial catalysis; (b) pseudo first order plot of the data in Fig. 5.

{{ 6 }}

Iron oxide mineralization of polymers

In a study involving the Banfield and Emerson groups, work by Clara Chan, formerly NAI-supported graduate student in the Banfield lab, involved samples from Contrary Creek and other field sites. She used advanced spectroscopic techniques to investigate the role that biopolymers produced by FeOM play in controlling the crystallization of iron oxides. The work suggests that carboxyl groups of acidic polysaccharides are produced by different microorganisms to create a wide range of iron oxide biomineral structures. This intimate and potentially long-term association controls the crystal growth, phase, and reactivity of iron oxyhydroxide nanoparticles in natural systems. This work greatly extends studies reported in Science by Chan et al. and is currently in revision at Geochimica et Cosmochimica Acta

Research led by the Emerson group has investigated the population dynamics of FeOMs in a nearby small spring that contains an Fe(II)-rich flow of water. Copious Fe-rich mats develop downstream of the source water, which contains approximately 150μM Fe(II), and there is a natural gradient of Fe(II) which is all consumed within 10-15 m from the source. The team has monitored iron, pH, and temperature as well general trends in the microbial population that live in the site for nearly two years. The population undergoes a significant shift from a sheath-forming, Fe(II)-oxidizing population in the summer months similar to Leptothrix spp. to a stalk-forming population of Gallionella-like organisms in the winter and spring. It is currently unclear what drives these changes in population structure, while temperature is the most consistent variable, we have other study sites with very similar temperature regimens that do not exhibit these same kinds of population shifts. [Fig. 7,8]

{{ 7 }}

{{ 8 }}

Spectral studies

In research led by the Bishop group, the NAI team is analyzing spectral data of Mars including i) CRISM images for the presence of phyllosilicates and sulfates and ii) MER Gusev crater Pancam data of the bright salty soils. This also involves characterizing the spectral properties of i) phyllosilicates and sulfates having a variety of mineral structures, and ii) altered volcanic material containing phyllosilicates and sulfates. During this year they completed a study on alteration at Haleakala, Maui, where both phyllosilicates and sulfates formed near cinder cones. Field work was also perfomed at and near Kilauea, Hawaii, where solfataric alteration of ash deposits is taking place and where orange-colored Fe-Ti-S-Si-bearing coatings are forming near vent sites on lava. [Fig. 9,10,11,12]

{{ 9 }}

{{ 10 }}

Metal sulfides, cluster formation, and sulfide biominerals

The Luther group has continued to work on metal sulfide clusters, MSaq, as a source of reduced metals including Fe(II) that may be used by microorganisms instead of the aqueous Fe(II) cation. Recent work on inert CdS clusters was presented at the American Society of Mass Spectrometry conference. Clusters up to Cd4S6 were found in laboratory experiments. These clusters have relevance for formation of metal sulfide nanoparticles, which have been proposed as a possible Mars-relevant biosignature due to the low rate of inorganic sulfate reduction at low temperature (Banfield et al. 2003). In related work to that on metal sulfide clusters, John Moreau, a NAI-supported Ph.D. student in the Banfield lab, examined iron sulfide nanoparticle aggregates produced by sulfate-reducing bacteria. In recently published work (Moreau et al. Science, 2007), the team documented a surprisingly high concentration of protein intimately associated with micron-scale iron sulfide nanoparticle aggregates. Such aggregates have also been reported in ancient rocks, and may entrap organic molecules or traces of carbon that are indicative of their biological origin. The team also identified gradients in sulfur isotope composition across micron-scale sulfide mineral aggregates (Moreau et al. in prep.).

The Fe/S catalytic cycle was investigated in two environments by the Luther group. One site is a local Inland Bay area where several deep anoxic holes occur and the other area is the hydrothermal vent sites at Lau Bain off Fiji. At the oxic-anoxic interface of the Inland Bay, FeSaq clusters as well as dissolved Fe(II) and H2S interact with O2, leading to sulfide oxidation and up to 30 M elemental sulfur (published in 2006). Several hydrothermal sites at Lau Basin off Fiji were investigated which showed significant differences in Fe/S chemistry that correlated with macrofauna and their endosymbiotic microbes (still under investigation). Certain locations at a given site showed only FeS as the dominant redox chemical species. However, thiosulfate, polysulfides and Fe(III) were also detected indicating that FeS is being oxidized (the energy from this oxidation should also be beneficial for life forms). The partially oxidized sulfur species are produced as sulfide from a diffuse flow area penetrates the basalt substrate and reduces Fe(III) in the solid phase. The latter work is now being written for two journal publications. The electrode technology developed by the Luther group has been transferred to the U Hawaii NASA NAI as a former graduate student in the Luther lab has been hired there as an assistant Professor in the Oceanography department.

Iron sulfide mineral dissolution and a subsurface biosphere

In the past year, we have continued work at the Iron Mountain Richmond Mine site. This is a subsurface system in which solutions attain pH values of around 1.0 as the result of microbially promoted dissolution of iron sulfide (pyrite, FeS2). The focus is on understanding microbial biofilm consortia that are supported by iron and sulfur cycling in this extreme environment. An important goal for studies led by the Banfield group is molecular-level analysis of microbial community assembly, adaptation and evolution. Key to this effort has been the development of community (meta)genomics and community proteomic analysis methods that will be transferred to the hypersaline lake project described elsewhere in this report (see publications by Baker et al. 2006 and Lo et al. and Denef et al. 2007). An important finding by Brett Baker, an NAI-supported team member, has been the detection of five lineages of archaea that are highly divergent from previously described euryarchaeaota. In addition to possessing 16S rRNA genes that have mismatches to primers used for PCR-based surveys, the cells are unusual in a number of ways. The cells are remarkably small, apparently smaller than any previously described cells, and have an unusual morphology, with folded membrane protrusions of unknown function (this work was highlighted on the front page of the New York Times in December 2006). Based on their size, it has been possible to concentrate these cells from biofilms sufficiently to allow comprehensive population genomic analysis. The sequencing was funded by a sequencing proposal to DOE (Joint Genome Institute) and completed in 2007. The genome size of several characterized lineages appears small, no more than around 1 Mb (near complete genomes have been obtained for three of the five lineages). This work in is preparation for publication in 2007 or 2008.