Molecular phylogenetic analysis of a bacterial mat community, Le Grotte di Frasassi, Italy

Bess Koffman Senior Integrative Exercise

10 March 2004

Submitted in partial fulfillment of the requirements for a Bachelor of Arts degree

from Carleton College, Northfield, Minnesota

Table of Contents

Abstract…………………………………………………………………………………….i Keywords…………………………………………………………………………………..i Introduction………………………………………………………………………………..1 Methods Cave description and sampling……………………………………………………4 DNA extraction, PCR amplification, and cloning of 16S rRNA genes…………...4 Phylogenetic analysis……………………………………………………………...6 Results Sampling environment and mat structure…………………………………………8 Phylogenetic analysis of 16S rRNA genes derived from mat……………………..8 Discussion………………………………………………………………………………..10 Acknowledgments………………………………………………………………………..14 References Cited…………………………………………………………………………15

i

Molecular phylogenetic analysis of a bacterial mat community, Le Grotte di Frasassi, Italy

Bess Koffman

Carleton College Senior Integrative Exercise Advisor: Jenn Macalady

10 March 2004 Abstract The Frasassi Caves are a currently forming limestone karst system in which biogenic sulfuric acid may play a significant role. High concentrations of sulfide have been found in the Frasassi aquifer, and gypsum deposits point to the presence of sulfur in the cave. White filamentous microbial mats have been observed growing in shallow streams in Grotta Sulfurea, a cave at the level of the water table. A mat was sampled and used in a bacterial phylogenetic study, from which eleven 16S ribosomal RNA (rRNA) gene clones were sequenced. The majority of 16S clones were affiliated with the δ-proteobacteria subdivision of the Proteobacteria phylum, and many grouped with 16S sequences from organisms living in similar environments. This study aims to extend our knowledge of bacterial diversity within relatively simple geochemical environments, and improve our understanding of the biological role in limestone corrosion. Georef Keywords: Sulfide oxidation, sulfate reduction, caves, bacteria, phylogeny, karst

1

Introduction Karst is the name given to large-scale limestone porosity—caves, sinkholes, and

other effects of significant limestone dissolution. Karst usually occurs as water

percolates through limestone. The water dissolves calcium carbonate, CaCO3, and reacts

with it to form bicarbonate, HCO3-; eventually pockets and large cavities form. However,

karst processes are not as simple as this popular explanation. In many cases, geologists

have found evidence of sulfur in limestone karst cave systems—often as gypsum

crystallized on the cave walls (Northrup and Lavoie, 2001). Some of the largest

limestone caves in the world, including Cueva de la Villa Luz in Tabasco, Mexico and

Lechuguilla Cave of Carlsbad Caverns National Park in New Mexico, seem to have been

formed through processes involving both bicarbonate and biogenic sulfuric acid

(Cunningham et al., 1995; Northrup and Lavoie, 2001).

Sulfuric acid, as a strong acid, can dissolve limestone more effectively than can

bicarbonate or its protonated cousin, carbonic acid. Sulfuric acid forms through the

oxidation of hydrogen sulfide, H2S, to sulfate, SO42- (Fig. 1a). When sulfate reacts with

water, it forms sulfuric acid, H2SO4. When sulfuric acid reacts with calcium carbonate,

CaCO3, it makes gypsum, CaSO4 (Fig. 1b). Water and carbon dioxide released through

the formation of gypsum react with calcium carbonate to make bicarbonate or carbonic

acid (Fig. 1c) (Galdenzi et al., 1999).

The Frasassi cave in Italy is a well-known and unusual karst limestone system, in

that, unlike most other well-known limestone caves, it is currently undergoing formation.

Observations of microbial biofilms covering the limestone surfaces within the cave and

of macroscopic microbial mats floating in cave streams have led researchers to postulate

that H2S oxidation is the driving factor behind karstification (Galdenzi et al., 1999).

2

Understanding the role of microorganisms—in particular, sulfur and sulfide oxidizers and

sulfate reducers—in the Frasassi cave may unearth biogeochemical interactions with a

broader application to karst systems.

A valuable aspect of working in Frasassi is that the geochemistry is simpler than

in many other environments (such as a lake or soil sample), and microbial metabolism is

limited by the absence of sunlight and organic material from the surface. Stable isotope

ratio analysis conducted by Galdenzi et al. on organic samples found a distinct

fractionation signature with consistently lighter isotopes of carbon and nitrogen in the

sulfidic sections of the caves compared with those at the caves’ entrances and the surface

(1999). This suggests that all organic compounds present in the cave are autochthonous,

and is consistent with our picture of the cave as an ecosystem completely isolated from

surface nutrients and photosynthesis. For this reason Frasassi is an ideal environment

that allows us to see discrete links between in situ organisms and the cave’s

geochemistry.

Given the geochemical and metabolic requirements of organisms such as those in

Frasassi, it is difficult to follow the traditional microbiological method of culturing

organisms in the lab in order to study their morphologies. This technique places heavy

selective pressure on organisms, with the result that what grows is often out of proportion

with the original microbial mat. Angert et al. note that typically less than 1% of all

microbes from a particular environment can be grown in the lab using standard

enrichment techniques (1998). In particular, organisms that rely on symbiosis or

syntrophy, whereby a substance is catabolized through the metabolic efforts of two or

more organisms that could not catabolize the substance on their own, may not grow at all

3

in the lab (Madigan et al., 2003). These two relationships often produce microbial mats

with intricate layering according to molecule-scale gradients of substances such as

oxygen and hydrogen sulfide (Stal, 2001). Therefore, it is important to rely not only on

culturing but on molecular phylogenetic work to identify organisms and determine

evolutionary relationships.

Some genes change over time faster than others depending on the necessity of

their function to the organism. Genes that code for the most important functions in an

organism are highly conserved and thus are used for molecular work. The 16S rRNA

gene codes for ribosomal RNA, a molecule that plays an important structural and

functional role in the ribosome (Madigan et al., 2003). For this reason it is highly

conserved and is the gene most often used for phylogenetic work. Using divergences

within this one gene, researchers can create phylogenetic trees that show evolutionary

relationships. On these trees, internal nodes represent common ancestors, and the lengths

of branches correspond to evolutionary changes, or more specifically, base pair

substitutions per site. Widespread use of 16S rDNA sequences for phylogenetic analysis

has resulted in a systematic and reliable way to diagram evolutionary relationships and

survey microorganisms present in natural environments (Hall, 2001a).

This paper presents a molecular phylogenetic analysis of 16S rDNA from bacteria

growing in a cave stream microbial mat in the Frasassi Grotta Sulfurea. Results suggest

that the mat contains organisms closely related to known sulfur oxidizing and sulfur

reducing bacteria. These bacteria likely play a significant role in the production of

sulfuric acid and subsequent karstification.

4

Methods Cave description and sampling

Le Grotte di Frasassi (Frasassi Caves) are located in the Marche region of Italy,

approximately 150 km NE of Rome, in the Sentino river gorge near the village of Genga.

The cave, at about 200-360 m above sea level, includes four main layers within the

massive Calcare Massiccio limestone formation (Galdenzi and Maruoka, 2003). Frasassi

includes over 100 caves in a network of about 35 cave passages (Fig. 2). The highly

permeable Massiccio (which is 600-1000 m thick) rests on the Anidriti del Burano, an

evaporite formation estimated to be 2000 m thick (Galdenzi and Maruoka, 2003). This is

thought to be the sulfur source for the Frasassi caves (Galdenzi and Maruoka, 2003).

Sodium, chloride and sulfur ions (i.e., sulfate and sulfide) are brought to the cave by

groundwater (Fig. 3) (Galdenzi et al., 1999). Sulfide concentrations of up to 0.4 mmol/l

and sulfate concentrations of up to 2.5 mmol/l have been found in the Frasassi aquifer

(Galdenzi et al., 1999). The cave streams keep a constant temperature of 13˚ C and

range in depth from 10 cm to several meters, with flow rates ranging from 0.5 to 50 l/s

(Vlasceanu et al., 2000). Galdenzi et al. found that conductivity of the streams was 1200-

1900 µS and that limestone dissolution occurred both in the streams and in the air at a

rate of 50 µm/yr over the study period of 5 years (1999).

In 2002, Jenn Macalady1 collected samples (“GS tm”) from a macroscopic

filamentous white mat in a cave stream of Grotta Sulfurea, at about 200 m elevation (the

water table) (Fig. 4). Samples were frozen at -20˚ C until transport to the U.S. on dry ice.

Upon arrival in the laboratory, they again were stored at -20˚ C.

1 Assistant Professor, Department of Geology, Carleton College.

5

DNA extraction, PCR amplification, and cloning of 16S rRNA genes Total community DNA was extracted from approximately 30 g (wet wt) frozen

sample using a MoBio Ultraclean Mega Prep Soil DNA kit, following the manufacturer’s

procedure. Presence of DNA was confirmed using agarose gel electrophoresis (1% agar)

with a secondary ethidium bromide staining solution. 16S rRNA genes were amplified

using the polymerase chain reaction (PCR) with 8-27 forward (5´-

AGAGTTTGATCCTGGCTCAG) bacterial and 1510-1492 reverse (5´-

GGTTACCTTGTTACGACTT) universal primers (Fig. 5). The PCR cycle was as

follows: initial denaturation at 94° C for 5 min., denaturation at 94° C for 1 min.,

annealing at 45° C for 45 sec., extension at 72° C for 1.5 min., final extension at 72° C

for 20 min. and storage at 4° C; the PCR ran for 30 cycles. Tests were run against

archaeal and blank controls, and against bacterial DNA extracted from a local stream

microbial mat as a positive control. Clones were transformed with a TOPO TA Cloning®

Kit for Sequencing, according to the manufacturer’s procedure. I used chemically

competent Escherichia coli, which were cultured in Ampicillin LB broth overnight and

then inoculated onto agarose plates.

Colonies from these smear plates and from later streak plates were sampled

randomly. Plasmids were extracted from cells in each colony and 16S rDNA inserts

purified using a Wizard® Plus SV Minipreps DNA Purification System (Promega). I

checked for concentration using agarose gel electrophoresis with 3.6 mL culture,

following the manufacturer’s procedure (Fig. 6). PCR with 27f and 1492 primers was

used to amplify the inserts, with the following PCR conditions: initial denaturation at 4°

C for 10 min., denaturation at 94º C for 30 sec., annealing at 47º C for 1 min., extension

6

at 72º C for 2 min., final extension at 72º C for 10 min. and storage at 4º C; the PCR ran

for 30 cycles.

The 16S gene inserts were sequenced at the University of Wisconsin-Madison’s

Biotechnology Center, using cycle sequencing with thermostable polymerases and

fluorescently labeled dideoxy terminators. The primers used were T3 (5’-

ATTAACCCTCACTAAAGGGA) and T7 (5’-TAATACGACTCACTATAGGG).

Phylogenetic analysis

The T3 reverse-complement and T7 ends of each sequence were matched

manually using the multiple alignment program Se-Al (Rambaut, 1996). Sequences were

submitted to the Ribosomal Database Project’s (RDP) online CHECK_CHIMERA

software to test for possible PCR artifacts (sequences recombined from more than one

organism’s 16S gene) as mediated by Taq polymerase (Maidak et al., 1997). In order to

compare the BK clones with evolutionarily similar sequences, my advisor Jenn Macalady

and I searched the ARB 16S rDNA database, which contains over 5000 bacterial

sequences (Ludwig and Strunk, unpublished data). Sequences similar to each of the BK

clones were used for automatic and manual alignments in Seaview and BioEdit (Galtier

et al., 1996; Hall, 2001b; Ludwig and Strunk, unpublished data). Aligning sequences

well is crucial to creating reasonable phylogenies, as poor alignments can lead to

misleading or meaningless trees (Hall, 2001a). I used the Blast program on the website

for the National Center of Biotechnology Information (NCBI) to calculate percent

similarity between cultivated species and my clones (Altschul et al.).

Tree-building methods vary depending on their assumptions about genetic

evolution. Since we as scientists do not know exactly how evolution works, it is

7

important to use multiple methods to construct trees and analyze phylogenetic

relationships. I used both the neighbor-joining algorithm in ARB and Bayesian analysis

to analyze my sequences (Huelsenbeck and Hall, 2000; Ludwig and Strunk, unpublished

data). Neighbor-joining uses a distance matrix derived from a multiple alignment. The

matrix records the calculated distance, or fraction of differences, between each pair of

sequences or lineal groups. In this way it can give a value and a corresponding branch

length to the calculated evolutionary distance (Hall, 2001a). Neighbor-joining is the only

practical tree-building algorithm to use with large data sets, such as the one in ARB.

Large data sets are valuable because they provide a broad context for the placement of

new sequences.

My sequences were aligned using the ARB aligner, checked manually, and added

to the database using ARB’s “QuickAdd” parsimony method. Parsimony is a “minimum

change” method that assumes that the most likely tree is the one with the least changes

between related taxa (Hall, 2001a). It produces trees that show approximate placement of

the organisms and is useful when adding a limited number of sequences to a neighbor-

joining tree. The ARB tree was rooted using several archaeal species.

Trees specific to each bacterial lineage were created from the ARB libraries using

Bayesian inference in MrBayes and were formatted graphically using PAUP*

(Huelsenbeck and Hall, 2000; Swofford, 1999). Bayesian analysis uses posterior

probabilities to find the best set of trees for the given sequence alignment and a

postulated model of evolution that maximizes the probability of observing the given data.

Rather than arriving at a single best tree, Bayesian inference samples trees using the

Markov chain Monte Carlo method, which produces a set of converging consensus

8

trees—trees for which accepting or rejecting any particular change becomes essentially

random. The frequency of any particular tree being sampled is related to the probability

that it is the best tree for the data. Bayesian analysis is quite powerful and is becoming

increasingly popular among phylogeneticists (Hall, 2001a).

I used MrBayes to run Bayesian inference searches for 105 cycles with four

Markov chains to check for convergence. Trees were saved every 100 generations; the

first 350 trees from each cold chain (the Markov chain used for sampling) were

discarded. These were rooted using monophyletic bacterial outgroups.

Results Sampling environment and mat structure

On the sampling expedition of October 2003 to Grotta Sulfurea, Jenn Macalady

and I measured the geochemistry of the shallow flowing cave stream containing the white

microbial mat sampled in 2002. We found cave water temperature to be 13.9-14.6˚ C.

The water had pH = 7.0 ± 0.03. The white filamentous microbial mat was floating on or

below the surface of the stream, and in places was directly on top of the black sulfidic

mud that lined the streambed. The strong “rotten egg” smell of hydrogen sulfide was

quite recognizable.

Phylogenetic analysis of 16S rRNA genes derived from mat From the Grotta Sulfurea cave microbial mat samples, eleven 16S rRNA clones

were sequenced. None of the sequences were positively identified as chimeric using the

RDP software CHECK_CHIMERA. Two sequences (clones BK9 and BK111) were over

99.5% identical and so only one of these (BK9) was used for phylogenetic work. The

remaining ten sequences fit into four phyla, all in the Bacterial domain (Fig. 7):

Proteobacteria (6 sequences), Cytophaga/Flavobacteria/Bacteroides (CFB) (2

sequences), WS6 (1 sequence), and Termite Group 1 (1 sequence). Of the bacteria in the

9

Proteobacteria phylum, 1 fit in the γ-proteobacteria subdivision, and the remaining 5 in

the δ-proteobacteria subdivision (Table 1). Preliminary results of further phylogenetic

work on clones from cave wall microbial films in Grotta Sulfurea (“GSgm”) are also

shown on ARB trees (Figs. 8, 10, 12, and 14).

Clone BK1 fit well with the γ-proteobacterium “Thiobacillus baregensis,” a

sulfur oxidizing bacterium found in sulfurated thermal waters in Bareges, France (Hedoin

et al., unpublished 1996). It also grouped with a number of Sulphur River clones, clones

isolated from a white filamentous microbial mat in Sulphur River, Parker Cave, Kentucky

(Angert et al., 1998). Clone BK1 grouped most notably with clone SRang 2.5, which

itself grouped closely with “T.barengensis” (Figs. 8 and 9).

The clones BK4, BK5, BK7, BK9, and BK10 grouped in the δ-proteobacteria, a

lineage of predominantly sulfur-reducing bacteria (Figs. 10 and 11). Clones BK4 and

BK10 clustered together, with their closest cultivated neighbor being Desulfosarcina

singaporensis, a sulfate-reducing bacterium isolated from sulfide-rich black marine mud

(Lie et al., 1999). Clone BK5 was closely related to Geobacter sulfurreducens and to

Geobacter metallireducens, both metal ion reducers (Methe et al., 2003). Clone BK7 fit

near Desulfonema ishimotoei, a sulfate reducer isolated from organic-rich sulfidic marine

and freshwater sediment (Fukui et al., 1999). Clone BK9 was similar to Syntrophobacter

wolinii, a sulfate reducer isolated from a culture enriched from anaerobic granular sludge

(Harmsen et al., 1993).

Clone BK2 and a partial sequence of BK108 (T3) fit in the CFB group (Figs. 12

and 13). Clone BK2 was most similar to Cytophaga fermentans. Clone BK108T3 was

closely associated with the Sulphur River clone SRang 1.29, the contaminated aquifer

10

clone WCHB1.32, and the Cesspool Cave clone CC.8. The only cultivated bacteria in the

same group as clone BK108T3 were the fermenter Bacteroides putredinis and the termite

gut bacterium Rikenella microfusus, both anaerobes (Madigan et al., 2003; Ohkuma et al.,

2002).

Clone BK6 fit in the WS6 phylum, near the OP11 clone OPB92 (Figs. 14 and 15).

Clone BK8 fit in Termite Group 1 (Figs. 15 and 16). Since I had only a partial sequence

of Clone BK8 (T3), I could not do rigorous phylogenetic work on it. The phyla WS6 and

Termite Group 1 are newly defined and contain no cultivated species.

Discussion The macroscopic filamentous white microbial mat found in the cave stream of

Grotta Sulfurea represents a distinct community isolated from photosynthetically derived

carbon. The observed high sulfide and sulfate levels in the water and air are sufficient to

support a sulfide-metabolizing community (Madigan et al., 2003). Given the sequence

similarities of 88-94% between the Grotta Sulfurea BK clones and known sulfur/ide

oxidizers and sulfate reducers (Table 1), it is reasonable to believe that my 16S clone

sequences came from respective sulfur/ide oxidizers and sulfate reducers, and that the

microbial community in turn includes both these types of bacteria. Since there is no input

of organic material from the surface to Grotta Sulfurea, chemolithoautotrophy likely

forms the base of the food chain. Sulfur oxidizing autotrophic bacteria thus likely

comprise a significant portion of the base-level biomass, a biomass capable of supporting

fifteen invertebrate species, of which seven are endemic to Frasassi (Galdenzi et al.,

1999).

A sulfur-rich environment such as Frasassi would be expected to support both

types of metabolism—especially where sulfur/ide oxidizers and sulfate reducers may

11

form close symbiotic relationships. This has been seen in hydrothermal vent

communities, where bacteria (especially sulfur/ide oxidizers and sulfate reducers) play an

important role in mineralization of vent chimneys (Lovley et al., 2000; McCollom and

Shock, 1997). In fact, sulfur isotope ratios analyzed by Galdenzi and Maruoka (2003)

suggest that sulfate reducing bacteria are present in Frasassi. They found sulfur isotope

values of δ34S ~ -19.60 ‰ in gypsum deposits in the cave, δ34S ~ -14 in H2S and δ34S ~

+21.00 ‰ in sulfate within sulfidic groundwater (Galdenzi and Maruoka, 2003).

Galdenzi and Maruoka (2003) propose that the size of these isotope fractionations can

best be explained by the sulfur cycling through sulfate reducing bacteria, which are

known to produce high fractionations. The presence of sulfate reducers, shown through

phylogenetic work, fits with the sulfur isotope analysis.

Not all clones from similar microbial mats in sulfidic streams have this split

between sulfur/ide oxidation and sulfate reduction. The Sulphur River clones from a

white filamentous mat in Parker Cave, Kentucky are predominantly ε- and γ-

proteobacteria, both lineages known for their sulfur-oxidizing metabolism (Angert et al.,

1998). Similarly, clones from a white filamentous microbial mat in the sulfidic waters of

Lower Kane Cave, Wyoming are almost entirely from the ε-proteobacteria (Engel et al.,

2003). It would seem that these results would be more expected for the Frasassi Grotta

Sulfurea BK clones given the similarities in limestone geology and microbial mat

morphology. There was one γ-proteobacteria, but the majority of clones (5) were δ-

proteobacteria, putative sulfate reducers.

The differences between the Frasassi Grotta Sulfurea clones and those of Sulphur

River and Lower Kane Cave may be due to a difference in sampling technique or an

12

introduced experimental bias. In the DNA extraction there may be lysing bias, whereby

certain cells are more easily opened and their DNA extracted. Similarly, the PCR

primers may anneal with certain sequences better than others. This latter potential source

of difference can be eliminated entirely from the Lower Kane Cave clones, as they were

PCR amplified using the same primers as for the BK clones (27f, 1492r) (Engel et al.,

2003). The Sulphur River clones were amplified using a 515 forward primer and the

same1492 reverse primer (Angert et al., 1998). It is also possible that differences in the

PCR programs used could have affected the results.

Even considering these possible sources of disparity between the Grotta Sulfurea

BK clones and the Sulphur River and Lower Kane Cave clones, a difference between the

phylogenetic results found in each location is evident. Why were there no sulfate

reducers found in either the Sulphur River or the Lower Kane Cave clones? Since sulfur

oxidizers produce sulfate, it would seem that their presence would provide the sulfate

necessary for sulfate reduction metabolism. This apparent difference between the sulfur-

cycling species present in each environment needs to be investigated further.

The presence of only Bacteria in my clone libraries can be attributed to the use of

bacterial PCR primers, which amplify only the bacterial 16S rRNA gene. We would

need to use archaeal primers and sequence a greater number of bacterial clones in order

to explore the full diversity of the Grotta Sulfurea stream microbial mat. Since 16S genes

evolve slowly relative to most other genes, they often do not keep up with changes in the

genes that code for proteins such as those involved in metabolism (Woese, 1987). It

would therefore be germane to sequence protein-coding genes—and even better, entire

genomes. Another important and ongoing aspect of this research is Fluorescence In-Situ

13

Hybridization (FISH), which will help us visualize the population numbers and

distribution of organisms between the two prokaryotic domains.

This study expands the known diversity of the Frasassi Grotta Sulfurea cave

stream microbial mat community based on phylogenetic analysis, and is an excellent

launching point for further physiological and phylogenetic study. The high percent

similarities between my 16S clones and the 16S genes of known sulfur/ide oxidizers and

sulfate reducers point to a biological source of sulfuric acid through biogenic sulfur

oxidation, which implies a crucial role played by bacteria in carving the Frasassi caves.

The presence of both sulfur/ide oxidizers and sulfate reducers fits with sulfur isotope

fractionation data and testifies to the complex sulfur cycle of the Frasassi cave system.

14

Acknowledgments Alessandro Galdenzi accompanied Jenn Macalady and me to Grotta Sulfurea in

October 2003, and for this I am grateful. Alessandro Montanari kindly shipped our

samples from Italy. The computer analysis work would have been impossible without the

help of Doug Foxgrover, who worked hard to install ARB on Jenn’s computer. Leah

Morgan answered my many questions about computer programs and let me use her

computer for BioEdit. My academic advisor Bereket Haileab provided moral support and

perspective. Cam Davidson tutored me in Adobe Illustrator. Tim Vick offered constant

technical support. Funding for this project came from the Bernstein Student Research

Endowment and the NSF Biogeosciences Program. This work builds on research done by

Jenn Macalady, Alessandro Galdenzi2, and Teruyuki Maruoka,3 among others. Finally, I

would like to thank Jenn Macalady, my advisor on this project. Without her vision,

expertise, and support, this research would not have happened.

2 Instituto Italiano di Speleologia, Frasassi Section. 3 Department of Geological Sciences, University of Vienna.

15

References Cited Altschul, S. F., Madden, T. L., Schaffer, A. A., Zhang, J., Zhang, Z., Miller, W., and

Lipman, D. J., Blast 2.04 of the NCBI Toolkit. Angert, E. R., Northup, D. E., Reysenbach, A.-L., Peek, A. S., Goebel, B. M., and Pace,

N. R., 1998, Molecular phylogenetic analysis of a bacterial community in Sulphur River, Parker Cave, Kentucky: American Mineralogist, v. 83, p. 1583-1592.

Cunningham, K. I., Northup, D. E., and Pollastro, R. M., 1995, Bacteria, Fungi, and Biokarst in Lechuguilla Cave, Carlsbad-Caverns-National-Park, New-Mexico: Environmental Geology, v. 25, no. 1, p. 208.

Engel, A. S., Lee, N., Porter, M. L., Stern, L. A., Bennett, P. C., and Wagner, M., 2003, Filamentous "Epsilonproteobacteria" dominate microbial mats from sulfidic cave springs: Applied and Environmental Microbiology, v. 69, no. 9, p. 5503-5511.

Fukui, M., Teske, A., Assmus, B., Muyzer, G., and Widdel, F., 1999, Physiology, phylogenetic relationships, and ecology of filamentous sulfate-reducing bacteria (genus desulfonema): Arch. Microbiol., v. 172, p. 193-203.

Galdenzi, S., and Maruoka, T., 2003, Gypsum deposits in the Frasassi Caves, central Italy: Journal of Cave and Karst Studies, v. 65, no. 2, p. 111-125.

Galdenzi, S., Menichetti, M., Sarbu, S., and Rossi, A., 1999, Frasassi caves: a biogenic hypogean karst system?: Colloque europeen-Karst 99-European Conference, p. 101-106.

Galtier, N., Gouy, M., and Gautier, C., 1996, SEAVIEW and PHYLO_WIN: two graphic tools for sequence alignment and molecular phylogeny.: Comput Appl Biosci, v. 12, no. 6, p. 543-548.

Hall, B. G., 2001, Phylogenetic Trees Made Easy: A How-To Manual for Molecular Biologists: Sunderland, M.A., Sinauer Associates, Inc.

Hall, T., 2001, BioEdit: Raleigh, N.C., North Carolina State University. Harmsen, H. J., Wullings, B., Akkermans, A. D., Ludwig, W., and Stams, A. J., 1993,

Phylogenetic analysis of Syntrophobacter wolinii reveals a relationship with sulfate-reducing bacteria: Arch. Microbiol., v. 160, no. 3, p. 238-240.

Hedoin, H., Dauga, C., Abadie, M., Kaiser, P., and Cerning, J., unpublished 1996, Candidatus Thiobacillus baregensis a new sulfoxidizing bacteriumfrom sulfurated thermal waters of Bareges (France). Applied and Environmental Microbiology, v. In press (unpublished as of 10 March 2004).

Huelsenbeck, J. P., and Hall, B., 2000, MrBayes: A program for the Bayesian inference of phylogeny: Rochester, N.Y., University of Rochester.

Invitrogen life technologies, 2003, TOPO TA Cloning Kit for Sequencing, Version K: Invitrogen Corporation.

Komatsoulis, G. A., and Waterman, M. S., 1997, A new computational method for detection of chimeric 16S rRNA artifacts generated by PCR amplification from mixed bacterial populations: Applied and Environmental Microbiology, v. 63, no. 6, p. 2238-2346.

Lie, T. J., Clawson, M. L., Godchaux, W., and Leadbetter, E. R., 1999, Sulfidogenesis from 2-aminoethanesulfonate (taurine) fermentation by a morphologically unusual sulfate-reducing bacterium, Desulforhopalus singaporensis sp. nov.: Applied and Environmental Microbiology, v. 65, p. 3328-3334.

Ludwig, and Strunk, unpublished data, ARB: Muenchen.

16

Madigan, M. T., Martinko, J. M., and Parker, J., 2003, Brock Biology of Microorganisms: Upper Saddle River, NJ, Pearson Education, Inc.

Maidak, B. L., Olsen, G. J., Larsen, N., Overbeek, R., McCaughey, M. J., and Woese, C. R., 1997, The RDP (Ribosomal Database Project): Nucleic Acids Research, v. 25, no. 1, p. 109-111.

Methe, B. A., Nelson, K. E., Eisen, J. A., Paulsen, I. T., Nelson, W., Heidelberg, J. F., Wu, D., Wu, M., Ward, N., Beanan, M. J., Dodson, R. J., Madupu, R., Brinkac, L. M., Daugherty, S. C., DeBoy, R. T., Durkin, A. S., Gwinn, M., Kolonay, J. F., Sullivan, S. A., Haft, D. H., Selengut, J., Davidsen, T. M., Zafar, N., White, O., Tran, B., Romero, C., Forberger, H. A., Weidman, J., Khouri, H., Feldblyum, T. V., Utterback, T. R., Van Aken, S. E., Lovley, D. R., and Fraser, C. M., 2003, Genome of Geobacter sulfurreducens: Metal reduction in subsurface environments: Science, v. 302, no. 5652, p. 1967-1969.

Northrup, D., and Lavoie, K., 2001, Geomicrobiology of Caves: A Review: Geomicrobiology Journal, v. 18, p. 199-222.

Ohkuma, M., Noda, S., Hongoh, Y., and Kudo, T., 2002, Diverse bacteria related to the bacteroides subgroup of the CFB phylum within the gut symbiotic communities of various termites: Biosci Biotechnol Biochem., v. 66, no. 1, p. 78-84.

Rambaut, A., 1996, Se-Al: Sequence Alignment Editor: Oxford, UK. Stal, L. J., 2001, Coastal microbial mats: the physiology of a small-scale ecosystem:

South African Journal of Botany, v. 67, no. 3, p. 399-410. Swofford, D. L., 1999, PAUP*. Phylogenetic Analysis Using Parsimony (*and Other

Methods). Version 4.: Sunderland, M.A., Sinauer Associates. Vlasceanu, L., Sarbu, S., Engel, A., and Kinkle, B., 2000, Acidic Cave-Wall Biofilms

Located in the Frasassi Gorge, Italy: Geomicrobiology Journal, v. 17, p. 125-139.

(a) H2S + 2 O2 2 H+ + SO42-

(b) 2 H+ + SO42- + CaCO3 + H2O CaSO4 * 2 H2O + CO2

(c) CO2 + CaCO3 + H2O Ca2+ + 2 HCO3-

Figure 1. Various pathways for limestone corrosion: (a) oxidation of sulfide tosulfate, which can bond with water to form sulfuric acid, H2SO4; (b) formation of gypsum, which frees CO2 to form bicarbonate; (c) typical karst-style formation

of bicarbonate. This can become protonated to form carbonic acid, H2CO3.

Adapted from Galdenzi et al., 1999.

17

Grotta SulfureaSentino River

Grotta Grande del Vento

Grotta del Fiume

Ancona*

Firenze

Roma

Napoli

Artificial accesstunnel

N

250 m0

Adriatic Sea

Tyrrhenian Sea

LegendSampling site

45°10° 16°

41°

Figure 2. Map of Frasassi Caves showing sampling site in Grotta Sulfurea. Adapted from Vlasceanu, et al., 2000.

18

Figure 3. Schematic cross section of Frasassi Gorge showing water table and movement of ground water. Adapted from Galdenzi et al., 1999.

~100 m elevation

19

Figure 4. Photographs from Frasassi Gorge and Caves: a) entrance to Grotta Sulfurea from SS 76 in Genga; b) Sandro Galdenzi lowering the rope into the first chute of Grotta Sulfurea; c) sampling the white mat; d) white mat (scale bar is about 0.3 m); e) measuring pH of drips from wall gypsum crystals. Photos by Jenn L. Macalady.

a)

b)

c)

e) d)

20

Clo

ne n

ame

Div

isio

n (S

ubdi

visi

on)

Nea

rest

taxo

n* (

% id

entit

y**)

Infe

rred

phy

siol

ogy

BK

1Pr

oteo

bact

eria

(ga

mm

a)"T

hiob

acill

us b

areg

ensi

s"

(94%

)Su

lfur

/ide

oxid

izer

BK

4Pr

oteo

bact

eria

(de

lta)

Des

ulfo

rhop

alus

sin

gapo

rens

is

(93%

)Su

lfat

e re

duce

rB

K5

Prot

eoba

cter

ia (

delta

)G

eoba

cter

sul

furr

educ

ens

(95

%)

Sulf

ate

redu

cer

BK

7Pr

oteo

bact

eria

(de

lta)

Des

ulfo

nem

a is

him

otoe

i (8

8%)

Sulf

ate

redu

cer

BK

9 (B

K11

1)Pr

oteo

bact

eria

(de

lta)

Synt

roph

obac

ter

wol

inii

(93%

)Su

lfat

e re

duce

rB

K10

Prot

eoba

cter

ia (

delta

)D

esul

forh

opal

us s

inga

pore

nsis

(9

4%)

Sulf

ate

redu

cer

BK

2C

FBC

ytop

haga

ferm

enta

ns

(88%

)A

naer

obe

BK

108T

7 (p

artia

l)C

FBSu

lphu

r R

iver

clo

ne S

Ran

g 1.

29 (

92%

)A

naer

obe

BK

6W

S6O

P11

clon

e O

PB92

(79

%)

BK

8T3

(par

tial)

Ter

mite

Gro

up 1

Ter

mite

gut

clo

ne T

G13

(79

%)

Ana

erob

e

*Nea

rest

cul

tivat

ed ta

xon

foun

d in

pub

lic d

atab

ases

**Pe

rcen

t ide

ntity

cal

cula

ted

usin

g nu

mbe

r of

bas

e pa

ir m

atch

es p

er a

lignm

ent i

n th

e N

CB

I B

last

pro

gram

(A

ltsch

ul, S

.F.,

et a

l.)

Tab

le 1

. Su

mm

ary

of 1

6S r

DN

A r

esul

ts.

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

----

---

24

LacZ

α-cc

dB

Kan

amyc

in

Am

pici

llin

pUC

ori

P lac

CA

CA

CA

GG

AA

A

CA

GC

TATG

A

CC

ATG

ATT

AC

G

CC

AA

GC

TCA

G

AA

TTA

AC

CC

TC

AC

TAA

AG

GG

TGTG

TCC

TT

TGTC

GA

TAC

TG

GTA

CTA

ATG

C

GG

TTC

GA

GT

CTT

AA

TTG

GG

A

GTG

ATT

TCC

GA

CTA

GTC

CT

GC

AG

GTT

TAA

A

CG

AA

TTC

GC

C

CTT

G

GG

C

GA

ATT

CG

CG

GC

TGA

TCA

GG

A

CG

TCC

AA

ATT

TG

CTT

AA

GC

G

GG

A

TT

CC

CG

C

TTA

AG

CG

CC

CC

GC

TAA

ATT

C

AA

TTC

GC

CC

TA

TAG

TGA

GT

CG

TATT

AC

AA

TT

CA

CTG

GC

C

GTC

GTT

TTA

CG

GC

GA

TTTA

A

GTT

AA

GC

GG

G

ATA

TCA

CTC

A

GC

ATA

ATG

TT

AA

GTG

AC

CG

G

CA

GC

AA

AA

TG

A A

PC

R

Pro

duct

T3

prim

ing

site

T7

prim

ing

site

Figu

re 5

. M

ap o

f pC

R®

4-

TOPO

® pl

asm

id, s

how

ing

nucl

eotid

e se

quen

ce su

rrou

ndin

g th

e cl

onin

g si

te w

ith b

ase

pair

num

bers

giv

en o

n th

e le

ft.

Plas

mid

con

tain

s 395

4 ba

se p

airs

w

ithou

t the

PC

R p

rodu

ct.

Ada

pted

fr

om In

vitro

gen

(200

3).

201

261

311

Eco

R I

Eco

R I

Pm

e I

Pst

I

LacZ

α in

itiat

ion

codo

n

21

1550 bp

1400 bp

Ladde

r

BK1BK2

BK6BK7

BK8BK10

BK1 1:10

BK2 1:10

BK6 1:10

BK7 1:10

BK8 1:10

BK10 1:

10

Water b

lank

Bacteri

al co

ntrol

Culture

blan

k

BK108

BK111BK111

1:10

BK108 1

:10

Figure 6. Gel of PCR product (27f, 1492r) showing BK clones from plasmid prep template at full concentration and 1:10 dilution. I used 10 µl Hi-Lo DNA Marker as my ladder, with 5 µl DNA template per well. Note size of band at ~1500 bp. Clones BK4, BK5, and BK9 were amplified during a later PCR run and are not shown.

22

TM

7 O

P10

Dei

noco

ccus

-The

rmus

Chl

orof

lexi T

herm

otoga

e

The

rmod

esul

foba

cter

ia

Aqu

ifica

e

Cyanobacteri

a

Actinobact

eria

OP9 Firmicutes

Spirochaetes

Planctomycetes

OP3 Chlamydiae Verrucomicrobia

Defe

rriba

cteres

Synerg

istes

Fusobacteria

TM

6 Ter

mite

grou

p I

Aci

doba

cter

ia

Fib

roba

cter

es

Gem

mim

onas

Nitr

ospi

ra C

hlor

obi

ε Pro

teob

acte

ria

Euca

rya

FCG2 Korarchaeota

FCG1

Thermoprotei C2

YNPFFA

C1

Sedim

ent-1

Halo

bacte

ria p

MC2

Met

hano

bact

eria

Met

hano

cocc

ales A

rcha

eogl

obi

Met

hano

pyru

s T

herm

ococ

ci

SA

GM

EG1

pM

C1

The

rmop

lasm

ata

Met

hano

mic

robi

a

0.10

nuc

leot

ide

chan

ges p

er si

te

Bac

teria

Arc

haea

Bac

tero

idet

es δ

Prot

eoba

cteria

α, β

, γ P

roteo

bacte

ria

Figu

re 7

. Ph

ylog

enet

ic tr

ee sh

owin

g th

e th

ree

dom

ains

of l

ife.

Hig

hlig

hted

ph

yla

deno

te c

lade

s in

the

Bac

teria

th

at c

onta

in B

K c

lone

s. C

andi

date

ph

yla

Term

ite G

roup

1 a

nd W

S6 n

ot

show

n. T

ree

adap

ted

from

Mac

alad

y an

d B

anfie

ld, 2

003.

23

Pseu

dom

onad

acea

e 5

6

Terr

edin

ibac

ter s

peci

es, Z

3165

8, 1

494

Hal

omon

adac

eae

24

Oce

anos

piril

lum

min

utul

um, 1

500

Oce

anos

piril

lace

ae 9

Flec

toba

cillu

s sp.

, U14

584,

147

0 M

arin

obac

ter h

ydro

carb

onoc

last

icus

, X67

022,

146

1 O

cean

ospi

rillu

m ja

poni

cum

, 149

8 SA

R86

7

Sym

bion

ts 3

acid

min

e dr

aina

ge c

lone

MIM

4138

, 148

4 C

oxie

lla b

urne

tii, M

2129

1, 1

471

Rhi

pice

phal

us sa

ngui

neus

sym

bion

t, D

8455

9, 1

460

acid

min

e dr

aina

ge c

lone

MIM

451,

148

5

Mor

axel

lace

ae 3

1

Fran

cise

llace

ae 7

Psiri

cket

tsia

Sym

bion

ts 7

Leuc

othr

ix m

ucor

, X87

277,

150

5 Th

ioth

rix 6

Legi

onel

lale

s 8

4

Xan

thom

onad

ales

13

GSg

m c

lone

EL5

8G

Sgm

clo

ne E

L9G

Sgm

clo

ne E

L2G

Sgm

clo

ne E

L54

GSg

m c

lone

EL7

0G

Sgm

clo

ne E

L8 GSg

m c

lone

EL4

2en

viro

nmen

tal c

lone

gro

up 8

PCB

8con

tam

inat

ed so

il cl

one

WD

284,

AJ2

9267

5, 1

489

rape

rhiz

osph

ere

clon

e w

r000

8, A

J295

469,

134

1 Th

iotri

chac

eae

5co

al ta

r8co

ntam

inat

ed g

roun

dwat

er c

lone

482

5, A

F351

230,

144

7 ac

id m

ine

drai

nage

clo

ne M

IM44

4, 1

467

"Thi

obac

illus

ferr

ooxi

dans

" stra

in m

81 (B

G),

1329

G

STM

clo

ne B

K1

SRan

g cl

one

1.33

, AF0

4761

9, 9

61SR

ang

clon

e 1.

28, A

F047

617,

963

GSg

m c

lone

EL5

GSg

m c

lone

EL7

7"T

hiob

acill

us b

areg

ensi

s", Y

0928

0, 1

405

SRan

g cl

one

2.5,

AF0

4762

4, 9

61C

ardi

obac

teria

les

3Fr

asas

si sn

ottit

e is

olat

e 1C

("Th

ioba

cillu

s sp.

"), A

F213

054,

629

Thio

baci

llus s

p., X

9753

4, 1

446

Hal

othi

obac

illus

kel

lyi,

AF1

7041

9, 1

494

Thio

baci

llus h

ydro

ther

mal

is, M

9066

2, 1

504

Met

hylo

cocc

ales

20

Cyc

locl

astic

us p

uget

ii, U

1262

4, 1

492

Cyc

locl

astic

us p

uget

ii, L

3495

5, 1

492

Thio

mic

rosp

ira 5

Met

hylo

phag

a 5

Sole

mya

reid

i sym

bion

t 2, L

2570

9, 1

407

Sym

bion

ts 8

Ach

rom

atiu

m o

xalif

erum

, L42

543,

147

9 M

ethy

loco

ccus

4aq

uacu

lture

pon

d cl

one

SF40

.049

, 149

8 sy

mbi

ont o

f Sol

emya

vel

um, M

9041

5, 1

459

Sole

mya

occ

iden

talis

gill

sym

bion

t, U

4104

9, 1

374

acid

min

e dr

aina

ge c

lone

MIM

486,

149

5 "T

hiob

acill

us p

rosp

erus

", un

publ

is, 1

419

Ecto

thio

rhod

ospi

ra sh

apos

hnik

ovi,

M59

151,

145

1 C

hrom

atia

ceae

6

Ecto

thio

rhod

ospi

race

ae 4

Nitr

osoc

occu

s oce

anus

, M96

395,

139

2 N

itros

ococ

cus o

cean

us ,

M96

398,

134

7

Figu

re 8

. Pa

rtial

AR

B n

eigh

bor-j

oini

ng tr

ee sh

owin

g re

lativ

e po

sitio

n of

clo

ne B

K1.

Tree

con

tain

s gam

ma-

prot

eoba

cter

ial s

eque

nces

and

is ro

oted

with

a m

onop

hyle

tic a

rcha

eal o

utgr

oup.

N

umbe

rs o

n tre

e ar

e G

enba

nk a

cces

sion

num

bers

for o

rgan

ism

s in

the

publ

ic d

ata

base

s.

0.10

subs

titut

ions

/site

25

Thiobacillus hydrothermalis

BK1 clone

Sulphur river clone 2.5

Beggiatoa alba

Acid mine drainage clone MIM444

"Thiobacillus ferooxidans"

Agrobacterium tumefaciens0.05 substitutions/site

100

100

96

95

Figure 9. Bayesian inference phylogram showing relative position of clone BK1. Tree rooted with Agrobacterium tumefaciens. Bootstrap values displayed as percentages of 1000 replications.

"Thiobacillus baregensis"

26

Des

ulfu

rom

onad

acea

e/Pe

loba

cter

11

"Geo

bact

er c

hape

lleii"

, U41

561,

146

5 Pe

loba

cter

pro

pion

icus

, X70

954,

154

9 G

STM

clo

ne B

K5

Geo

bact

er su

lfurr

educ

ens,

U13

928,

140

8 "G

eoba

cter

hyd

roge

noph

ilus"

, U46

860,

143

8 "G

eoba

cter

hyd

roge

noph

ilus"

, U28

173,

139

1 G

eoba

cter

met

allir

educ

ens,

L078

34, 1

537

sulfa

te re

duci

ng b

iore

acto

r clo

ne S

RB

AC

37, 1

477

GST

M c

lone

BK

10G

STM

clo

ne B

K4

GSg

m c

lone

EL3

GSg

m c

lone

EL6

2co

ntam

inat

ed a

quife

r clo

ne W

CH

B17

67, A

F050

536,

148

1 D

esul

fofu

stis

gly

colic

us, X

9970

7, 1

545

aqua

cultu

re p

ond

clon

e SF

39.4

58, 1

444

Des

ulfo

rhop

alus

sing

apor

ensi

s, A

F118

453,

149

8 aq

uacu

lture

pon

d cl

one

SF40

.063

, 150

4 D

esul

forh

opal

us v

acuo

latu

s, L4

2613

, 148

2 G

Sgm

clo

ne E

L28

Des

ulfo

bulb

us e

long

atus

, X95

180,

145

3 D

esul

fobu

lbus

rhab

dofo

rmis

, U12

253,

149

8 Sy

ntro

phac

eae

4G

Sgm

clo

ne E

L21

GSg

m c

lone

EL4

9D

esul

fom

onile

4D

esul

foar

culu

m b

aars

ii, M

3440

3, 1

355

Des

ulfo

arcu

lus s

p. B

G74

, U85

477,

151

5 G

Sgm

clo

ne E

L4D

esul

foba

cter

ium

ani

lini,

AJ2

3760

1, 1

527

Des

ulfo

bact

eriu

m sp

. mX

yS1,

AJ0

0685

3, 1

528

GSg

m c

lone

EL3

8ho

t spr

ing

clon

e O

PB16

, AF0

2699

8, 1

490

hot s

prin

g cl

one

OPB

55, A

F026

993,

149

5 D

esul

foba

cter

hal

otol

eran

s, Y

1474

5, 1

472

Des

ulfo

bact

er sp

ecie

s, M

3441

6, 1

436

Des

ulfo

bact

er sp

ecie

s, M

3441

5, 1

373

Des

ulfo

bact

er h

ydro

geno

philu

s, M

3441

2, 1

358

Des

ulfo

bact

er c

urva

tus,

M34

413,

134

4 D

esul

foba

cter

vib

riofo

rmis

, U12

254,

151

5 D

esul

foba

cter

pos

tgat

ei, M

2663

3, 1

499

delta

pro

teob

acte

rium

WN

, U51

844,

137

2 D

esul

foba

cula

tolu

olic

a, X

7095

3, 1

550

"Des

ulfo

bact

eriu

m n

iaci

ni",

U51

845,

142

7 "D

esul

foba

cter

ium

vac

uola

tum

", M

3440

8, 1

345

Des

ulfo

bact

eriu

m a

utot

roph

icum

, M34

409,

131

9 G

STM

clo

ne B

K7

Des

ulfo

cocc

us m

ultiv

oran

s, M

3440

5, 1

347

Des

ulfo

nem

a lim

icol

a, U

4599

0, 1

496

GSg

m c

lone

EL2

2D

esul

fosa

rcin

a va

riabi

lis, M

2663

2, 1

499

Des

ulfo

sarc

ina

varia

bilis

, M

3440

7, 1

504

GSg

m c

lone

EL7

5D

esul

fofa

ba g

elid

a, A

F099

063,

148

4 D

esul

fobo

tulu

s sap

ovor

ans,

M34

402,

133

9 D

esul

fone

ma

ishi

mot

oei,

U45

991,

144

0 D

esul

fone

ma

ishi

mot

oei,

U45

992,

144

0 D

esul

fobo

tulu

s sp.

BG

14, U

8547

0, 1

500

Des

ulfo

cella

hal

ophi

la, A

F022

936,

154

4 G

STM

clo

ne B

K9

Des

ulfo

virg

a ad

ipic

a, A

J237

605,

152

4 D

esul

forh

abdu

s am

nige

nus,

X83

274,

151

5 un

iden

tifie

d su

lfate

7red

ucin

g ba

cter

ium

, U49

429,

151

8 U

nkno

wn,

X82

875,

147

5 U

nkno

wn,

X82

874,

148

6 Sy

ntro

phob

acte

r wol

inii,

X70

905,

142

4 D

esul

foac

inum

infe

rnum

, L27

426,

152

6 ho

t spr

ing

clon

e O

PB33

, AF0

2698

7, 1

486

Ther

mod

esul

forh

abdu

s nor

vegi

cus,

U25

627,

155

9 sl

udge

clo

ne H

16, A

F234

748,

155

5 ac

id m

ine

drai

nage

clo

ne M

IM41

8, 1

520

Myx

ococ

cale

s 1

2

Bde

llovi

brio

nale

s 2

TM6

9

Aci

doba

cter

ia 8

9

acid

min

e dr

aina

ge c

lone

MIM

4123

, 151

6 ac

id m

ine

drai

nage

clo

ne M

IM44

0, 1

533

Des

ulfo

vibr

iona

les

44

slud

ge c

lone

195

9, A

F097

803,

147

9 la

b7sc

ale

slud

ge c

lone

SB

R10

93, A

F269

002,

148

5 co

ntam

inat

ed a

quife

r clo

ne W

CH

B17

27, A

F050

538,

148

5

delta

-pro

teob

acte

ria

Figu

re 1

0. P

artia

l AR

B n

eigh

bor-j

oini

ng tr

ee sh

owin

g re

lativ

e po

sitio

ns o

f clo

nes B

K4,

BK

5, B

K7,

BK

9, a

nd

BK

10 a

s mem

bers

of t

he d

elta

-pro

teob

acte

ria.

Tree

root

ed w

ith m

onop

hyle

tic a

rcha

eal o

utgr

oup.

Num

bers

show

n ar

e G

enba

nk a

cces

sion

num

bers

for o

rgan

ism

s in

the

publ

ic d

atab

ases

.

0.10

subs

titut

ions

/site

27

Desulfobacter postgatei

Syntrophobacter wolinii

BK9 clone

Desulfosarcina sp.

Desulforhopalus singaporensis

BK4 clone

BK10 clone

Desulfomonile tiedjei

Nitrospina gracilis

Desulfobacterium anilini

Geobacter metaloreducens

Geobacter sulfurreducens

BK5 clone

Desulfonema ishimotoei

BK7 clone

0.05 substitutions/site

Figure 11. Bayesian inference phylogram showing relative positions of clones BK4, BK5, BK7, BK9, and BK10. Tree rooted with Desulfobacter postgatei. Bootstrap values displayed as percentages of 100 replications.

100

100

100

100

94

59

78

100

94

100

51

90

28

Bac

tero

ides

, Pre

vote

lla 6

9

Porp

hyro

mon

adac

eae

13

Bac

tero

ides

dis

taso

nis,

M25

249

M, 1

511

Bac

tero

ides

mer

dae,

X83

954,

136

6 su

lfate

redu

cing

bio

reac

tor c

lone

SR

BA

C29

, 144

9 B

acte

roid

es fo

rsyt

hus,

L164

95, 1

455

Bac

tero

ides

fors

ythu

s, X

7396

2, 1

450

CD

C G

roup

DF.

3, U

4135

5, 1

467

sulfa

te re

duci

ng b

iore

acto

r clo

ne S

RB

AC

55, 1

446

Bac

tero

ides

sp. s

p4, A

B00

3389

, 143

2 M

arin

ilabi

lia a

garo

vora

ns, M

6242

2, 1

454

Cyt

opha

ga fe

rmen

tans

, M58

766,

146

3 B

acte

roid

es sp

lanc

hnic

us, L

1649

6, 1

458

G

STM

clo

ne B

K2

7co

ntam

inat

ed a

quife

r clo

ne W

CH

B1.

32, A

F050

543,

144

0 SR

ang

clon

e 1.

29, A

F047

618,

944

GST

M c

lone

108

T3C

essp

ool c

lone

CC

.8, A

F207

533,

626

"Ana

erof

lexu

s mar

itim

us",

1386

co

ntam

inat

ed a

quife

r clo

ne W

CH

B1.

53, A

F050

539,

144

8

anae

robi

c di

gest

or c

lone

vad

inH

A17

, U81

712,

145

2 G

Sgm

clo

ne E

L26

cont

amin

ated

aqu

ifer c

lone

WC

HB

1.69

, AF0

5056

7, 1

445

GSg

m c

lone

EL2

0an

aero

bic

dige

stor

clo

ne B

C27

, U81

676,

144

8

sulfa

te re

duci

ng b

iore

acto

r clo

ne S

RB

AC

60, 1

446

sulfa

te re

duci

ng b

iore

acto

r clo

ne S

RB

AC

32, 1

446

cont

amin

ated

aqu

ifer c

lone

WC

HB

1.29

, AF0

5054

4, 1

446

Bac

tero

ides

put

redi

nis,

L164

97, 1

448

R

iken

ella

mic

rofu

sus,

L164

98, 1

463

Flav

obac

teria

les

64

Flex

ibac

tera

ceae

9

Sphi

ngob

acte

riace

ae 9

cool

ing

coil

colo

nise

r iso

late

PC

5.1A

, 138

1 Sp

iroso

ma

lingu

ale,

M27

802

M, 1

423

Flex

ibac

ter e

lega

ns ,

1432

Fl

exib

acte

r ele

gans

, M

5878

3, 1

432

Mic

rosc

illa

mar

ina,

M58

793,

142

5 Fl

exib

acte

r rub

er, M

5878

8, 1

450

aero

bic

slud

ge c

lone

PH

OS.

HE2

1, A

F314

419,

141

7 R

unel

la sl

ithyf

orm

is, M

2779

9 M

, 144

4 Fl

exib

acte

r lito

ralis

, M58

784,

147

0 Fl

exib

acte

r pol

ymor

phus

, M58

786,

144

9 Fl

exib

acte

r ros

eolu

s, M

5878

7, 1

439

Flex

ibac

ter a

ggre

gans

Pers

icob

acte

r diff

luen

s, M

5876

5, 1

463

envi

ronm

enta

l iso

late

470

77, A

F227

830,

140

7 ho

t spr

ing

clon

e O

PB73

, AF0

2700

8, 1

449

hot s

prin

g cl

one

env.

OPS

17, A

F018

199,

143

7 Th

erm

onem

a la

psum

, L11

703,

149

0 Sa

pros

pira

ceae

11

Rie

mer

ella

ana

tipes

tifer

, U10

877,

150

4 R

hodo

ther

mus ho

t spr

ing

clon

e O

PB88

, AF0

2700

6, 1

433

Chl

orob

i 1

4

Epsi

lon

prot

eoba

cter

ia 1

28

Des

ulfu

rella

les

2

CFB

Figu

re 1

2. P

artia

l AR

B n

eigh

bor-j

oini

ng tr

ee sh

owin

g re

lativ

e po

sitio

ns o

f clo

nes B

K2

and

par

tial B

K10

8T3

as

mem

bers

of t

he C

ytop

haga

/Fla

voba

cter

ia/B

acte

roid

es.

Tree

root

ed w

ith m

onop

hyle

tic a

rcha

eal o

utgr

oup.

N

umbe

rs sh

own

are

Gen

bank

acc

essi

on n

umbe

rs fo

r org

anis

ms i

n th

e pu

blic

dat

abas

es.

0.01

subs

titut

ions

/site

29

Bacteroides forsythus

Bacteroides distasonis

sulfate reducing bioreactor clone SRBAC29

sulfate reducing bioreactor clone SRBAC5

contaminated aquifer clone WCHB1

Rikanella microfusus

sulfate reducing bioreactor clone SRBAC32

BK2 clone

Cytophaga fermentans

Cytophaga lytica

Flavobacterium saccharophilum

0.05 substitutions/site

100

64

100

85

100

100

58

100

Figure 13. Bayesian inference phylogram showing relative position of clone BK2. Tree rooted with Cytophaga lytica and Flavobacterium saccharophilum as monophyletic sister clade. Bootstrap values displayed as percentages of 1000 replications.

30

sulfa

te re

duci

ng b

iore

acto

r clo

ne S

RB

AC

17, 1

392

co

ntam

inat

ed a

quife

r clo

ne W

CH

B1-

64, A

F050

606,

141

4 su

lfate

redu

cing

bio

reac

tor c

lone

SR

BA

C27

, 140

0 co

ntam

inat

ed a

quife

r clo

ne W

CH

B1-

07, A

F050

600,

151

2 co

ntam

inat

ed a

quife

r clo

ne W

CH

B1-

26, A

F050

599,

141

7 ac

id m

ine

drai

nage

clo

ne M

IM45

0, 1

425

hot s

prin

g cl

one

OPB

92, A

F027

030,

143

8 co

ntam

inat

ed a

quife

r clo

ne W

CH

B1-

11, A

F050

603,

142

3

cont

amin

ated

aqu

ifer c

lone

WC

HB

1-56

, AF0

5060

5, 1

517

tere

phth

alat

e-de

grad

ing

cons

ortiu

m c

lone

TA

18, A

F229

791,

131

8 ac

id m

ine

drai

nage

clo

ne M

IM49

0, 1

378

aqua

cultu

re p

ond

clon

e SF

39.3

80, 1

376

OP1

1-5

cont

amin

ated

aqu

ifer c

lone

WC

HB

1-01

, AF0

5059

7, 1

472

cont

amin

ated

aqu

ifer c

lone

WC

HB

1-15

, AF0

5059

6, 1

474

cont

amin

ated

aqu

ifer c

lone

WC

HB

1-06

, AF0

5059

5, 1

470

cont

amin

ated

aqu

ifer c

lone

CA

1-11

5, A

F172

911,

971

un

cultu

red

bact

eriu

m B

OL-

01, A

F172

883,

974

co

ntam

inat

ed a

quife

r clo

ne C

A1-

02, A

F172

904,

971

co

ntam

inat

ed a

quife

r clo

ne C

A1-

03, A

F172

905,

105

1 m

arin

e se

dim

ent c

lone

BM

S1-2

8, A

F172

878,

104

6 G

STM

clo

ne B

K6

sedi

men

t clo

ne B

OL-

130,

AF1

7289

6, 9

68

uncu

lture

d ba

cter

ial c

lone

D1-

11, A

F172

882,

964

la

b-sc

ale

slud

ge c

lone

SB

R10

71, A

F268

996,

142

0 la

b-sc

ale

slud

ge c

lone

SB

R20

04, A

F268

997,

139

4 de

ep-s

ea se

dim

ent c

lone

BD

4-8,

AB

0155

58, 1

568

met

al-c

onta

min

ated

soil

clon

e K

20-2

7, A

F145

827,

151

5 so

il cl

one

C1-

29, 1

433

soil

clon

e S1

-197

, 142

9 la

b-sc

ale

slud

ge c

lone

SB

R20

13, A

F269

000,

142

0 so

il cl

one

C1-

105,

148

4 m

etal

-con

tam

inat

ed so

il cl

one

K20

-12,

AF1

4581

5, 1

484

full-

scal

e sl

udge

clo

ne G

C1,

AF2

6899

4, 1

381

lab-

scal

e sl

udge

clo

ne S

BR

2060

, AF2

6899

9, 1

429

oral

cav

ity c

lone

BE1

09, A

Y00

5446

, 144

3 or

al c

avity

clo

ne B

S003

, AY

0054

48, 1

448

arid

soil

clon

e C

0-26

, AF1

2871

7, 1

437

lab-

scal

e sl

udge

clo

ne S

BR

2096

, AF2

6899

8, 1

487

or

al c

avity

clo

ne F

061,

AF1

2520

5, 1

449

lab-

scal

e sl

udge

clo

ne S

BR

2113

, AF2

6900

1, 1

395

oral

cav

ity c

lone

I025

, AF1

2520

6, 1

413

past

ure

soil

clon

e EB

1087

, 134

4 G

Sgm

clo

ne E

L59

petro

leum

-con

tam

inat

ed g

roun

dwat

er c

lone

102

5, A

B03

0611

, 138

4 co

ntam

inat

ed a

quife

r clo

ne W

CH

B1-

60, W

S5, A

F050

598,

176

9 SR

ang

clon

e 2.

3, A

F047

623,

930

trick

ling

filte

r bio

film

clo

ne k

oll6

, AJ2

2453

9, 1

423

OP1

1

WS6

TM7

Figu

re 1

4. P

artia

l AR

B n

eigh

bor-j

oini

ng tr

ee sh

owin

g re

lativ

e po

sitio

n of

clo

ne B

K6

as m

embe

r of t

he W

S6.

Tree

ro

oted

with

mon

ophy

letic

arc

haea

l out

grou

p. N

umbe

rs sh

own

are

Gen

bank

acc

essi

on n

umbe

rs fo

r org

anis

ms i

n th

e pu

blic

dat

abas

es.

0.01

subs

titut

ions

/site

31

Cam

pylo

bact

er je

juni

Des

ulfu

rella

ace

tivor

ans

hot s

prin

g cl

one

OPB

92

BK6

clon

e

cont

amin

ated

aqu

ifer c

lone

CA

1-03

mar

ine

sedi

men

t clo

ne B

MS1

-28

uncu

lture

d ba

cter

ial c

lone

D1-

11

Aci

d m

ine

drai

nage

clo

ne M

IM49

0

deep

-sea

sedi

men

t clo

ne B

D7-

4

CL5

00-1

3 cl

one

trick

ling

filte

r bio

film

clo

ne k

oll6

full-

scal

e sl

udge

clo

ne G

C1

soil

clon

e S1

-197

lab-

scal

e sl

udge

clo

ne S

BR

2004

met

al-c

onta

min

ated

clo

ne K

20-2

7

soil

clon

e C

1-10

5

oral

cav

ity c

lone

F06

1

oral

cav

ity c

lone

BE1

09

arid

soil

clon

e C

0-26

lab-

scal

e sl

udge

clo

ne S

BR

2113

past

ure

soil

clon

e EB

1087

petro

leum

-con

tam

inat

ed g

roun

dwat

er c

lone

1025

BK8T

3 cl

one

0.05

subs

titut

ions

/site

Fig

ure

15.

Bay

esia

n in

fere

nce

phyl

ogra

m sh

owin

g re

lativ

e po

sitio

ns o

f clo

nes B

K6

and

BK

8T3.

Tre

e ro

oted

with

C

ampy

loba

cter

jeju

ni a

nd D

esul

fure

lla a

cetiv

oran

s as m

onop

hyle

tic si

ster

cla

de.

Boo

tstra

p va

lues

dis

play

ed a

s pe

rcen

tage

s of 1

000

repl

icat

ions

.

90

84

92

10010

089

71

100 95

9287

83

95

52

7873

72

85

32

term

ite g

ut c

lone

TG

14, A

B00

4576

, 134

2 te

rmite

gut

clo

ne T

G33

, AB

0045

78, 1

342

term

ite g

ut c

lone

TG

25, A

B00

4577

, 134

1 te

rmite

gut

clo

ne T

G13

, AB

0045

75, 1

341

GST

M c

lone

BK

8T7

rum

en c

lone

4C

0d$3

, AB

0340

17, 1

505

anae

robi

c sl

udge

clo

ne v

adin

HB

76, U

8175

4, 1

446

lab$

scal

e sl

udge

clo

ne S

BR

2095

, AF2

6900

3, 1

453

OP5

3ho

t spr

ing

clon

e O

PB25

, OP2

, AF0

2709

7, 1

478

Figu

re 1

6. P

artia

l AR

B n

eigh

bor-j

oini

ng tr

ee sh

owin

g re

lativ

e po

sitio

n of

par

tial c

lone

BK

8T7

as m

embe

r of t

he

Term

ite G

roup

1 p

hylu

m.

Num

bers

show

n ar

e G

enba

nk a

cces

sion

num

bers

for o

rgan

ism

s in

the

publ

ic d

atab

ases

.

0.01

subs

titut

ions

/site

33