Recent Results from Studies of Extremophilic Anoxygenic Phototrophs and Other Sulfur-Cycling Prokaryotes

In the past 5 years the Madigan lab has isolated and characterized various prokaryotes from extreme environments. Under work supported by NSF this has involved anoxygenic phototrophs, sulfate-reducing bacteria, and sulfur-oxidizing chemolithotrophs. The following pages contain a summary of our results to date. Other results can be found in the publications cited at the end of the description.

Overview. NSF support within the past five years: OPP9809195 (October, 1998­September, 2001, Biodiversity and Biogeochemistry of Antarctic Photosynthetic Bacteria), and OPP0185481 (January 2001­December 2002, Diversity, Vertical Distribution, and Metabolic Activities of Inorganic Sulfur- Cycling Prokaryotes in Lake Fryxell, Antarctica). Fieldwork in Antarctica (McMurdo) occurred in November­December of '99 and 01'. From these studies we have: (1) obtained cultures of psychrophilic phototrophic bacteria and know that several phylotypes of such organisms exist (Figs. 1­3); (2) detected species of Archaea as well as Bacteria; (3) cultured cold-active sulfur-cycling organisms (Figs. 4 and 5); (4) obtained depth profiles of some of these organisms (Fig. 3); (5) developed molecular probes for the study of the diversity and distribution of specific physiological groups (Fig. 3); and (6) isolated and characterized alkaliphilic anoxygenic phototrophs (Fig. 6).
Purple bacteria from Antarctic microbial mats and Lake Frxyell. We isolated and described a purple phototrophic bacterium, Rhodoferax antarcticus (Fig. 1a) (Madigan et al. 2000), from an Antarctic microbial mat. Rfx. antarcticus is the first example of a psychrophilic anoxygenic phototroph; it grows optimally at 18°C, will not grow above 25°C, and grows slowly at 0°C. Physiologically, Rfx. antarcticus is capable of both photoautotrophy (CO2 + H2 Æ CH2O) and nitrogen fixation (N2 Æ NH3, Fig. 1b). Thus, Rfx. antarcticus may be ecologically important as a source of both organic carbon and fixed nitrogen to other dry valley lakes microorganisms. In terms of basic research, however, the isolation of Rfx. antarcticus makes available a model system for studying how evolution has shaped metabolism and pigment-protein complexes in phototrophic bacteria to function optimally in the cold.
Rfx. antarcticus is closely related to the mesophilic species Rfx. fermentans (97.5% 16S rDNA sequence identity), and because of this, DNA hybridization experiments were performed to better characterize the two organisms (Madigan et al. 2000). Interestingly, dot blots (Fig. 1c) of genomic DNA from these two bacteria showed no detectable relationship, indicating that they were indeed distinct species and perhaps even separate genera, despite their close phylogenetic relationship. It is hypothesized that Rfx. antarcticus contains genes, absent from Rfx. fermentans, which encode cold-active proteins and other biomolecules critical for life at cold temperature. Additional examples of novel genomes from dry valley lakes prokaryotes that have close mesophilic relatives will almost certainly emerge from cultures to be isolated in the MO.

Fig. 1. Rhodoferax antarcticus. (a) Scanning electron micrograph of cells of Rfx. antarcticus. An average cell dimension is 0.7 x 2-3 mm. (b) Growth of Rfx. antarcticus at 18°C on N2 as sole nitrogen source vs. an Argon control. (c) Dot blot genomic DNA:DNA hybridization of digested DNA from A, Rfx. antarcticus and B, Rfx. fermentans probed with digoxigenin-labeled genomic DNA from 1, Rfx. antarcticus or 2, Rfx. fermentans.

Molecular Probes Targeted to Anoxygenic Phototrophs. Following the isolation of Rfx. antarcticus a molecular approach was taken to determine the diversity of anoxygenic phototrophs in Lake Fryxell and as a roadmap for further enrichment cultures. Primer sets were designed for all known groups of phototrophic bacteria (Achenbach et al. 2001). Ribosomal RNA primers were made specific to the heliobacteria (Gram-positive phylum), the green sulfur bacteria, and the green nonsulfur bacteria. However, due to the phylogenetic and metabolic diversity of Proteobacteria, the group that includes the purple bacteria, the 16S rRNA gene was not a useful tool. Instead, pufM, a gene that encodes the M subunit of the photosynthetic reaction center and which is universally distributed among purple phototrophic bacteria, was used as the target (see Fig. 3). Appropriate controls confirmed the specificity and reliability of the pufM probe (Achenbach et al. 2001).
Using this suite of probes, several different anoxygenic phototrophs were discovered in Antarctic lakes. Green sulfur bacteria were identified in samples of several Vestfold Hills lakes (east Antarctica, 68° S) obtained from John Bowman (Tasmanian Antarctic Research Center); green bacteria were not detected in dry valley lakes (Achenbach et al. 2001). From the Vestfold Hills samples, green sulfur bacteria were cultured (Fig. 2a) and bacteriochlorophyll c identified as their main pigment (Fig. 2b). Thus far we have determined that these organisms are halophilic, requiring marine salinities (Vestfold Hills lakes are the Vestfold Hills green bacteria represent a new temperature class of these organisms (Fig. 2c). Another unusual feature of these organisms was the fact that their chlorosomes (structures that contain the light-harvesting pigments in green bacteria) were unusually large, some 25% larger than those of mesophilic or thermophilic green bacteria (Fig. 2d). This enlargement of the photosynthetic apparatus may be an adaptation to the extreme low light conditions that exist in Antarctic lakes (light must penetrate 4-6 m of ice and 15 m of lake water before reaching the bloom of green bacteria).
A technical glitch has slowed progress with the green bacteria and is the reason that a formal characterization of these organisms is yet to be published. Phylogenetic analyses revealed that the putative pure culture from Pendant Lake (strain PDL, see Fig. 2) was indeed free of nonphototrophic contaminants, but contained two different phylotypes of green bacteria. Phylogenetic analyses showed that either of the phylotypes could be considered new species, one of the genus Chlorobium and one of Prosthecochloris. However, since the culture was not axenic, we had to repeat our purification protocols. The usual method of purifying phototrophic sulfur bacteria employs agar dilution tubes. However, this has proven unsuccessful with Antarctic green bacteria, since unlike mesophilic or thermophilic green bacteria, cold-active green bacteria do not survive even brief exposure to molten agar. Thus, we have pursued pure cultures using extincting dilution. One strain (strain AW from Ace Lake, Vestfold Hills) is approaching axenic culture and characterization of this organism should begin soon.

Fig. 2. Green sulfur bacteria from the Vestfold Hills, Antarctica. (a) Color phase photomicrograph and (b) absorption spectra of Chlorobium sp. strain PDL. Inset shows tube culture of strain PDL. (c) Growth as a function of temperature for strain PDL vs. mesophilic (C. limi-cola) and thermophilic (C. tepidum) species of chlorobia. (d) Transmission electron micrograph of a cell of Chlorobium strain PDL showing the large chlorosomes (arrow) of this organism.

Cultures of purple bacteria from Lake Fryxell. Spectral analyses of concentrated Lake Fryxell water collected from beneath the oxycline clearly showed the signature of phototrophic purple bacteria; absorption maxima were obtained near 800 and 850 nm, characteristic of bacteriochlorophyll a. Isolation of bulk DNA and amplification with pufM gave a PCR product at all depths tested (Fig. 3a). Resolution of these bands by DGGE showed that several phylotypes of purple bacteria were present and that their distribution was depth dependent (Fig. 3b). Enrichment cultures (4°C in dim light) yielded two morphotypes, a filamentous form and a club-shaped form, both of which are now in culture (Figs. 3c and 3d). Absorption spectra of cells from these enrichments (Fig. 3f) were similar to those of concentrated Lake Fryxell water. Physiological characterizations and analyses of pufM products (Fig. 3b) indicated that these organisms were purple nonsulfur bacteria (b-Proteobacteria). Unexpectedly, these organisms were found to contain gas vesicles (Fig. 3e), structures heretofore unknown in purple nonsulfur bacteria. Gas vesicles are likely essential for positioning these phototrophs in the water column in such a way as to receive optimal light and sulfide for photosynthesis. An axenic culture of the filamentous form (Fig. 3c) has recently been obtained and detailed studies of it are ongoing. The club-shaped form (Fig. 3d) grows only at very low temperatures and its isolation is a work in progress.

Fig. 3. Purple phototrophic bacteria from Lake Fryxell. (a) Amplification using pufM primers of DNA isolated with depth from Lake Fryxell. (b) DGGE analysis of pufM from the 9 and 10 m depths. At least three phylotypes are apparent, with the B-type dominating. (c), (d) Phase photomicrographs of enrichments of purple nonsulfur bacteria obtained by conventional culturing techniques from a 10 m water sample. (e) Thin section TEM of a cell shown in d. Note the abundant gas vesicles. (f) Absorption spectrum of cells from d-e.

Nonphototrophic sulfur-cycling prokaryotes from Lake Fryxell. Unlike Lake Bonney, Lake Fryxell contains significant levels of sulfide and thus an active sulfur-cycling microflora is present. Preliminary culturing studies from samples collected this past December have confirmed this. We now have enrichment cultures of sulfate-reducing bacteria and DMSO-reducing bacteria, and a pure culture of a Thiobacillus-like sulfur chemolithotroph from Lake Fryxell (Figs. 4 and 5).
Enrichment cultures for sulfate-reducing bacteria were established at 4º, 10º, 18º, and 25ºC using a lactate/sulfate medium containing a sulfide trap (Fe2+) to prevent sulfide toxicity (Fig. 4a). Enrichments developed rapidly at 4° and 10º, more slowly at 18º, and not at all at 25º. This indicates that cold-adapted sulfate-reducers are indeed present in Lake Fryxell. At least two morphotypes of sulfate-reducing bacteria were obtained. At 4° and 10°, large rod-shaped organisms predominated (Fig. 4b), while at 18º, lemon-shaped organisms resembling cells of the genus Desulfobulbus predominated (Fig. 4c). Isolations are proceeding and pure cultures of these organisms are anticipated soon, since unlike for the phototrophs, isolation in agar dilution tubes has been possible. Lake Bonney is unusual among Taylor Valley lakes in its abundance of organic sulfur compounds, primarily dimethylsulfoxide (DMSO) and its reduced product, DMS (see Introduction). Lake Fryxell lacks these compounds, yet enrichments for DMSO-reducing bacteria were readily obtained from this source. As was found with sulfate-reducers from Lake Fryxell (Fig. 4), growth of these organisms only occurred below 25ºC and growth was rapid at 10ºC. The DMSO-reducers were facultative Gram-negative rods (Fig. 4d) and ribosomal RNA sequencing and physiological characterizations of them are in progress.
Primary productivity measurements in Lake Fryxell have shown a significant fraction of the autotrophy to occur from non-phototrophic sources. Measurements of primary productivity with depth have shown a major spike of CO2 fixation in the sulfide-containing zones (Priscu et al. 1987), and we hypothesized that this activity is due to sulfur chemolithotrophs. From our 01' samples we established enrichment cultures for such organisms using thiosulfate as substrate. Primary liquid enrichments (10ºC) developed within one week and we now have pure cultures of these organisms (Fig. 5). Oxidation of thiosulfate leads to the production of sulfur (S2O32­ Æ S0, Fig. 5), and all isolates obtained thus far are short Gram-negative rods resembling species of Thiobacillus (Fig. 5b). The 9 m isolate is currently undergoing 16S sequencing and physiologically appears to be an obligate chemolithotroph,since tests for growth on various organic media have been negative. Interestingly, unlike the sulfate-reducing and DMSO reducers, Lake Fryxell sulfur chemolithotrophs grow best at 20­25ºC; no growth occurs at 30ºC and slow growth occurs at 4°C. More psychrophilic species of sulfur chemolithotrophs will be pursued in the MO, including actual sulfide-oxidizing species (see Culturing).

Fig. 4. Sulfate-reducing (SR) and DMSO-reducing prokaryotes from Lake Fryxell. (a) Bottles (uninoculated left; inoculated right) of SR medium from a 10°C enrichment. Note FeS production in right bottle from the reaction H2S + Fe2+ Æ FeS + 2H+. (b) Phase photomicrograph of cells from a 10C SR enrichment. (c) Cells from an 18C SR enrichment. (d) Cells from a 10C DMSO-reducing bacteria enrichment.

Fig. 5. Sulfur chemolithotrophs. (a) Colonies on a mineral salts/thiosulfate plate of a Thiobacillus species isolated from 9 m Lake Fryxell water. Note the yellow color of the colonies from elemental sulfur produced from the reaction: S2O32- Æ S0 + SO32-. (b) Phase photomicrograph of cells from the colonies in a. Note refractile sulfur globules.

Other extremophilic phototrophic bacteria. Our ongoing interest in the biodiversity of extremophilic anoxygenic phototrophs has also borne fruit; new genera and several species of alkaliphilic phototrophic bacteria have been described from the Madigan lab. These organisms, Heliorestis daurensis (Bryantseva et al. 1999) (Fig. 6a) and Rhodobaca (Milford et al. 2000) (Fig. 6b) are strongly alkaliphilic and inhabit cold Siberian and warm African soda lakes, respectively. In addition, another isolate, Heliorestis strain HH (Fig. 6c), was isolated from samples obtained from an expedition to Egyptian soda lakes that took place before an NSF-sponsored Egyptian/USA conference on microbial ecology held in Cairo, Egypt (May, 2001).

Fig. 6. Alkaliphilic anoxygenic phototrophs. (a) SEM of a cell of Heliorestis daurensis (helio-bacteria). (b) TEM of a cell of Rhodobaca bogoriensis (purple bacteria). (c) SEM of a cell of Heliorestis strain HH (heliobacteria).

Conclusions. In summary our work has: (1) given us valuable Antarctic experience and a taste of the exciting types of organisms that exist there, and (2) allowed us to troubleshoot some of the key problems associated with culturing cold-dwelling prokaryotes. Geochemical and cultural studies described in this proposal clearly indicate that C-N-S cycling is occurring in dry valley lakes, and thus the stage is set for an in depth study of the indigenous prokaryotes that participate in these cycles. It is predicted that the dry valleys MO will reveal the microbial diversity that underlies these nutrient cycles from both a molecular and cultural perspective. Moreover, the MO will plug a number of gaps in our understanding of nutrient cycling in Antarctic lakes, especially of C-and N-cycling organisms, where little data currently exist. Finally, the results presented here are proof that conventional culturing techniques can still yield interesting organisms. Extension of this work in the MO by the Madigan lab coupled with the high throughput system of Lanoil/Giovannoni, should many additional new organisms encompassing a variety of physiologies. Cultures of these organisms will make available a physiologically comprehensive set of psychrophilic and psychrotolerant prokaryotes from a single cold environment for further microbiological and genetic studies.

Publications Citing OPP9809195 or OPP 0185481

Madigan, M.T. 1998. Isolation and characterization of psychrophilic purple bacteria from Antarctica, pg. 699-706. In: Peschek, G.A., W. Löffelhardt and G. Schmetterer (eds.), The Phototrophic Prokaryotes. Plenum Press, New York.
Madigan, M.T. and A. Oren. 1999. Thermophilic and halophilic extremophiles. Curr. Opin. Microbiol. 2: 265-269.
Bryantseva, I.A., V.M. Gorlenko, E.I. Kompantseva, L.A. Achenbach, and M.T. Madigan. 1999. Heliorestis daurensis gen. nov. sp. nov., an alkaliphilic coiled to rod-shaped phototrophic heliobacterium from an alkaline Siberian soda lake. Arch. Microbiol. 172: 167-174.
Madigan, M.T., D.O. Jung, C.R. Woese and L.A. Achenbach. 2000. Rhodoferax antarcticus, sp. nov. a moderately psychrophilic purple nonsulfur bacterium from an Antarctic microbial mat. Arch. Microbiol. 173: 269-277.
Milford, A.D., D.O. Jung, L.A. Achenbach and M.T. Madigan. 2000. Rhodobaca bogoriensis gen. nov. and sp. nov., an alkaliphilic purple nonsulfur bacterium from African Rift Valley soda lakes. Arch. Microbiol. 174: 18-27.
Madigan, M.T. 2000. Bacterial habitats in extreme environments, pp. 61­72. In: J. Seckbach (ed.), Journey to Diverse Microbial Worlds­adaptation to exotic environments. Kluwer, Dordtrecht, The Netherlands.
Madigan, M.T. 2001. Physiological ecology and evolution of thermophilic anoxygenic phototrophs, pp. 103­124. In: A.L. Reysenbach, M. Voytek and R. Mancinelli (eds.), Thermophiles: Biodiversity, Ecology, and Evolution. Plenum Press, New York.
Madigan, M.T. 2001. Heliobacteriaceae, pp. 625­630. In Boone, D., Castenholz, R.W., and Garrity G.M. (eds.), Bergey's Manual of Systematic Bacteriology, 2nd.ed.,Vol. 1, Springer-Verlag, New York.
Takaichi, S., D.O. Jung and M.T. Madigan. 2001. Accumulation of unusual carotenoids in the spheroidene pathway, demethylspheroidene and demethylspheroidenone, in an alkaliphilic purple nonsulfur bacterium Rhodobaca bogoriensis. Photosyn. Res. 67: 207-214.
Madigan, M.T. 2001. The family Heliobacteriaceae. The Prokaryotes (electronic version). Springer-Verlag, New York. http://www.prokaryotes.com
Achenbach, L.A., J.R. Carey and M.T. Madigan. 2001. Photosynthesis and phylogenetic primers for the detection of anoxygenic phototrophs in natural environments. Appl. Environ. Microbiol. 67: 2922-2926.
Nübel U., M.M. Bateson, M.T. Madigan, M. Kuhl and D.M. Ward. 2001. The diversity and distribution of bacteria phylogenetically related to Chloroflexus in hypersaline microbial mats. Appl. Environ. Microbiol. 67: 4365­4371.
Madigan, M.T. 2002. Anoxygenic phototrophs from extreme environments. Photosyn. Res. In press.