Microbial Metabolism Resulting from the Mixing of Sulfate ... · WL Wood-Ljundahl Pathway TCA...

POSIVA OY

Olki luoto

FI-27160 EURAJOKI, FINLAND

Phone (02) 8372 31 (nat. ) , (+358-2-) 8372 31 ( int . )

Fax (02) 8372 3809 (nat. ) , (+358-2-) 8372 3809 ( int . )

February 2020

Working Report 2020-2

Emma Bell , Ti ina Lamminmäki,

Petteri Pitkänen, Rizlan Bernier-Latmani

Microbial Metabolism Resulting from theMixing of Sulfate-rich and Methane-rich Deep

Olkiluoto Groundwaters in DrillholesOL-KR11, OL-KR13 and OL-KR46

February 2020

ONKALO is a registered trademark of Posiva Oy

Working Reports contain information on work in progress

or pending completion.

Emma Bell , Rizlan Bernier-Latmani

Environmental Microbio logy Laboratory,

École Polytechnique Fédérale de Lausanne (EPFL)

Tiina Lamminmäki, Petteri Pitkänen

Posiva Oy

Working Report 2020-2

Microbial Metabolism Resulting from theMixing of Sulfate-rich and Methane-rich Deep

Olkiluoto Groundwaters in DrillholesOL-KR11, OL-KR13 and OL-KR46

MICROBIAL METABOLISM RESULTING FROM THE MIXING OF SULFATE-RICH AND METHANE-RICH DEEP OLKILUOTO GROUNDWATERS IN DRILLHOLES OL-KR11, OL-KR13 AND OL-KR46

ABSTRACT

Olkiluoto, an island in south-west Finland, has been selected as the site for a deep geological repository for the final storage of spent nuclear fuel. It is therefore important to understand the geomicrobial processes underway at this site to ensure long-term, safe storage of the fuel. Of particular concern is the generation of sulfide, as it can induce corrosion of the waste-bearing copper canisters. Groundwater at Olkiluoto is hydrogeochemically stratified with depth, and sulfide production is observed when shallower sulfate-rich groundwater mixes with deeper more saline methane-rich groundwater. To constrain the electron donor(s) driving the production of sulfide (sulfidogenesis), three groundwaters from different depths with high, intermediate and low concentrations of sulfide were investigated. The three groundwaters vary in their bicarbonate, sulfate, methane, and chloride concentration. Detailed metaproteogenomic characterisation coupled to hydrogeochemical and isotopic analyses uncovered distinct communities involved in sulfur, carbon, nitrogen and iron cycling. The data shows that, in the deepest of the three groundwaters investigated (OL-KR46_570), sulfate reduction is fuelled by hydrogen as well as organic carbon from primary production and fermentation. At the transition between sulfate-rich and methane-rich groundwaters (OL-KR13_405), sulfate-reducing bacteria oxidise hydrogen and small organic compounds. Methane oxidising archaea are also detected, but it was not possible to clearly constrain the electron acceptor coupled to this process. In the third groundwater (OL-KR11_411), also a mixture of sulfate-rich and methane-rich waters, sulfide is often below detection limit, but evidence points towards the reduction of sulfate and subsequent oxidation of sulfide in a cryptic sulfur cycle. The results contribute towards our understanding of microorganisms in deep terrestrial subsurface ecosystems and their role in hydrogeochemical cycling. Keywords: bedrock, deep groundwater, sulfate-reducing bacteria, sulfide, metagenomics, metaproteomics

MIKROBIOLOGISET AINEENVAIHDUNTAREITIT OLKILUODON SYVIEN SULFAATTI- JA METAANIPITOISTEN VESIEN SEKOITTUMISEN YHTEYDESSÄ KAIRAREI'ISSÄ OL-KR11, OL-KR13 JA OL-KR46

TIIVISTELMÄ

Olkiluodon saari Lounais-Suomessa on valittu käytetyn ydinpolttoaineen geologiseksi loppusijoituspaikaksi. Tästä syystä on tärkeä ymmärtää paikan geomikrobiologisia prosesseja, jotta voidaan varmistua loppusijoituksen pitkäaikaisturvallisuudesta. Yksi erityinen prosessi on pohjaveden sulfaatin pelkistyminen sulfidiksi, koska liuennut sulfidi voi aiheuttaa korroosiota kuparille, josta valmistetaan loppusijoituskapselin ulkokuori. Olkiluodon pohjavesikoostumus on kerroksellinen syvyyden suhteen ja sulfidin muodostumista on havaittu tapahtuvan kallion yläosassa olevan sulfaattipitoisen veden sekoittuessa syvemmällä olevan suolaisen metaanipitoisen veden kanssa. Sulfidin tuotannossa (sulfidogenesis) käytettävän elektronin luovuttajan selvittämiseksi tutkittiin kolmelta eri syvyydeltä pohjavesinäytteitä, joissa oli havaittu Olkiluodon olosuhteissa matala, keskitasoinen ja korkea sulfidipitoisuus. Näiden kolmen eri näytteenottokohteen pohjavesikemia vaihteli mm. bikarbonaatin, sulfaatin, metaanin ja kloridin pitoisuuksien suhteen. Yksityiskohtainen metaproteogenomien karakterisointi yhdistettynä hydrogeokemiallisiin ja isotooppi-analyyseihin osoitti erilaisten mikrobipopulaatioiden liittyvän rikin, hiilen, typen ja raudan kierrättämiseen tutkituissa pohjavesissä. Tulosten perusteella syvimmässä näytekohteessa (OL-KR46) sulfaatin pelkistyksen elektronin luovuttajana toimivat vety sekä orgaaninen hiili, joka muodostui alkutuotanto- ja fermentaatioprosesseissa. Kairareiän OL-KR13 näytevälissä sulfaatti- ja metaanipitoisen veden sekoittumisen yhteydessä sulfaatinpelkistäjäbakteerit hapettivat vetyä sekä pieniä orgaanisia yhdisteitä. Myös metaania hapettavia arkeoneja havaittiin, mutta tähän prosessiin liittyvää elektronin vastaanottajaa ei pystytty täysin määrittämään. Kolmannessa pohjavesinäytekohteessa (OL-KR11), jossa on todettu myös sulfaatti- ja metaanipitoisen pohjavesien sekoittuminen, sulfidipitoisuus on ollut yleisesti alle määritysalarajan. Näytekohteessa havaittiin kuitenkin monimutkainen rikinkiertoprosessi, jossa samanaikaisesti sulfaattia pelkistyi ja sulfidia hapettui. Tulokset auttavat ymmärtämään paremmin syvien kalliopohjavesien ekosysteemejä ja mikrobien merkitystä hydrogeokemiallisissa redox-vuorovaikutuksissa. Avainsanat: kallioperä, syväpohjavesi, sulfaatinpelkistäjäbakteeri, sulfidi, metagenomiikka, metaproteomiikka

1

TABLE OF CONTENTS

ABSTRACT

TIIVISTELMÄ

ABBREVIATIONS USED IN THIS REPORT. ................................................................ 3

GLOSSARY .................................................................................................................. 5

PREFACE ..................................................................................................................... 7

1 INTRODUCTION................................................................................................... 9

1.1 Aim ................................................................................................................. 9 1.2 Background .................................................................................................... 9

1.2.1 Drillhole OL-KR11 .................................................................................. 10 1.2.2 Drillhole OL-KR13 .................................................................................. 11 1.2.3 Drillhole OL-KR46 .................................................................................. 13

1.3 Approach ...................................................................................................... 15 2 MATERIALS AND METHODS............................................................................. 19

2.1 Samples ....................................................................................................... 19 2.2 Geochemical characterisation ....................................................................... 21

2.2.1 Anions ................................................................................................... 21 2.2.2 Organic Compounds .............................................................................. 21 2.2.3 Sulfide ................................................................................................... 21 2.2.4 Dissolved gases (methane and hydrogen) ............................................. 21 2.2.5 Dissolved organic carbon (DOC) ........................................................... 21 2.2.6 Sulfur isotopes (δ34SSO4) ........................................................................ 21 2.2.7 Carbon isotopes (δ13CDIC and δ13CCH4) ................................................... 22

2.3 Microbiological characterisation .................................................................... 23 2.3.1 Cell counts ............................................................................................. 23 2.3.2 Biomass collection ................................................................................. 23 2.3.3 DNA extraction ...................................................................................... 24 2.3.4 16S rRNA gene amplicon sequencing ................................................... 24 2.3.5 Metagenomics and construction of metagenome assembled genomes (MAGs) .............................................................................................................. 25 2.3.6 Metaproteomics ..................................................................................... 25

2.4 Groundwater Incubations .............................................................................. 26 2.4.1 Catalysed reporter deposition-fluorescence in situ hybridization (CARD-FISH). .............................................................................................................. 26 2.4.2 Nanoscale secondary ion mass spectrometry (nanoSIMS) .................... 26

3 RESULTS AND DISCUSSION OL-KR11_411 .................................................... 27

3.1 Hydrogeochemistry ....................................................................................... 27 3.2 Microbiological analyses ............................................................................... 33

3.2.1 Community composition ........................................................................ 33 3.2.2 Metaproteogenomics ............................................................................. 34 3.2.3 Metagenome assembled genomes (MAGs) ........................................... 39

3.2.3.1 Sulfidogenic bacteria ...................................................................... 39

3.2.3.2 Sulfide-oxidising bacteria ................................................................ 42 3.2.3.3 Nitrate reducing bacteria ................................................................. 45 3.2.3.4 Iron(III)-reducing bacteria ............................................................... 46 3.2.3.5 Fermentation .................................................................................. 46

2

3.3 Summary of metabolism in OL-KR11_411 groundwater ............................... 47 4 RESULTS AND DISCUSSION OL-KR13_405 .................................................... 49

4.1 Hydrogeochemistry ....................................................................................... 49 4.2 Microbial community analysis ....................................................................... 52

4.2.1 Metagenome assembled genomes (MAGs) ........................................... 54 4.2.1.1 Sulfate-reducing bacteria ................................................................ 55

4.2.1.2 Methane cycling archaea ................................................................ 57

4.2.1.3 Acetogenic bacteria ........................................................................ 58

4.2.1.4 Fermentation .................................................................................. 58

4.3 Summary of metabolism in OL-KR13_405 groundwater ............................... 59 5 RESULTS AND DISCUSSION OL-KR46 ............................................................ 61

5.1 Hydrogeochemistry ....................................................................................... 61 5.2 Microbial community analysis ....................................................................... 64

5.2.1 Metagenome assembled genomes (MAGs) ........................................... 66 5.2.1.1 Sulfate-reducing bacteria ................................................................ 67

5.2.1.2 Acetogenic bacteria ........................................................................ 68

5.2.1.3 Methane cycling archaea ................................................................ 68

5.2.1.4 Sulfide oxidising bacteria ................................................................ 68

5.2.2 Fermentation ......................................................................................... 69 5.3 Summary of metabolism in OL-KR46 groundwater ....................................... 70

6 INCUBATION EXPERIMENTS ............................................................................ 73

6.1.1 Sulfide oxidation in OL-KR11_411 ......................................................... 73 6.1.2 Anaerobic oxidation of methane in OL-KR13_405 ................................. 81 6.1.3 Acetate oxidation in OL-KR46_570 ........................................................ 89

7 SYNTHESIS ........................................................................................................ 97

8 REFERENCES ................................................................................................. 101

9 SUPPLEMENTARY INFORMATION ................................................................. 109

3

ABBREVIATIONS USED IN THIS REPORT.

Abbreviation Explanation DIC Dissolved inorganic carbon DOC Dissolved organic carbon Mbsl Metres below sea level OL-KR Core drilled drillhole VPDB Vienna Pee Dee Belemnite (‰) VCDT Vienna Canyon Diablo Troilite (‰) OTU Operational taxonomic unit MAG Metagenome assembled genome SRB Sulfate-reducing bacteria SRM Sulfate-reducing microorganisms SOB Sulfide oxidising bacteria MRB Metal reducing bacteria NRB Nitrate reducing bacteria DNRA Dissimilatory nitrate reduction to ammonia WL Wood-Ljundahl Pathway TCA Tricarboxylic Acid Cycle rTCA Reductive Tricarboxylic Cycle PP Pentose Phosphate Pathway rPP Reductive Pentose Phosphate Pathway ED Entner-Duodoroff Pathway bp Nucleotide base pairs Mbp Mega base pairs (1,000,000 bp) Gbp Giga base pairs (1,000,000,000 bp) SI Supplementary Information

5

GLOSSARY

Explanation Acetoclastic methanogen An organism that oxidises acetate and produces

methane. Acetogen Obligately anaerobic bacteria that produce

acetate. Autotroph An organism able to grow with inorganic carbon

(carbon dioxide or bicarbonate) as its sole source of carbon.

Assimilatory A process which generates molecules that are incorporated into cell material.

Calvin Cycle A series of biosynthetic reactions by which autotrophic organisms convert CO2 to organic compounds which can be assimilated.

Carbon fixation The conversion of inorganic carbon to organic carbon.

Chemolithotroph An organism that obtains its energy from the oxidation of inorganic compounds.

Chemoorganotroph An organism that obtains its energy from the oxidation of organic compounds.

Dehydrogenase An enzyme that catalyses oxidation reactions by transferring hydrogen from a substrate to an acceptor.

Denitrification The microbial reduction of nitrate to dinitrogen gas (NO3

- to N2). Dissimilatory An energy-gaining respiratory process. Dissimilatory nitrate reduction The microbial reduction of nitrate to ammonia

(NO3- to NH3).

Dissimilatory sulfate reduction The microbial reduction of sulfate to sulfide (SO4

2- to S2-). Endospore A differentiated cell formed within certain gram-

positive bacteria that is dormant and extremely resistant to heat as well as to other harmful agents.

Enzyme A catalyst for chemical reactions that occur within a cell, enzymes are typically proteins and end with the suffix -ase e.g. hydrogen-ase.

Fermentation The anaerobic degradation of an organic compound which serves as both an electron acceptor and an electron donor and in which ATP is usually produced by substrate level phosphorylation.

Fermenter An organism that carries out the process of fermentation.

Gene A region of DNA (deoxyribonucleic acid) that specifies a protein. DNA is transcribed into RNA which is translated into protein. Genes are written with a lowercase first letter and are italicised e.g. dsrAB.

Glycolysis The conversion of glucose to pyruvate via the Embden-Meyerhof-Parnas pathway.

Hydrogenase An enzyme, widely distributed in anaerobic microorganisms, capable of H2-uptake or evolution.

6

[NiFe] hydrogenases A type of hydrogenase with a [NiFe] active site. Usually an uptake hydrogenase (hydrogen oxidation).

Heterotroph An organism that requires organic carbon as its carbon source; also a chemoorganotroph.

Hydrogenotroph An organism that oxidises H2 as a source of energy.

Metagenomics DNA is extracted from an environmental sample with a mixed microbial community. Metagenomics analyses the genes present (metabolic potential) within the whole microbial community.

Metaproteomics Protein is extracted from an environmental sample with a mixed microbial community. Metaproteomics is the measurement of whole community gene expression (which genes are active). Proteins correspond to genes encoded in the metagenome.

Metaproteogenomics The combined analysis of the metagenome and the metaproteome.

Methanogen An organism that produces CH4. Mixotroph An organism that uses organic compounds as

carbon sources but uses inorganic compounds as electron donors for energy metabolism.

Nitrification The microbial oxidation of ammonia to nitrite (NH3 to NO3

-). Oligotrophic An environment with low nutrients. An

oligotroph is an organism that can survive in an environment with low levels of nutrients.

Peptide A peptide is a short chain of ≥2 amino acids. Proteins are made up of multiple peptides, also known as polypeptides.

Primary producer An organism that synthesises new organic material from CO2 and obtains energy from light or from the oxidation of inorganic compounds. Also, an autotroph.

Protein A molecule consisting of one or more polypeptides (see peptide). Proteins are important for cell function; the two major types are catalytic proteins (enzymes; see enzyme) and structural proteins. Proteins are written with a capital letter and are not italicised e.g. DsrAB.

Reductase An enzyme that catalyses a biochemical reduction reaction.

Sulfidogen An organism that produces sulfide. Sulfidogenesis The biological generation of sulfide. Spore A general term for dormant structures formed by

many prokaryotes and fungi. See endospore. Spore maturation Part of the sporulation process, where spores are

formed. Sporulation The process whereby an organism in a vegetative

state forms a spore. Wood-Ljungdahl Pathway A pathway of autotrophic CO2 fixation and

acetate oxidation widespread in anaerobes including methanogens, acetogens and sulfate-reducing bacteria.

7

PREFACE

This Working Report is supervised by Tiina Lamminmäki and Petteri Pitkänen (Posiva Oy). The work was carried out by Emma Bell, supervised by Rizlan Bernier-Latmani at the Swiss Federal Institute for Technology in Lausanne (EPFL).

Maarit Yli-Kaila and Raila Viitala (Posiva Oy) contributed to the organisation of field work at Olkiluoto. Manon Frutschi, Guillaume Sommer and Louise Balmer (EPFL) contributed to the collection of samples from Olkiluoto. Louise Balmer also performed some of the hydrogeochemical analyses.

Metagenomic analysis was performed by Emma Bell with the help of Johannes Alneberg and Anders F. Andersson at KTH Royal Institute of Technology, Science for Life Laboratory, Stockholm. Protein was extracted and analysed at the Oak Ridge National Laboratory, Chemical Sciences Division, Tennessee by Chen Qian and Weili Xiong, supervised by Robert L. Hettich.

9

1 INTRODUCTION

1.1 Aim

The concentration of sulfide is relatively high in groundwater from some water conductive fractures at Olkiluoto. Microbial activity produces sulfide, as sulfide is produced by sulfate-reducing microorganisms that use sulfate as an electron acceptor during anaerobic respiration. The electron donor fuelling this process at Olkiluoto has not been clearly identified. To gain a better understanding of the electron donors and metabolic processes contributing to the production of sulfide in Olkiluoto groundwater, hydrogeochemical, metagenomic and metaproteomic analyses were combined to identify potential energy and electron source(s) driving the production of sulfide (sulfidogenesis). Three hydrogeochemically distinct groundwaters which vary in their sulfide concentration were investigated from three different drillholes (OL-KR11_411, OL-KR13_405 and OL-KR46_570; Table 1-1). Of these three groundwaters, OL-KR46_570 has the highest concentration of sulfide, OL-KR13_405 has an intermediate concentration of sulfide and OL-KR11 has little to no sulfide. By this approach we aim to identify which microorganisms produce sulfide and what electron donor(s) they use.

1.2 Background

Groundwater chemistry at Olkiluoto is stratified with depth. In baseline conditions, where deep drilling or construction of underground tunnels has not affected the groundwater, fresh groundwater (total dissolved solids (TDS) <1 g/L) is found at shallow depths (<30 m). Salinity increases with depth and brackish groundwater (TDS >1 <10 g/L) abundant in sulfate is found up to ~400 m depth, beyond which sulfate-free saline groundwater (TDS >10 g/L) dominates (Posiva, 2013). OL-KR11_411 and OL-KR13_405 are both brackish and OL-KR46_570 is saline (Table 1-1). The number at the end of the drillhole name refers to drillhole length of the sampling section in metres (upper packer) e.g. the drillhole OL-KR11 was sampled from the section starting at 411 m drillhole length. All water conductive fractures sampled during this study (Table 1-1) have been affected by mixing, where sulfate-rich brackish groundwater has mixed with sulfate-free saline groundwater due to open drillhole phases or drawdown caused by excavated ONKALO tunnels. Therefore, all water types represent a mixture. This means that sulfate is detected in deep saline groundwater from OL-KR46_570. This drillhole has the greatest concentration of sulfide (Table 1-1). The concentration of sulfate is greatest in OL-KR11, but no sulfide is detected (Table 1-1). Both sulfate and sulfide are detected in OL-KR13 (Table 1-1). The concentration of dissolved inorganic carbon (DIC) is also a distinguishing characteristic of different groundwater types at Olkiluoto. DIC is generally present at greater concentrations at shallower depths. In this study, DIC is most abundant in OL-KR11 (Table 1-1).

10

Table 1-1. Hydrogeochemical characteristics of drillholes OL-KR11_411, OL-KR13_405 and OL-KR46_570, taken prior to this study (2013–2016). Chemical measurements were carried out by TVO (methods in Table S1).

OL-KR11_411 (411–430 m drillhole length)

Depth (mbsl) Date Sulfate (mM)

Sulfide (µM)

TDS (g/L)

Chloride (g/L)

DIC (mg/L)

366.7–383.5 14/07/14 3.25 <0.63 5.26 2.89 26 366.7–383.5 08/02/16 3.54 <0.63 5.20 2.81 28

OL-KR13_405 (405.5–414.5 m drillhole length)


Sulfide (mM)

TDS (g/L)

Chloride (g/L)

DIC (mg/L)

330.5–338.0 25/11/13 0.32 0.38 6.27 3.77 14 330.5–338.0 20/02/14 0.40 0.50 6.67 4.05 13 330.5–338.0 24/11/14 0.39 0.44 6.79 4.12 12 330.5–337.9 22/02/16 0.52 0.44 7.07 4.27 11

OL-KR46_570 (570.5–573.5 m drillhole length)


Sulfide (mM)

TDS (g/L)

Chloride (g/L)

DIC (mg/L)

528.2–531.5 11/10/13 3.80 0.97 11.91 6.92 16 528.7–531.5 02/06/14 5.82 0.56 11.48 6.69 13 528.7–531.5 08/12/14 4.03 0.94 12.71 7.44 10 528.7–531.5 22/06/15 2.15 1.53 14.54 8.71 3.6 528.7–531.5 28/01/16 1.28 1.38 15.13 9.14 4.6

mbsl = metres below sea level

1.2.1 Drillhole OL-KR11

OL-KR11 was drilled during 27/05/1999–03/07/1999. Multipacker systems have been installed in OL-KR11 during multiple periods: 16/12/2002 – 4/10/2007 20/04/2008 – 04/05/2009 20/05/2009 – 11/03/2010 17/01/2013 – present (2018) In the interim periods, the drillhole was open, allowing flow from fractures with a higher hydraulic head to fractures with a lower hydraulic head. The flow direction of the sampling section is from drillhole to bedrock in open drillhole conditions and the transmissivity is ~1.5–1.7 × 10-8 m2/s (measured in 2010). Long-term pumping of the section sampled during this study (411–430 m) at 366.7−383.5 mbsl began on 07/01/2016, two months prior to the beginning of microbiological sampling on 01/03/2016. The volume of water pumped prior to sampling should have been sufficient to remove contaminating drillhole water from the fracture, however the elevated concentration of sulfate at this depth relative to baseline values (Figure 1-1) suggests that the section has not fully recovered from open-hole phases. A connection of this fracture to ONKALO is not known to exist. However, hydraulic head data suggested that a minor connection to a shallower fracture may have formed via tube leakage during the sampling

11

period in 2016 (Figure 1-1; 110–112 mbsl (125–127 drillhole length)). The minor leakage was confirmed 2019 when packers were lifted up and inspected.

Figure 1-1. Drillhole OL-KR11. The fracture at 366–383 mbsl (411–430 m drillhole length) was sampled. Depths are shown in mbsl. Comparative baseline values for Olkiluoto are indicated by thatched boxes (Posiva, 2013). HZ = modelled hydrogeological zones at Olkiluoto (Vaittinen et al., 2011).


OL-KR13 was drilled during 04/04/2001–02/05/2001. A multipacker system was installed on 07/09/2007 which is in place to date. During the six-year period prior (May 2001–September 2007) the drillhole was open. In open drillhole conditions, the flow direction is from drillhole to bedrock and the transmissivity of the fracture is ~3.3–5.0 × 10-7 m2/s (measured in 2010). The packer system has been in place for 10 years, so minimal contamination from the open drillhole phase is expected. Long-term pumping of

12

section sampled during this study (405.5–414.5 m) at 330.5−337.9 mbsl began on 07/01/2016, two months prior to microbiological sampling on 01/03/2016. Chloride and sulfate values for this section are consistent with the baseline values (Figure 1-2), however some contamination from the upper part of the drillhole is possible as the calculated volume of water moving into the fracture during the open phase (90 m3) is greater than the volume pumped during sampling (17 m3).

Figure 1-2. Drillhole OL-KR13. The fracture at 335 mbsl corresponds to the sampled section at 330–338 mbsl (405.5–414.4 m drillhole length). Depths are shown in mbsl. Comparative baseline values for Olkiluoto are indicated by thatched boxes (Posiva, 2013). HZ = modelled hydrogeological zone, BF = brittle fracture zone (Vaittinen et al., 2011). OL-KR13 is modelled to intersect the fracture zone OL-BFZ045 (Figure 1-2 & 1-3). This fracture zone also intersects several open tunnels in ONKALO. This could result in groundwater movement caused by drawdown towards open tunnels via this fracture zone.

13

Figure 1-3. Fracture zones intersecting OL-KR13 and ONKALO (Vaittinen et al., 2011).


OL-KR46 was drilled during 25/04/2007–04/06/2007. The section sampled during this study (570.5–573.5 m) at 530.6 mbsl was isolated with double packers on 18/12/2013. During the six-year interval prior to the installation of packers, the drillhole was left open. The flow direction during open-hole phase is from drillhole to bedrock, causing drawdown of shallower sulfate-rich groundwater during the open phase. Contamination of shallower brackish sulfate-rich groundwater is still evident from the hydrogeochemical measurements (Figure 1-4). Since the packers have been installed there have been leakages and the packers have been taken out and replaced: Packers removed 06/03/2014 Packers replaced 12/03/2014 Leakage observed 05/07/2016 Leak prevented 21/07/2016 Long-term pumping started on 16/12/2015, three months prior to the first microbiological sampling (01/03/2016). Transmissivity of the fracture has reduced from 2.2 × 10-9 to 5.3 × 10-10 m2/s (measured in 2017), suggesting some minerals may have formed in the fracture (possibly calcite and/or iron sulfide/pyrite).

14

OL-KR46 is drilled through the hydrological zone OL-HZ056 (Figure 1-4 and 1-5). This fracture zone intersects ONKALO (Figure 1-5) which causes drawdown and groundwater movement towards open tunnels.

Figure 1-4. Drillhole OL-KR46. The fracture at 530 mbsl corresponds to the sampled packer interval at 529–532 mbsl (570.5–573.5 m drillhole length). Depths are shown in mbsl. Comparative baseline values for Olkiluoto are indicated by thatched boxes (Posiva, 2013). HZ = modelled hydrogeological zones at Olkiluoto (Vaittinen et al., 2011).

15

Figure 1-5. Hydrological zone OL-HZ056 intersecting OL-KR46 and ONKALO (Vaittinen et al., 2011).

1.3 Approach

To identify which microorganisms produce sulfide and what electron donor(s) they use, a microbial metabolic model for Olkiluoto groundwater was reconstructed using a combination of hydrogeochemical and microbiological methods (Figure 1-6). The biodiversity is explored by extracting total genomic DNA from the groundwater. A single gene from the DNA can be sequenced e.g. the 16S rRNA gene which encodes the small subunit (16S) of ribosomal RNA in bacteria and archaea. 16S rRNA gene amplicon sequencing provides taxonomic information of the community composition. When all of the genes from a community are sequenced this is termed metagenomics. Metagenomics provides information about the gene contents and potential metabolic processes (what microorganisms are able to do) present in a microbial community. In the cell, DNA (encoding a gene) is transcribed into RNA, then RNA is translated into proteins. Proteins may be structural or catalytic (an enzyme). When microorganisms are actively performing a function, the necessary genes will be translated into proteins. Analysing the proteins from a microbial community (the metaproteome) therefore tells us what metabolic processes are active. Using metagenomics and metaproteomics together is termed metaproteogenomics.

16

Figure 1-6. Schematic workflow of microbiological analyses. When sequencing the metagenome, it is also useful to determine which genes come from which microorganisms. Assembled metagenomic sequence reads can be binned using bioinformatics tools (Figure 1-7). The bins represent the genomes of different taxa found in the groundwater. Good quality bins with high completeness and low contamination are considered to be metagenome assembled genomes (MAGs; i.e. genomes assembled in silico from metagenomic sequence data). If genes from the MAG are identified in the metaproteome, the proteins belonging to that MAG are referred to as a proteome i.e. the proteins from a single organism, opposed to the proteins from the whole community (metaproteome).

17

Figure 1-7. Schematic of the workflow for the construction of metagenome assembled genomes (MAGs), adapted from Kang et al., (2015). First metagenomic sequencing of the microbial community is performed. The short (~150 bp) sequence reads are then assembled into longer contigs. Contigs are placed into bins based on their sequence composition and coverage across multiple samples using CONCOCT (Alneberg et al., 2014). High quality MAGs are constructed representing a single lineage within the microbial community.

19

2 MATERIALS AND METHODS

2.1 Samples

Samples were taken during 2016 for 16S rRNA gene amplicon sequencing, metagenomics and metaproteomics (Table 2-1, 2-2 & 2-3). The microbial community composition was monitored throughout the sampling period. Multiple metagenomes were generated during certain sampling months (Table 2-1 & 2-2). This was to improve the recovery of MAGs, as binning uses the abundance across different samples to bin assembled contigs (Figure 1-7). Samples were also collected in 2018 for 16S rRNA gene amplicon sequencing and groundwater incubations. The groundwater incubation experiments were designed to confirm metabolic processes determined by the metaproteogenomic approach. Table 2-1. Microbiological samples taken from OL-KR11_411. Blue shaded boxes indicate that a sample was taken (x1). If multiple samples were taken during the same period this is indicated (e.g. two samples = x2). White boxes indicate that no sample was taken.

Sampling Period 16S rRNA

gene amplicon

Metagenome Metaproteome Incubation

01/03/2016 – 10/03/2016 x1 03/05/2016 – 13/05/2016 x1 07/06/2016 – 15/06/2016 x1 x1 05/07/2016 – 14/07/2016 x1 09/08/2016 – 17/08/2016 x1 27/09/2016 – 05/10/2016 x3 x3 x1 08/11/2016 – 17/11/2016 x2 x2 +1 x1 21/11/2018 – 23/11/2018 x1 x1

20

Table 2-2. Microbiological samples taken from OL-KR13_405. Green shaded boxes indicate that a sample was taken (x1). If multiple samples were taken during the same period this is indicated (e.g. two samples = x2). White boxes indicate that no sample was taken.

Sampling Period 16S rRNA gene

amplicon


01/03/2016 – 10/01/2016 x1 03/05/2016 – 13/05/2016 x1 07/06/2016 – 15/06/2016 x1 x1 05/07/2016 – 14/07/2016 x1 x1 09/08/2016 – 17/08/2016 x1 27/09/2016 – 05/10/2016 x3 x3 x1 08/11/2016 – 17/11/2016 x2 x2 x1 21/11/2018 – 23/11/2018

Table 2-3. Microbiological samples taken from OL-KR46_570. Orange shaded boxes indicate that a sample was taken (x1). White boxes indicate that no sample was taken. The asterisk (*) indicates that no samples were taken, due to the packer leakage.

Sampling Period 16S rRNA gene

amplicon


01/03/2016 – 10/01/2016 x1 x1 03/05/2016 – 13/05/2016 x1 07/06/2016 – 15/06/2016 x1 x1 05/07/2016 – 14/07/2016 * * * 21/08/2016 – 23/08/2016 x1 x1 x1 27/09/2016 – 05/10/2016 x1 08/11/2016 – 17/11/2016 x1 21/11/2018 – 23/11/2018 x1 x1

21

2.2 Geochemical characterisation

The methods for hydrogeochemical analysis performed at EPFL are described.

2.2.1 Anions

Water collected for the determination of sulfate and thiosulfate was filtered (0.2 µm pore size) into a sterile 50 mL Hungate tube and stored at 4°C. Sulfate and thiosulfate were measured by ion chromatography using a Dionex Integrion HPIC system with an IonPac AS18 analytical column. The flow rate was set to 0.25 mL/min and the eluent was KOH (40 mM).

2.2.2 Organic Compounds

Water was filtered (0.2 µm) and collected into a sealed sterile 10 mL Balch tube by inserting a needle through the butyl rubber stopper. Samples were collected into sealed tubes to prevent loss of volatile compounds and were stored at 4°C. Organic acids (acetate, lactate, propionate, butyrate) and glucose were measured by HPLC with an Agilent Hi-Plex H column. Alcohols (ethanol, methanol, propanol and 2-butanol) and acetone were measured by gas chromatography-mass spectrometry (GC-MS). 1-butanol was used as the internal standard.

2.2.3 Sulfide

Sulfide was fixed in solution by filtering (0.2 µm) groundwater into a sterile Falcon tube containing a 5% zinc acetate solution (1% final concentration). Samples were stored at -20°C. Sulfide was measured spectrophotometrically on a Shimadzu UV-2501PC using the cline assay (Cline, 1969).

2.2.4 Dissolved gases (methane and hydrogen)

Water was collected by inserting a needle through the butyl rubber stopper of a sealed N2-filled Wheaton glass serum bottle containing the biocide NaN3. Serum bottles were stored inverted, to minimise gas loss, at room temperature. The gas phase was sampled using a gas tight syringe and H2 and CH4 measured on a Varian 450-GC (Agilent, Santa Clara, USA) with a flame ionisation (FID) and thermionic specific (TSD) detector. Dissolved gas concentrations were calculated as described previously (Bagnoud et al., 2016). Dissolved gases can be difficult to measure when samples are taken at ground level as the change in pressure results in degassing therefore gas results may not accurately represent the concentration in situ.

2.2.5 Dissolved organic carbon (DOC)

Groundwater was filtered (0.2 µm) into a sterile 50 mL Hungate tube and stored at 4°C. DOC was measured on a Shimadzu TOC-V.

2.2.6 Sulfur isotopes (δ34SSO4)

Sulfide was fixed in solution with zinc acetate (final concentration 1%). The zinc sulfide precipitate was removed by centrifugation at 4,500 x g for 10 minutes. The zinc sulfide

22

pellet was discarded and the supernatant was used for measurement of isotopic sulfate. The supernatant was first acidified to ~pH 3 with 1 M HCl. Acidified samples were then heated to 90°C in a water bath and sulfate was precipitated as BaSO4 by the addition of BaCl2 (10% solution). Samples were left to cool overnight then centrifuged at 4,500 xg for 10 minutes. The supernatant was discarded and the BaSO4 pellet air-dried overnight. The 34S/32S isotope ratio of sulfate (δ34SSO4 ‰ VCDT) was measured at the Stable Isotope Laboratory (Faculty of Geosciences and Environment, Université de Lausanne, Lausanne, Switzerland). The S-isotope compositions were measured with He carrier gas and a Carlo Erba (CE 1100) elemental analyser linked to a Thermofisher Delta V mass spectrometer. Samples were reacted at 1050°C in a stream of He-carrier gas spiked with oxygen gas. External reproducibility of standards was better than 0.15 ‰ and samples were calibrated against IAEA standards S1 and S3 (Ag2S) and NBS-127 (BaSO4) with accepted values of –0.3, –32.1 and 20.3 ‰, respectively.

2.2.7 Carbon isotopes (δ13CDIC and δ13CCH4)

For analysis of the 13C/12C ratio of dissolved inorganic carbon (DIC) 30 mL Wheaton glass serum bottles were rinsed three times in filtered water and sealed with a butyl rubber stopper and crimp. Filtered (0.2 µm) groundwater was collected into the serum bottles through a needle inserted into the stopper. Serum bottles were filled so that there was no headspace and stored at 4°C. The 13C/12C ratio of dissolved inorganic carbon (δ13CDIC ‰ VPBD) was measured at the Stable Isotope Laboratory (Faculty of Geosciences and Environment, Université de Lausanne, Lausanne, Switzerland) with a Thermo Finnigan Delta Plus XL IRMS equipped with a GasBench II for analyses of carbonates. The C- and O- isotope composition of carbonates were measured with a GasBench II connected to a Finnigan MAT DeltaPlus XL mass spectrometer, using a He-carrier gas system (Spötl and Vennemann, 2003). Water (0.2 to 1.2 mL) was injected into sample vials containing six drops of orthophosphoric acid that were flushed with He prior to injection. After 1 hour of reaction and several minutes of agitation, the sample vials were inserted into the GasBench II at room temperature for analyses. Solid carbonate in-house standards were reacted with the same acid at 70ºC for one hour but their CO2 was also extracted at room temperature. Samples were normalised using an in-house standard calibrated against δ13C values of NBS-19 (+1.95‰, relative to VPDB). External reproducibility for the analyses estimated from replicate analyses of the in-house standard reacted at 110ºC (n=6) was ±0.04‰ for δ13C. The signal area of mass 44 of the samples was also used to calculate the concentration of dissolved inorganic carbon in solution. The precision of this method is 5% based on reproducibility of standard samples. For analysis of the 13C/12C ratio of methane (δ13CCH4 ‰ VPBD), sealed serum bottles containing the biocide NaN3 were filled with groundwater so that there was no headspace. A known volume of sample was taken with a syringe flushed with synthetic air. The syringe was then shaken for ~1 minute to create a headspace. The 13C/12C ratio of methane in the headspace was measured using a Picarro G2201-I Analyser.

23

2.3 Microbiological characterisation

2.3.1 Cell counts

Groundwater was filtered through a 0.3 µm filter to remove any particulates in the groundwater and immediately fixed in paraformaldehyde at a final concentration of 1–2%. Fixed samples were stored at 4°C. For cell enumeration, 10 mL of fixed sample was filtered onto a black polycarbonate 0.22 µm pore size filter membrane and washed three times with PBS (8 g/L NaCl, 0.2 g/L KCl, 1.4 g/L Na2HPO4, 0.24 g/L KH2PO4, pH 7.4). The filter membrane was stained with SYBR Green (Thermofisher) for 20 minutes and cells imaged on an epifluorescent Nikon Eclipse E800 microscope at 1,000x magnification. A minimum of ten fields of view were imaged per sample.

2.3.2 Biomass collection

Groundwater was pumped directly into a chilled sterile Nalgene filtration unit fitted with a filter and connected to a vacuum pump (Figure 2-1). Biomass was collected on a 0.22 µm Isopore polycarbonate filter membrane (Merck). Filters collected for DNA were removed aseptically and stored in a 1.5 mL screwcap tube containing 750 µL of LifeGuard Soil Preservation Solution (MoBio, Carlsbad, CA, USA). Filters collected for protein analysis stored in 1.5 mL screwcap tubes and flash-frozen in a dry ice and ethanol mixture. Filter tubes were stored on dry ice immediately on site. Filters for DNA were subsequently stored at -20°C and filters for protein were stored at -80°C. Biomass was also collected on a 0.1 µm pore size Isopore polycarbonate filter (Merck) to capture ultrasmall microorganisms. Filtrate from the 0.22 µm filtration was transferred to a second chilled Nalgene filtration unit fitted with a 0.1 µm filter. Filters were removed and preserved for DNA extraction as described above.

24

Figure 2-1. Sterile tubing is attached with a swagelock to the groundwater outflow (1) and collected directly into a sterile Nalgene filtration unit (2) fitted with a vacuum pump (3). Collected water is filtered through a 0.22 µm pore size polycarbonate filter (4) and filtrate collected into a sterile Duran bottle (5). During filtration, filter units are kept chilled by placing them into a cool bag and filled with ice packs. Flexible ice packs are wrapped around the filter holder.

2.3.3 DNA extraction

DNA was extracted using a phenol-chloroform method described previously (Bagnoud et al., 2016) with the following modifications; 1) filter pieces were subject to bead-beating (2 x 15 seconds) prior to incubation in lysozyme for 2 hrs at 37°C; 2) lysate was incubated in Proteinase K (200 µg/mL final concentration) for 2 hrs. The DNA concentration of the extracts were quantified on a Qubit 3.0 Fluorometer using the Qubit dsDNA High Sensitivity Assay kit (Thermo Fisher Scientific). The volume of groundwater filtered and total extracted DNA is provided in the Supplementary Information (Table S2).

2.3.4 16S rRNA gene amplicon sequencing

Extracted DNA was used as a template for PCR amplification using primers 515F/806R that target the V4 region of the 16S rRNA gene (Caporaso et al., 2011). Libraries were generated on an Illumina MiSeq at either RTL Genomics or the Lausanne Genomics Technologies Facility at Université de Lausanne (UNIL) with a 2 x 250 bp read configuration resulting in an ~200 bp overlap. Sequence reads were merged and quality filtered with Usearch v11 (Edgar, 2010). Operational taxonomic units (OTUs) were assigned and chimeras removed using UNOISE2 (Edgar, 2016). Taxonomy was assigned to OTUs with RDP classifier (Cole et al., 2009) in QIIME (Caporaso et al., 2010) using

25

the Silva 132 database (Quast et al., 2013). Rarefaction to the lowest number of sequences was used to normalise the sample count prior to analysis of the dataset.

2.3.5 Metagenomics and construction of metagenome assembled genomes (MAGs)

Metagenomic sequencing was performed at three facilities: (1) RTL Genomics, TX, USA, (2) the Joint Genome Institute (JGI), Walnut Creek, CA, USA, and (3) the Marine Biological Laboratory (MBL), Woods Hole, MA, USA. Each facility prepared metagenomic libraries from raw DNA using their standard protocol for low DNA input samples, which requires 25 ng of DNA. DNA was sheared into fragments (~275 bp) and were sequenced with a 2 x 150 bp read configuration. Samples sequenced at RTL Genomics and the JGI were sequenced on an Illumina HiSeq. Samples sequenced at MBL were sequenced on an Illumina NextSeq. Full details of the samples sequenced are provided in the Supplementary Information (Table S3). Pre-processing of metagenomic sequence reads was performed using FastUniq (Xu et al., 2012). A separate assembly was performed for each sample using MegaHit (Li et al., 2015). Assembled contigs were quantified using Kallisto (Bray et al., 2016). Contigs were then binned using CONCOCT (Alneberg et al., 2014), which groups contigs together based on their sequence composition and frequency across multiple samples. Contigs with a similar frequency pattern (coverage) over multiple samples likely belong to the same microorganism. Microorganisms with a similar abundance pattern can be further separated on the sequence composition (tetranucleotide frequency) (Figure 1-7). Bins were assessed for contamination and completeness using CheckM, which uses lineage specific marker genes to assess bin quality (Parks et al., 2015). Ribosomal sequences were identified and taxonomically classified with Metaxa2 (Bengtsson-Palme et al., 2015). 16S rRNA gene sequences were also compared to the NCBI 16S ribosomal RNA database for bacteria and archaea using BLAST (Altschul et al., 1990). Genes were identified using Prodigal (Hyatt et al., 2010) and annotated with KO (KEGG Orthology) assignments using GhostKOALA (Kanehisa et al., 2016). Each query gene was assigned functional annotation as well as a taxonomic category according to the best-hit gene. This taxonomic information contributed to taxonomic placement of bins without a 16S rRNA gene sequence. Taxonomic classification was also performed using Kaiju which compares sequences to the NCBI BLAST non-redundant protein database (nr) (Menzel et al., 2016). Determination of metabolic pathways was aided with KEGG (Kanehisa et al., 2012) and MetaCyc (Caspi et al., 2018) databases. Bins that were considered high quality (>90% completeness, <5% contamination) or good quality (>75% completeness, <10% contamination) were considered to a MAG (Figure 1-7).

2.3.6 Metaproteomics

Protein was extracted at the Oak Ridge National Laboratory (Oak Ridge, Tennessee, USA) according to the protocol outlined previously (Chourey et al., 2010, 2013). Sample details are provided in the Supplementary Information (Table S4). Peptides from OL-KR46 were run by Multi-Dimensional Protein Identification Technology (MudPIT) (11 salt fractions). OL-KR11 and OL-KR13 peptides were run by mini-MudPIT (3 salt

26

fractions). Peptide fragmentation was measured on an Orbitrap-Elite mass spectrometer. The metagenomic dataset was used to construct a database of genes for each groundwater. Proteins were identified by searching the raw spectra (peptides) against the protein database to identify which genes were detected in the metaproteome. Proteins were clustered into protein groups (the same protein detected in multiple samples) post database search. Taxon specific proteomes were determined by identifying peptides that corresponded to genes that were binned into a MAG.

2.4 Groundwater Incubations

To collect biomass for incubation, groundwater was collected directly into sterile, N2 flushed 1 L glass serum bottles. Groundwater was transported in coolers to the laboratory where it was stored at 4°C. Within 1 week of sampling, groundwater was transferred into sterile N2 flushed 120 mL Wheaton glass serum bottles. The transfer was done with an N2 flushed syringe through a butyl rubber stopper. The groundwater was amended with an electron donor, electron acceptor and a nitrogen source from sterile anaerobic stock solutions. Serum bottles were incubated at 15°C and periodically subsampled with an N2 flushed syringe. DNA from groundwater incubations was extracted using the QIAamp UCP Pathogen Mini kit, using the pathogen lysis bead tubes (S) (QIAGEN, AG, Hombrechtikon, Switzerland). Extracted DNA was quantified on a Qubit 3.0 Fluorometer using the Qubit dsDNA High Sensitivity Assay kit (Thermo Fisher Scientific, Inc.). Extracted DNA was used for 16S rRNA gene amplicon sequencing as described in section 2.3.4.

2.4.1 Catalysed reporter deposition-fluorescence in situ hybridization (CARD-FISH).

To prepare samples for CARD-FISH, samples were fixed in paraformaldehyde (4% diluted in phosphate-buffered saline (PBS) to a final concentration of 1% [v/v]) and incubated overnight at 4°C. Fixed samples are then filtered through a 0.22 µm gold-coated polycarbonate membrane and washed three times with PBS. Hybridization buffer containing the CARD-FISH probes targeting specific taxonomic groups were added to the filters and incubated at 35°C overnight. All filters were counter stained with DAPI (4′6-diamidino-2-phenylindole). Target cells (cells that fluoresce when hybridised with the probe) were sought by viewing with an epifluorescence microscope.

2.4.2 Nanoscale secondary ion mass spectrometry (nanoSIMS)

Nitrogen (15N/14N) and carbon (13C/12C) isotopic compositions of individual cells were determined via nanoSIMS. The surface of the sample was sputtered with an ion beam, the resulting secondary ions released from the sample are characteristic of the chemical composition of the analysed sample. The secondary ions are collected, and the secondary ion beam is directed into a magnetic sector mass analyser. This information is then used to create a quantitative mass image of the analysed surface. With this, microorganisms that are metabolically active can be identified by the incorporation 15N and/or 13C into cell biomass.

27

3 RESULTS AND DISCUSSION OL-KR11_411

3.1 Hydrogeochemistry

Hydrogeochemical parameters were measured throughout the sampling period in 2016, during which time the drillhole was continuously pumped. Pressure data from before and after long-term pumping suggests that a connection to a shallower section at 125–127 m drillhole length (110–112 mbsl; Figure 1-1) occurred during long-term pumping in 2016. The possibility of mixing is therefore taken into consideration when interpreting data for OL-KR11_411. As pumping commenced and groundwater was drawn from within the fracture, an initial change in hydrogeochemistry was observed, followed by stabilization as pumping continued; chloride decreased from 2.8 g/L to 2.5 g/L (Figure 3-1A), total dissolved solids decreased from 5.2 g/L to ~4.7 g/L (Figure 3-1D), and sulfate decreased from 3.5 mM to 3.0 mM (Figure 3-1E). Two time points (June and October) could possibly represent a connection to a shallower fracture as chloride appears lower (Figure 3-1A) with a concurrent increase in sulfate (Figure 3-1E). The concentration of sulfate accounts for the total sulfur measured in OL-KR11_411 (cf. Figure 3-1E and 3-1G), which also decreases at the beginning of the sampling period. The concentration of sulfide detected (<5 µM; Figure 3-1F) does not account for the total loss of sulfate (~500 µM). This indicates that the change in sulfate concentration may result from pumping of the fracture, rather than the microbiological conversion of sulfate to sulfide. The concentration of sulfate measured (~3 mM) is similar to the concentration measured in 2014 (Table 1-1; 3.3 mM) indicating that sulfate has been relatively stable long-term.

28

Figure 3-1. OL-KR11_411 hydrogeochemical measurements. Error bars in (H) show standard deviation (n=2). Data in A–G were analysed by TVO. Data in (H) was analysed by EPFL. All data are provided in Supplementary Information (Table S5 & S6).

29

Figure 3-2. OL-KR11_411 hydrogeochemical measurements. DIC = dissolved inorganic carbon, DOC = dissolved organic carbon. Error bars show standard deviation in (A) methane, n=6, (B) δ13CCH4 ‰ VPDB, n=2, (D) δ13CDIC ‰ VPDB, n=2, (E) DOC, n=2. Data in (C) was analysed by TVO. Data in A, B, D, E and F were analysed by EPFL. All data are provided in Supplementary Information (Table S5 & S6).

30

Sulfate and sulfide measurements were also taken by EPFL, measurements were consistent with TVO measurements and also showed a decrease in sulfate at the beginning of the sampling period. Thiosulfate was measured but was not detected. Changes in the ratio between 34S/32S of sulfate (δ34SSO4 ‰ VCDT) can be used as an indicator of sulfate reduction by sulfate-reducing bacteria (SRB). This is because SRB preferentially use the light sulfur isotope in sulfate (32S) during sulfate respiration. This enriches the residual sulfate in the heavy isotope (34S) resulting in increased δ34SSO4 values. The isotopic ratio of sulfate was monitored over the sampling period (Figure 3-1H) to determine whether sulfate reduction may be on going in OL-KR11_411. The isotopic ratio varied between +23‰ and +26‰ over the nine-month time course but showed no overall increasing or decreasing trend. These δ34SSO4 values indicate an enrichment of 34S and are consistent with previous reports from Olkiluoto that show sulfate in brackish-SO4 type groundwater has an isotopic ratio of ~25‰ (Figure 3-3). This signature is thought to be the result of a past reduction event, when seawater infiltrated through organic rich bottom sediments (Posiva, 2013). The isotopic signature of OL-KR11_411 groundwater is therefore not indicative of further sulfate reduction.

Figure 3-3. Sulfate concentration vs. δ34SSO4 ‰ VCDT. The figure is taken from the Olkiluoto Site Description 2011 (Posiva, 2013). Processes influencing the isotopic signature of sulfate are indicated. Methane can be produced by both abiotic and biological processes. At Olkiluoto, the isotopic composition of methane indicates mixtures of thermally- and microbially-formed methane (Figure 3-4). Thermally-formed methane has an isotopic signature of ~-30‰. This ratio becomes lighter when mixed with microbially-formed methane (Figure 3-4). In OL-KR11_411, methane increased over the sampling period which coincided with a

31

decrease in δ13CCH4 (Figure 3-2A & 3-2B). A decrease in the isotopic signature of methane indicates an enrichment of light methane (12CH4). This can be an indication of microbial activity, as methane-producing archaea (methanogens) produce isotopically light methane. The change in methane concentration and δ13CCH4 could also represent a mixing effect due to pumping of this sampling section.

Figure 3-4. δ13CCH4 ‰ PDB vs. δ2HCH4 ‰ SMOW. The figure is taken from the Olkiluoto Site Description 2011 (Posiva, 2013). Processes influencing the isotopic signature of methane are indicated. DIC was relatively high in OL-KR11_411 (Figure 3-2C), consistent with observations of DIC in brackish-SO4 groundwater in Olkiluoto (Posiva, 2013). During the sampling period the concentration of DIC gradually increased (2.4 to 2.7 mM). When organic carbon is mineralised (broken down to CO2) by microbial activity, the isotopic signature of the organic carbon is transferred to the DIC pool. The δ13CDIC can therefore be used as an indicator of the origin of DIC, although it can be complex as many processes may contribute to the DIC pool (Figure 3-5). The δ13CDIC (~-15‰) in OL-KR11_411 (Figure 3-2D) is consistent with organic matter degradation as a source DIC in brackish-SO4 groundwater types (Figure 3-5). The isotopically light signature of δ13CDIC in brackish-SO4 groundwater is thought to derive from the same reduction event that resulted in isotopically heavier sulfate (Posiva, 2013).

32

Figure 3-5. δ13CDIC ‰ PDB vs. 14C, pmc. The figure is taken from the Olkiluoto Site Description 2011 (Posiva, 2013). Processes influencing the isotopic signature of inorganic carbon are indicated. DIC is an important source of carbon for autotrophic microorganisms. Autotrophs are microorganisms that fix inorganic carbon (e.g. carbon dioxide) to make organic compounds for biosynthesis and cell growth. Microorganisms that cannot fix inorganic carbon require an organic carbon source. Microorganisms that use organic carbon as their carbon source are called heterotrophs. Heterotrophic microorganisms often also oxidise organic carbon as a source of electrons (electron donor). Dissolved organic carbon (DOC) was present in OL-KR11_411 at <2 mM (Figure 3-2E). Organic compounds detected were acetone (~8 µM) and ethanol (~9 µM), both of which can be used as organic electron donors for microbial metabolism including sulfate reduction. The concentration of acetone and ethanol measured does not account for the concentration of DOC measured, so other organic compound(s) must also be present in OL-KR11_411. Glucose, butyrate, lactate, propionate and acetate were measured but not detected. Variability in the concentration of DOC (Figure 3-2E) was not reflected in the concentration of acetone or ethanol indicating that other organic compound(s) account for the observed change in concentration. Hydrogen is involved in many biogeochemical reactions and is a key component in anaerobic metabolism (Nealson et al., 2005). Hydrogen was detected in low concentration (<13 µM) in two samples from OL-KR11_411, but was not detected in replicate samples and was most often not detected (Figure 3-2F). Hydrogen is an important electron donor

33

in subsurface ecosystems, but it can be rapidly turned over and is often present at very low concentrations (Hoehler et al., 2001; Adhikari et al., 2016). In addition, hydrogen can be difficult to measure when samples are taken at ground level as the change in pressure results in degassing. The lack of measurable hydrogen in most OL-KR11_411 samples therefore does not rule out its presence in situ and the use of hydrogen as an electron donor.

3.2 Microbiological analyses

Cell counts indicated an abundance of 1.7 x 105 cells/mL in OL-KR11_411, consistent with previous reports of microbial abundance in groundwater at Olkiluoto (Pedersen et al., 2013). Samples were taken for microbial community analysis throughout the sampling period (Table 2-1). 16S rRNA gene amplicon libraries were produced for every sampling month. Six metagenomes were generated for OL-KR11_411, (one from June, three from September and two from November) with two metaproteomes (September and November) (Table 2-1). In addition, one metagenome was generated from the 0.1 µm pore size fraction collected in November.

3.2.1 Community composition

Analysis of 16S rRNA gene amplicon libraries showed that the microbial population captured on a 0.22 µm filter was predominantly bacterial (94.4 ± 3.0%) with Proteobacteria (64.5 ± 10.8%), Patescibacteria (8.9 ± 4.1%) and Omnitrophicaeota (6.1 ± 2.1%) representing the phyla with the greatest relative abundance. Archaeal sequences belonged to the candidate phylum Nanoarchaeaeota (4.2 ± 2.3%; formerly Euryarchaeota DHVEG-6 cluster) and Euryarchaeota (0.9 ± 0.6%). Omnitrophicaeota and Nanoarchaeaeota had a greater relative abundance in the 0.1 µm size fraction relative to the 0.22 µm (Figure 3-6). This is consistent with reports that these candidate phyla have small morphology and small genomes, possibly as an adaptation to life in the deep subsurface (Luef et al., 2015; Wu et al., 2016). Classes alpha-, beta-, delta- and epsilon- were represented within the Proteobacteria. Sulfidogenic bacteria (bacteria that produce sulfide) within the Deltaproteobacteria (Desulfobulbaceae, Desulfobacteraceae SEEP SRB1 and Desulfobacteraceae) were abundant (Figure 3-6). This appears contrary to the hydrogeochemical data, where little to no sulfide is observed (Figure 3-1F).

34

Figure 3-6. Genera detected in OL-KR11_411 groundwater by 16S rRNA gene amplicon sequencing (V4 region). All samples were collected on a 0.22 µm filter, except Nov (0.1) which was sequentially filtered on a 0.1 µm filter. For clarity, only genera with an average relative abundance of ≥1% over the sampling period are shown.

3.2.2 Metaproteogenomics

Metagenomes were generated to determine the metabolism of microorganisms detected in the groundwater. Seven metagenomes were generated from OL-KR11 samples (six from the 0.2 µm filtration and one from the 0.1 µm filtration; Table 2-1). Two metaproteomes were constructed (Table 2-1). Key genes from metabolic pathways of interest were sought within the metagenome and metaproteome of OL-KR11_411 (Figure 3-7).

35

Figure 3-7. Key genes from metabolic pathways of interest in OL-KR11_411 metagenomes. The abundance of genes is shown relative to the beta subunit of the single-copy RNA polymerase (rpoB) present in all microorganisms. Peptides from genes with an asterisk (*) were also detected in the metaproteome. Full gene names and descriptions are provided in the text. Sample IDs denote the sampling months June (J), September (S1, S2 and S3) and November (N2 and N3). The ID 0.1 denotes the metagenomic library generated from the 0.1 µm filtration taken in November. Sulfate reduction can occur by an energy-consuming assimilatory pathway whereby reduced sulfur compounds are used for biosynthesis of S-containing amino acids. This is in contrast to the energy-producing dissimilatory pathway which results in the excretion of sulfide as a product of anaerobic respiration. Dissimilatory sulfate reduction is therefore of interest at Olkiluoto as this metabolism impacts the groundwater hydrogeochemistry through the production of sulfide. Dissimilatory sulfite reductase catalyses the final step in sulfate reduction (Figure 3-8). The alpha and beta subunits of this enzyme (dsrAB) were sought within the metagenome as an indicator of dissimilatory sulfate reduction (Figure 3-7). The presence of dsrAB in the OL-KR11_411 metagenomes is consistent with the presence of SRB in 16S rRNA gene amplicon libraries (Figure 3-6). Corresponding peptides for DsrAB were also detected in the metaproteome showing that microorganisms with the enzyme dissimilatory sulfite reductase are active OL-KR11_411.

36

Figure 3-8. The pathway of dissimilatory sulfate reduction. Sulfate (SO4

2-) is activated by the enzyme sulfate adenylyltransferase (sat) to form adenylyl sulfate (APS). APS is directly reduced to sulfite (SO3

-) by adenylyl-sulfate reductase (apr). Sulfite is further reduced to sulfide (HS-) by dissimilatory sulfate reductase alpha (dsr). Dissimilatory sulfite reductase (dsr) is not exclusive to sulfate reduction. This enzyme can also be found in some microorganisms that carry out sulfide oxidation or disproportionation of intermediate oxidation state sulfur compounds (elemental sulfur, thiosulfate and sulfite) (Milucka et al., 2012). It is therefore necessary to consider the phylogeny of the organism the dissimilatory sulfite reductase comes from and to look for other genes required for sulfate reduction or sulfide oxidation e.g. the subunits dsrEFH are found in organisms that use dissimilatory sulfite reductase for sulfide oxidation (Dahl et al., 2008a). The Sox complex (soxXA, soxYZ, soxB and soxCD) mediates thiosulfate, sulfite, sulfur and sulfide oxidation (De Zwart et al., 1997) (Figure 3-9). The soxB gene from the Sox complex was therefore used an indicator of oxidative sulfur cycling (Figure 3-7). The soxB gene was detected in the OL-KR11_411 metagenome and corresponding peptides were detected in the metaproteome indicating that sulfide oxidising microorganisms (SOM) are active in OL-KR11_411.

Figure 3-9. The Sox multienzyme complex for thiosulfate (S2O3

2-) oxidation to sulfate (SO4

2-). Sox genes are shown in white boxes; sulfur-binding protein (soxYZ), haem enzyme complex (soxXA), sulfate thiohydroylase (soxB) and sulfane dehydrogenase (soxCD). The subunits soxCD are absent in some bacteria which utilise the sox pathway. The pathway with the absence of soxCD is shown by the dotted line. Pathway modified from the KEGG Pathway Database (Kanehisa et al., 2019). Hydrogenase proteins catalyse the reversible conversion of molecular hydrogen to protons and electrons (Figure 3-10). Hydrogenases with a [NiFe] active site are usually uptake hydrogenases i.e. they carry out hydrogen oxidation. They are therefore found in microorganisms that use hydrogen as an electron donor. A microorganism that oxidises hydrogen as an electron donor is called a hydrogenotroph. The beta subunits of [NiFe] hydrogenases (hyaB and hydB) were searched as a marker of hydrogenotrophic

37

respiration (Figure 3-7). Both beta subunits (hyaB and hydB) were detected in OL-KR11_411 but peptides were not detected in the metaproteome (Figure 3-7). This shows that there is the metabolic potential for hydrogen oxidation in OL-KR11_411.

Figure 3-10. Hydrogenotrophic respiration. A hydrogenase reversibly oxidises molecular hydrogen (H2) to protons and electrons. Methanogenesis (methane production) is carried out by anaerobic archaea (Figure 3-11). The pathway can also be reversed by anaerobic methane oxidising archaea (ANME) to catalyse the anaerobic oxidation of methane (AOM) (Figure 3-11). The alpha subunit of methyl-coenzyme M reductase (mcrA) catalyses production (methanogenesis) and consumption (AOM) of methane (Figure 3-11). The mcrA gene was therefore sought as an indicator of methane cycling (Figure 3-7). The mcrA gene had very low coverage in the metagenome of OL-KR11_411 suggesting that archaea are present in low abundance (Figure 3-7). Peptides for McrA were detected in the metaproteome indicating that archaea are active in OL-KR11_411. All of the mcrA genes identified in the metagenome were related to the methane-oxidising archaeon Candidatus Methanoperedens, suggesting that the mcrA enzyme operates in the methane-oxidising direction in OL-KR11_411.

Figure 3-11. Key reactions of the methanogenesis pathway from CO2 and H2. Some products are omitted for simplicity. Genes are shown in white boxes. The final step in methanogenesis is carried out by the key enzyme methyl-coenzyme M reductase (mcr). H4MPT = tetrahydromethanopterin. The enzyme formyl tetrahydrofolate synthetase (fhs) catalyses the ATP-dependent activation of formate in the Wood-Ljungdahl (WL) pathway (Figure 3-12). The WL pathway is found in a broad range of phylogenetic classes. In acetogens the pathway is

38

used in the reductive direction (production of acetate) for energy conservation and autotrophic carbon assimilation (Ragsdale and Pierce, 2008). Methanogens also use it in the reductive direction for CO2 fixation when growing on H2 and CO2, but they conserve energy by the conversion of H2 and CO2 to CH4 (Figure 3-11). During acetoclastic methanogenesis, part of the pathway is used as acetate is converted to acetyl-CoA by acetate kinase (ackA) and phosphotransacetylase (pta) (Figure 3-12) (Ragsdale and Pierce, 2008). Sulfate-reducing bacteria can utilise the Wood-Ljungdahl pathway in both directions to generate metabolic energy by oxidising acetate to CO2 or to fix CO2. The enzyme formyl tetrahydrofolate synthetase (fhs) was detected in both the metagenome and the metaproteome of OL-KR11_411 (Figure 3-7) indicating this pathway is utilised by microorganisms in OL-KR11_411.

Figure 3-12. The Wood-Ljungdahl pathway. The pathway can be utilised by acetogens, methanogens and sulfate-reducing bacteria. Genes are shown in white boxes, including the key enzyme formyl tetrahydrofolate synthetase (fhs). THF = tetrahydrofolate. The nitrate reductases nap and nar catalyse the first step in dissimilatory nitrate reduction to ammonia (DNRA) and denitrification to dinitrogen gas (Figure 3-12). The final step of denitrification (N2O to N2) is catalysed by the enzyme nitrous oxide reductase (nosZ) (Figure 3-12). These genes were sought as an indicator of nitrate metabolism. The genes subunits encoding napA, narG and nosZ were detected in OL-KR11_411 indicating the presence of nitrate reducing and/or denitrifying bacteria in OL-KR11_411. Furthermore,

39

peptides from NapA were detected in the metaproteome (Figure 3-7) suggesting nitrate reduction is active in OL-KR11_411.

Figure 3-13. Two pathways are shown; dissimilatory nitrate reduction to ammonia (DNRA) (NO3

- to NH3) and denitrification (NO3- to N2). The first step (nitrate reduction

to nitrite) is common to both pathways. Genes are shown in white boxes.

3.2.3 Metagenome assembled genomes (MAGs)

Assembled contigs from the OL-KR11_411 metagenomes were binned (Figure 1-7). The constructed bins were assessed for their completeness (how much of the genome they are estimated to cover) and their contamination (how many reads may be contaminating from another microorganism). Incomplete bins (<75%) and/or bins with a high level of contamination (>10%) were discarded. The remaining bins were considered to be metagenome assembled genomes (MAGs). MAGs were then searched for metabolisms of interest.

3.2.3.1 Sulfidogenic bacteria Sulfidogenic bacteria produce sulfide as a result of reductive sulfur cycling. This includes SRB as well as organisms which utilise other sulfur compounds such as sulfur, sulfite and thiosulfate. SRB reduce sulfate using the enzymes sulfate adenylyltransferase (sat), adenylyl-sulfate reductase (apr) and dissimilatory sulfite reductase (dsr) (Figure 3-8). SRB can be autotrophs (use inorganic carbon as a carbon source) or heterotrophs (use organic carbon as a carbon source). Autotrophic SRB often use hydrogen as an electron donor e.g., equation 1. During heterotrophic growth organic carbon can serve as both the electron donor and carbon source e.g., equation 2. Heterotrophic SRB can use a number of different organic compounds as an electron donor (e.g. lactate, butyrate, alcohols). SO4

2- + 4H2 + H+ → HS- + 4H2O Equation (1) SO4

2- + CH3COO- → HS- + 2HCO3- Equation (2)

40

Disproportionation is a chemolithotrophic process whereby sulfur compounds (sulfur, thiosulfate and sulfite) serve as both electron donor and acceptor, producing sulfide and sulfate (Finster, 2008). The ability to disproportionate thiosulfate (equation 3) and sulfite (equation 4) is relatively common among sulfate-reducing Deltaproteobacteria, but only a few strains have been reported to grow by the disproportionation of elemental sulfur (equation 5) (Thorup et al., 2017). For sulfur disproportionation to be thermodynamically favourable under standard conditions, the produced sulfide must be removed e.g. the sulfide can be scavenged by ferric iron (equation 6) (Janssen et al., 1996; Finster, 2008). S2O3

2- + H2O → SO42- + HS- + H+ Equation (3)

4SO3

2- + H+ → 3SO42- + HS- Equation (4)

4S0 + 4H2O → SO4

2- + 3HS- + 5H+ Equation (5) 3S0 + 4H2O + 2Fe2+ → SO4

2- + 2FeS + 8H+ Equation (6) Seven sulfate-reducing Deltaproteobacteria MAGs were recovered from the OL-KR11_411 metagenome (Table 3-1). All seven MAGs contained the complete pathway for dissimilatory sulfate reduction (sat, aprAB, dsrAB). Five of the seven Deltaproteobacteria MAGs expressed genes for dissimilatory sulfate reduction in their proteome, providing evidence that these microorganisms are actively respiring sulfate in Olkiluoto groundwater. Four of these MAGs also possessed the gene thiosulfate reductase (phsA) which catalyses the initial step in the disproportionation of thiosulfate (equation 3), where thiosulfate is cleaved into sulfite and sulfide. Table 3-1. Sulfidogenic MAGs recovered from OL-KR11_411. Presence (+) or absence (-) of genes within a MAG are noted. TCA = tricarboxylic acid cycle

Process SO42-

reduction

S2O32- disp.

Wood-Ljungdahl Pathway

Acetate oxidation H2 oxidation

Genes sat,

aprAB dsrAB

phsA acsB cooS citrate

synthase (TCA)

ACSS pta [NiFe] group

(1) Desulfocapsa + + + + + + - + (2) Desulfobacteraceae + - + + + + - + (3) Desulfobacula + + + + + + + + (4) Desulfurivibrio + + + + + + + + (5) Desulfobacterium + + + + + + + + (6) Desulfobulbaceae + + + + + + - + (7) Desulfobacteraceae + - + + + + - +

A near complete 16S rRNA gene sequence (1478 bp) was retrieved from a MAG classified as Desulfocapsa (Table 3-1). The 16S rRNA gene sequence shared 100% identity with a Desulfobulbaceae OTU detected by 16S rRNA gene amplicon sequencing (Figure 3-6). The longer 16S rRNA gene sequence recovered from the MAG shared greatest identity with Desulfocapsa thiozymogenes strain Bra2 (accession number NR_029306, 97% sequence identity).

41

Desulfocapsa species can grow by dissimilatory sulfate reduction and by the disproportionation of elemental sulfur, sulfite or thiosulfate with CO2 as their sole carbon source (Janssen et al., 1996; Finster et al., 2013). Accordingly, the Desulfocapsa proteome had expressed genes for both sulfate reduction and disproportionation as well as the WL pathway for carbon fixation. A genome sequenced strain of Desulfocapsa, Desulfocapsa sulfexigens strain SB164P1, is not able to grow by dissimilatory sulfate reduction despite encoding a complete set of genes necessary for sulfate reduction (Finster et al., 2013). This has also been reported for other sulfur disproportionating bacteria (Melton et al., 2016). It has been suggested that these genes could instead encode the reversed sulfate reduction pathway involved in the disproportionation (Frederiksen and Finster, 2003). Thus, from the genomic information, it is not possible to conclude if the OL-KR11_411 Desulfocapsa uses the dissimilatory sulfate reduction genes for sulfate reduction or disproportionation. A second MAG was classified as Desulfurivibrio (Table 3-1). This MAG contained a near complete 16S rRNA gene sequence (1482 bp) which corresponded to an OTU of Desulfurivibrio recovered by 16S rRNA gene amplicon sequencing (Figure 3-6). The closest relative was Desulfurivibrio alkaliphilus strain AHT2 (accession number NR_074971), which shared 90% sequence identity. Desulfurivibrio alkaliphilus can also grow by the disproportionation of sulfur (Sorokin et al., 2008; Poser et al., 2013; Melton et al., 2016). Thiosulfate reductase (phsA) was also recovered from the Desulfurivibrio MAG (Table 3-1). Two MAGs (Desulfobacteraceae and Desulfobacterium; Table 3-1) had 16S rRNA gene sequences that did not cover the variable region (V4) used for 16S rRNA gene amplicon sequencing. Thus, it was not possible to compare the sequences directly to our 16S rRNA gene amplicon results. A V1-V3 fragment of the 16S rRNA gene was recovered from the Desulfobacteraceae MAG. This gene fragment shared 86% identity with Desulfotalea psychrophilia strain LSv54 (accession number NR_028729). A V1 fragment of the 16S rRNA gene was recovered from Desulfobacterium. This gene fragment shared 90% 16S rRNA sequence bp identity with Desulfosarcina variabilis strain Montpellier (NR_044680). The low sequence identity of both 16S rRNA genes to those in the NCBI database suggests they may represent novel species. [NiFe] hydrogenases catalyse the reversible oxidation of hydrogen to protons and electrons (Figure 3-10). All seven SRB MAGs harboured Group 1 [NiFe] hydrogenases that catalyse hydrogen oxidation. This indicates that OL-KR11_411 SRB are able to use hydrogen as an electron donor. The hydrogenase small subunit (hyaA) from Desulfobacterium was detected in the metaproteome indicating that this microorganism is actively utilising hydrogen. No other [NiFe] hydrogenases from SRB were found in the metaproteome. The assimilation of inorganic carbon is known as carbon fixation, and microorganisms able to do this are called autotrophs. When SRB grow autotrophically with hydrogen, they often use the Wood-Ljungdahl (WL) pathway to incorporate CO2 into cell material.

42

The WL pathway is a series of reversible reactions used by many anaerobes for acetate synthesis (from CO2) or acetate oxidation (to CO2) (Figure 3-12). The net reaction of the WL pathway is shown in equation (7). 4H2 + 2HCO3

- + H+ ↔ CH3COO- + 4H2O Equation (7) The WL pathway employs the key enzymes carbon monoxide dehydrogenase (cooS) and acetyl-CoA synthase (acsB). All Deltaproteobacteria MAGs from OL-KR11_411 harboured these key genes from the Wood-Ljungdahl pathway (Table 3-1). The corresponding peptides from these genes were found in the proteome of Desulfocapsa, Desulfobacterium and Desulfobacteraceae indicating that this pathway is active. All Deltaproteobacteria MAGs also encoded the enzyme acetyl-CoA synthetase (ACSS), which oxidises acetate to acetyl-CoA. Acetate is a key intermediate in the anaerobic degradation of organic matter and can be used as an electron donor for SRB (equation 2). SRB can oxidise acetate to CO2 using the tricarboxylic acid cycle (TCA cycle). A complete, or near complete, TCA cycle was found in all SRB MAGs, including a key enzyme from the pathway called citrate synthase (Table 3-1).

3.2.3.2 Sulfide-oxidising bacteria The microbial oxidation of sulfide is a key reaction of the microbial sulfur cycle, whereby sulfide is converted to a higher oxidation state. Microorganisms that are able to use sulfur compounds as an energy source can use oxygen, nitrate and nitrite as terminal electron acceptors. Under anaerobic conditions, as found in the groundwater in OL-KR11_411, sulfide oxidation would be anaerobic and coupled to nitrate or nitrite (equations 8 and 9). 8NO3

- + 5HS- + 3H+ → 4N2 + 5SO42- + 4H2O Equation (8)

2NO2

- + HS- → N2 + SO42- + H+ Equation (9)

Anoxic sulfide oxidation can occur via different pathways (Figure 3-14). Sulfide can be oxidised by reversing the sulfate reduction pathway (Loy et al., 2009) (reverse of Figure 3-8). It can also be oxidised by sulfide:quinone oxidoreductase (sqr) which catalyses the oxidation of sulfide to intracellular sulfur (Griesbeck et al., 2002) (Figure 3-14). Sulfide (and other reduced sulfur compounds) can also be oxidised using the Sox complex (De Zwart et al., 1997; Poser et al., 2014) (Figure 3-14).

43

Figure 3-14. Oxidative sulfur cycling. Genes are indicated in white boxes; sulfide:quinone oxidoreductase (sqr), rhodanese (TST), dissimilatory sulfate reductase alpha and beta subunits (dsrAB), adenyl-sulfate reductase alpha and beta subunits (aprAB), sulfate adenyltransferase (sat), sulfite dehydrogenase (soeABC) and the Sox complex (soxXAYZBCD). The Sox pathway is also involved in the oxidation of other reduced sulfur compounds (thiosulfate, sulfur, sulfite and tetrathionate) Seven MAGs from OL-KR11_411 contained genes for the oxidation of reduced sulfur compounds (Table 3-2). This included two MAGs of Alphaproteobacteria (Rhizobiaceae and Brevundimonas), two Betaproteobacteria (Rhodocyclaceae and Sideroxydans), one Epsilonproteobacteria (Sulfurimonas), one Actinobacteria, and one Deltaproteobacteria (Desulfurivibrio). Both Betaproteobacteria MAGs (Rhodocyclaceae and Sideroxydans) encoded a complete sulfate reduction pathway (Table 3-2), which when operated in reverse catalyses the complete oxidation of sulfide to sulfate (Figure 3-14). Both MAGs encoded the dissimilatory sulfite reductase subunits dsrEFH on the same contig as dsrAB, which is essential for the function of reverse dsr (Dahl et al., 2008b). Both Betaproteobacteria MAGs also harboured a second sulfite oxidation pathway catalysed by the cytoplasmic sulfite oxidising enzymes soeABC (Watanabe et al., 2014) whereby sulfite is oxidised to sulfate (Figure 3-14). In addition, they possessed genes from the Sox multienzyme complex. The Sox system (soxXA, soxYZ, soxB and soxCD) mediates the oxidation of thiosulfate, sulfite, sulfur and sulfide (Friedrich et al., 2001) (Figure 3-9 & 3-14). Both Rhodocyclaceae and Sideroxydans lacked the genes encoding sulfur dehydrogenase (soxCD), which is common among sulfur oxidising Betaproteobacteria (Friedrich et al., 2005).

44

Table 3-2. OL-KR11_411 MAGs with genes involved in the oxidative sulfur cycle. Process HS- oxidation SO3-

oxidation Carbon fixation DNRA Denitri-

fication

Genes sat,

aprAB, dsrAB

Sox complex sqr soeABC PRK /

aclAB nar, nap, nir, nrf

nar, nap, nir, nor,

nos Rhizobiaceae - soxXAYZB

soxCD - - PRK narGHI

nirBD

narGHI norBC nosZ

Brevundimonas - - + - - narGHI napAB

narGHI napAB nirK norB

Rhodocyclaceae + soxAYZB - + PRK nirBD

napAB norBC nosZ

Sideroxydans + soxXAYZB - + PRK - nirS norBC

Sulfurimonas sat - +++ - aclAB napAB napAB Actinobacteria - - ++ - - - - Desulfurivibrio + - + - - napA,

nrfAH, nirB

-

The genus Sulfurimonas consists of a group of sulfur-oxidising bacteria that can grow with a broad range of reduced sulfur compounds (sulfide, sulfur and thiosulfate). The Sulfurimonas MAG from OL-KR11_411 did not contain a 16S rRNA gene sequence, but a Sulfurimonas OTU was identified in the 16S rRNA gene amplicon libraries from OL-KR11_411 (Figure 3-6). This OTU shared greatest sequence identity (98%) with Sulfurimonas denitrificans strain DSM 1251 (250 bp; accession number NR_074133). Sulfurimonas denitrificans grows with reduced sulfur compounds and hydrogen as electron donors, and nitrate or oxygen as an electron acceptor (Han and Perner, 2016). All sequenced Sulfurimonas genomes encode sulfide:quinone oxidoreductase (sqr) as well as Sox genes (soxXYZAB) (Han and Perner, 2016). The Sulfurimonas MAG recovered from OL-KR11_411 harboured three sqr genes, but unusually was lacking the sox gene cluster (Table 3-2). The Alphaproteobacteria MAG Rhizobiaceae encoded the complete Sox system (Table 3-2), which was also partially expressed in its proteome, indicating that Rhizobiaceae is actively oxidising reduced sulfur compounds in OL-KR11_411. The second Alphaproteobacteria MAG, Brevundimonas, also encoded a sqr gene but no other genes for sulfide oxidation were identified. It has been suggested that heterotrophic bacteria that contain only sqr may oxidise sulfide either as a strategy to detoxify sulfide, or as a sulfide conservation strategy, as sulfur is required for biosynthesis (Marcia et al., 2009; Xia et al., 2017). Alternatively it has been shown that bacteria with only sqr may collaborate in oxidising sulfide through interspecies transfer of polysulfide (𝑆𝑆𝑛𝑛

2−) (Xia et al., 2017). The Actinobacteria MAG encoded two sqr genes, but no other sulfide oxidising genes. As discussed above, the sqr gene may be involved in detoxification of sulfide rather than

45

primary energy metabolism. Pathways identified in the Actinobacteria MAG indicate that it grows by fermentation producing small organic compounds (section 3.2.2.5). During autotrophic growth, when sulfide is used as an energy source (equations 8 and 9) CO2 must be assimilated as a carbon source. Two carbon fixation pathways were identified among the sulfide-oxidising MAGs (Table 3-2). Rhodocyclaceae, Sideroxydans and Rhizobiaceae encoded near-complete pathways for carbon fixation via the reductive pentose phosphate (rPP) pathway, including the key enzyme phosphoribulokinase (PRK) (Table 3-2). This gene was found in the proteome of Rhodocyclaceae and Rhizobiaceae suggesting they are actively fixing CO2. Sulfurimonas harboured a different pathway for the fixation of CO2 called the reductive tricarboxylic acid cycle (rTCA), including the key enzyme ATP-citrate lyase (aclAB) (Table 3-2). Rhodocyclaceae, Sideroxydans and Sulfurimonas all harboured Group 3d [NiFe] hydrogenases which are bidirectional and can generate a reductant for hydrogenotrophic carbon fixation by coupling H2 oxidation to the reduction of nicotinamide adenine dinucleotide (NAD+) (Søndergaard et al., 2016). Rhizobiaceae harboured three hydrogenases; one [NiFe] Group 1d and two [NiFe] Group 2b. These are H2-uptake and H2-sensing hydrogenases, respectively. The H2-uptake hydrogenase is used for hydrogenotrophic growth. H2-sensing hydrogenases control expression of hydrogenase without providing energy for metabolism.

3.2.3.3 Nitrate reducing bacteria Under anoxic conditions, nitrate is the preferential electron acceptor for sulfide oxidation (equation 8). Many MAGs that harboured genes involved in the oxidation of sulfide also harboured genes for nitrate reduction (Table 3-2). Nitrate can be reduced to nitrogen (denitrification; equation 8) or to ammonium (dissimilatory nitrate reduction to ammonium ((DNRA) (equation 10). NO3

- + HS- + H+ → NH4+ + SO4

2- Equation (10) The periplasmic (nap) and cytoplasmic (nar) nitrate reductases reduce nitrate to nitrite (Figure 3-13). Nitrite can then be further reduced to ammonium (DNRA) or nitrogen (denitrification). The product of DNRA, ammonium, was detected in OL-KR11_411 (~4 µM) but neither nitrate nor nitrite are detected. Rhodocyclaceae and Rhizobiaceae possessed genes for both DNRA and denitrification (Table 3-2). Peptides from the alpha and beta subunits of the nitrate reductase NapAB were detected in the proteome of Rhodocyclaceae, showing active nitrate reduction in OL-KR11_411. Sideroxydans and Brevundimonas had an incomplete set of genes for denitrification (Table 3-2). Sideroxydans contained a contig with the nitrite reductase nirS (NO-forming) and nitric oxide reductase norBC (N2O-forming) but nitrous oxide reductase (nosZ, N2 forming), which catalyses the final step in denitrification (Figure 3-13), was missing. Brevundimonas was also missing nitrous oxide reductase (Table 3-2). Nitrate (napAB) and nitrite (nrfAH) reductases were also identified in three sulfidogenic MAGs; Desulfocapsa, Desulfurivibrio and Desulfobacteraceae. Many sulfate-reducing

46

bacteria that encode NapA can use nitrate as an alternative terminal electron acceptor (Marietou, 2016). Three other MAGs that harboured genes for nitrate reduction were recovered, but they did not encode genes for sulfide oxidation. These MAGs were classified as Hyphomonas, Sphingomonadales and Rhodoferax. Hyphomonas encoded the nitrate reductase (nar) and Sphingomonadales encoded genes for DNRA (nar and nir). Neither MAG encoded a pathway for carbon fixation, so these MAGs are inferred to be heterotrophic nitrate reducers (e.g. equation 11). Both MAGs encoded genes for acetate and lactate oxidation. 5CH3COO- + 8NO3

- + 13H+ → 10CO2 + 4N2 + 14H2O Equation (11) Rhodoferax had genes for both DNRA and denitrification. This MAG is discussed further in the following section.

3.2.3.4 Iron(III)-reducing bacteria The Rhodoferax MAG contained a 16S rRNA gene fragment which corresponded to the Comamonadaceae OTU detected by 16S rRNA gene amplicon sequencing (Figure 3-6), and shared greatest sequence identity with Rhodoferax ferrireducens strain T118 (99%, 1457 bp; accession number NR_074760). Rhodoferax ferrireducens can use nitrate and Fe(III) as electron acceptors, using organic compounds including acetate as a carbon source (Finneran et al., 2003) (equation 11 and 12). CH3COO- + 2H2O + 8Fe(III) → 2CO2 + 7H++ 8Fe(II) Equation (12) The Rhodoferax MAG encoded genes for both DNRA and denitrification, indicating that, like Rhodoferax ferrireducens, it has the ability to use nitrate as an electron acceptor. The Rhodoferax MAG also encoded the genes mtrA, mtrB and mtrC. The genes mtrA and mtrC encode c-type cytochromes which are involved in dissimilatory metal reduction, and mtrB is an associated outer membrane protein. Microorganisms that harbour mtrABC are able to reduce metals. Peptides from MtrB and MtrC were detected in the proteome of Rhodoferax, suggesting that it is actively using a metal (e.g. Fe(III)) as an electron acceptor. Rhodoferax possessed a complete TCA cycle, for the oxidation of organic carbon, including the key enzyme citrate synthase (CS), which was expressed in the proteome. It also possessed the enzymes acetyl CoA-synthetase (ACSS) and phosphotransacetylase (pta) for acetate oxidation.

3.2.3.5 Fermentation A Propionicimonas MAG belonging to the phylum Actinobacteria was recovered from OL-KR11_411. A full length 16S rRNA sequence was recovered from the MAG. Primer matching revealed that the primers used for 16S rRNA gene sequencing would not anneal to this 16S rRNA gene sequence. This explains why a corresponding OTU was not detected by 16S rRNA gene amplicon sequencing. Some members of the Propionibacteraceae family are known contaminants in subsurface 16S rRNA gene amplicon libraries, so their subsurface origin should be treated with caution. However, the OL-KR11_411 Propionicimonas was not detected in 16S rRNA gene amplicon libraries and was only recovered in the metagenome and metaproteome.

47

Propionicimonas expressed genes from the Embden-Meyerhof-Parnas (EMP) glycolysis pathway (conversion of glucose to pyruvate) and the TCA cycle in the proteome, suggesting that the metabolism is fermentative. Propionicimonas expressed genes to produce acetate and propionate. These short chain fatty acids (SCFA) can provide organic carbon for sulfate reduction (and other metabolisms).

3.3 Summary of metabolism in OL-KR11_411 groundwater

The microbial metabolisms identified in OL-KR11_411 are summarised in Figure 3-15. Sulfide is limited in OL-KR11_411, nevertheless, sulfidogenic microorganisms are detected in 16S rRNA gene amplicon libraries, and genes for sulfate reduction are expressed in the metaproteome. SRB in OL-KR11_411 possessed genes for the utilisation of both organic and inorganic compounds as electron and carbon sources. Four MAGs of Deltaproteobacteria also possessed a gene (phsA) involved in the disproportionation of sulfur compounds, and peptides from this gene were detected in the proteome of Desulfocapsa. Thiosulfate was not detected in OL-KR11_411, and the total sulfur concentration (Figure 3-1G) does not indicate that there is an abundance of other sulfur compounds that were not measured (e.g. sulfite and sulfur). This suggests that if other sulfur intermediates are being utilised by microorganisms in OL-KR11_411 they are present at low concentration and/or are rapidly utilised. The concentration of sulfide in OL-KR11_411 may be kept low by the activity of sulfide oxidising bacteria (SOB) which oxidise sulfide to sulfate (Figure 3-15). If sulfide is cycled back to sulfate by SOB, the concentration of sulfate may be maintained. While the fractionation associated with sulfide oxidation can be relatively small (Poser et al., 2014), the oxidation of sulfide to sulfate by sulfide-oxidizing bacteria, would transfer the light isotopic signature of sulfide produced by sulfate-reducing bacteria back to the sulfate pool (Mills et al., 2016). This could explain the observed values for δ34SSO4, which were not comparatively enriched in 34S, and the long-term stability of the concentration of sulfate. Disproportionation will also affect the fractionation between 34S/32S of sulfate. During disproportionation, 32S is preferentially used for the formation of sulfide, and 34S is preferentially used for the production of sulfate (Canfield et al., 1998). This results in 34S enriched sulfate, as with sulfate reduction, although the extent of the fractionation can be variable depending on sulfur compound, the enzymatic pathway and environmental conditions (Habicht et al., 1998; Poser et al., 2014). To maintain oxidative and reductive sulfur cycling in OL-KR11_411, an electron acceptor for sulfide oxidation must be maintained. Nitrate and nitrite were not detected in the groundwater, but ammonium (the product of DNRA, equation 10) was detected in low concentrations. Nitrate and nitrite can be produced by the anaerobic oxidation of ammonium. No known genes for ammonium oxidation were detected, but the enzymatic pathways for ammonium oxidation are not fully elucidated and genes are poorly represented in gene annotation databases. It is possible then, that ammonium in anaerobically oxidised to nitrate or nitrite in OL-KR11_411, but that it was not possible to identify the gene(s) responsible. The potential for iron reduction by Rhodoferax was also identified. Abiotically, ferrous iron (Fe2+) can react with sulfide (S2-) and precipitate as ferrous sulfide (FeS). This could potentially act as a sink for sulfide. Ferrous iron was detected in OL-KR11_411 (~1–2 µM). However, if a significant proportion of sulfide was being removed by

48

precipitation, this should be evident both in the concentration of sulfate and the sulfate isotopic signature. If sulfide is precipitating with ferrous iron, this process could be difficult to identify hydrogeochemically, if pumping is continuously bringing sulfate-containing groundwater into the fracture. Peptides from Candidatus Methanoperedens were identified in the metaproteome of OL-KR11_411 (section 3.2.2), suggesting that Candidatus Methanoperedens is active. However, the isotopic signature of DIC (Figure 3-2D) does not suggest that methane is a significant source of DIC in OL-KR11_411. Possible electron acceptors for AOM include sulfate, nitrate, iron and manganese. This is discussed in more detail in section 4.2.1.2 pertaining to OL-KR13_405. There was no genomic evidence for methanogenesis, suggesting that the increasing methane concentration (Figure 3-2A) and lighter δ13CCH4 (Figure 3-2B) is not microbially driven, and may result from pumping this section.

Figure 3-15. Metabolic pathways identified in OL-KR11_411. Electron acceptors (blue) and electron donors (green) are indicated. Possible electron acceptors are indicated in grey. Carbon sources (yellow) are only marked for SRB. Processes identified in the proteome are indicated by a solid line, unconfirmed processes or processes identified only in the metagenome only are indicated by a dashed line. SCFA = short chain fatty acids.

49

4 RESULTS AND DISCUSSION OL-KR13_405


The groundwater sampled from OL-KR13_405 is a mixture of sulfate-rich and methane-rich groundwater. Chloride (Figure 4-1A) and total dissolved solids (TDS) (Figure 4-1D) increased between February and June 2016. After June, the salinity remained stable. The concentration of sulfide and thiosulfate are relatively high in OL-KR13_405 (Figure 4-1F). The abundance of sulfide in the groundwater suggests that sulfate reduction is ongoing. The concentration of sulfate (Figure 4-1E), thiosulfate and sulfide (Figure 4-1F) account for the total sulfur detected in OL-KR13_405 (Figure 4-1G). The sulfate concentration did not show the same trend as salinity, and was relatively stable over the 9-month sampling period, with a slight decrease in the later sampling points (Figure 4-1E). The concentration of sulfide in the groundwater also decreased in later sampling points (Figure 4-1F) but this was not reflected in a loss of total sulfur (Figure 4-1G). The isotopic ratio of 34S/32S in sulfate was the most depleted in 32S of all the three groundwaters investigated in this study (average +44.55‰; Figure 4-1H). The δ34SSO4 is heavier than the value observed for Littorina derived sulfate and is consistent with further sulfate reduction (Figure 3-3). Methane is present in OL-KR13_405 at ~0.8 mM (Figure 4-2A). The concentration was variable over the sampling period and decreased between June and November. The δ13CCH4 of OL-KR13_405 groundwater indicates a mix of thermogenic and microbial methane (Figure 4-2B). The decreasing δ13CCH4 indicates an input of lighter methane (12C), either from microbial production of methane or from mixing with an isotopically lighter methane. The decrease in methane concentration and decrease in methane isotopic ratio could result from long term pumping of the fracture. Previous studies of Olkiluoto have shown that sulfide is often elevated where sulfate-rich and methane-rich groundwaters mix (Posiva, 2013), thus methane has been implicated as a potential electron donor for sulfate reduction. The isotopic signature of δ13CDIC can be used as an indication of the source of DIC (Figure 3-5). The isotopic signature of DIC in OL-KR13_405 decreased from -27 and -29‰ (Figure 4-2D). This is the lightest δ13CDIC of the three groundwaters investigated in this report. Negative values (≤-30‰) indicate that methane or another higher hydrocarbon could be a source of inorganic carbon (Figure 3-5). The values measured for OL-KR13_405 therefore indicate that CO2 from the anaerobic oxidation of methane could contribute a minor portion of the DIC. DOC was present in OL-KR13_405 at <0.6 mM (Figure 4-2E). Organic compounds detected were acetone (~6 µM) and ethanol (~3 µM), both of which can be used as organic electron donors. The concentration of acetone and ethanol does not account for the concentration of DOC measured, so other organic compound(s) must also be present in OL-KR13_405. Glucose, butyrate, lactate, propionate and acetate were measured but not detected.

50

Hydrogen was detected in OL-KR13_405 (Figure 4-2F). The concentration depleted over time and no hydrogen was detected in November. Long-term pumping may cause degassing as a result of pressure changes, and may account for the loss of hydrogen observed in OL-KR13_405. The abundance of hydrogen at this depth suggests it may be produced in situ, either biotically (e.g. from fermentation) or abiotically (e.g. water-rock interactions).

51

Figure 4-1. OL-KR13_405 hydrogeochemical measurements. Error bars show standard deviation in (H) (n=2). Data in A–G were analysed by TVO. Data in (H) and thiosulfate data in (F) was analysed by EPFL. All data are provided in Supplementary Information (Table S5 & S6).

52

Figure 4-2. OL-KR13_405 hydrogeochemical measurements. DIC = dissolved inorganic carbon, DOC = dissolved organic carbon. Error bars show standard deviation in (A) methane, n=6, (B) δ13

CH4 ‰ VPDB, n=2, (D) δ13CDIC ‰ VPDB, n=2, (E) DOC, n=2. Data in (C) was analysed by TVO. Data in A, B, D, E and F were analysed by EPFL. All data are provided in Supplementary Information (Table S5 & S6).

4.2 Microbial community analysis

Cell counts indicated an abundance of 2.3 x 105 cells/mL in OL-KR13_405. Eleven 16S rRNA gene amplicon libraries were generated from OL-KR13_405 groundwater throughout 2016 (Figure 4-3). Consistent with the observed sulfide (Figure 4-1F) and the isotopic sulfur isotopic signature of sulfate (Figure 4-1H), sulfate-reducing Deltaproteobacteria (Desulfurivibrio, Desulfobulbaceae, Desulfobacteraceae, Desulfatiglans and Desulfobacteraceae SEEP SRB1 were abundant (≥1%)) (Figure 4-3). Candidatus Methanoperedens was also among the abundant genera detected in OL-KR13_405 (Figure 4-3). Candidatus Methanoperedens is an anaerobic methane oxidising archaeon (Haroon et al 2013). Methane oxidation by Candidatus Methanoperedens could

53

contribute to the light isotopic signature of DIC observed in OL-KR13_405 groundwater (Figure 4-2D).

Figure 4-3. Genera detected in OL-KR13_405 groundwater by 16S rRNA gene amplicon sequencing (V4 region). All samples were collected on a 0.2 µm filter, except Nov (0.1) which was sequentially filtered on an 0.1 µm filter. For clarity, only genera with an average relative abundance of ≥1% over the sampling period are shown. To determine the metabolic activity of microorganisms detected in OL-KR13_405 groundwater, six metagenomes and two metaproteomes were examined. Key genes from metabolic pathways of interest were sought within the metagenome and the metaproteome (Figure 4-4). Peptides from dissimilatory sulfite reductase (DsrAB) were detected in the metaproteome confirming that sulfate reduction is active in OL-KR13_405. Hydrogen was detected in OL-KR13_405 (Figure 4-2F) and can be used as electron donor for sulfate reduction, as well as other metabolisms. Accordingly, genes for hydrogen oxidation (hyaB and hydB) were found in the metagenome and the associated peptides were identified in the metaproteome. Methyl-coenzyme M reductase (mcrA), which catalyses both the production and oxidation of methane (Figure 3-11) was present in the metagenome in relatively low abundance, nevertheless peptides from McrA were

54

detected in the metaproteome indicating that methane cycling is active in OL-KR13_405. The genomic potential for sulfide oxidation, nitrate reduction and denitrification were found in the metagenome but the activity of these metabolisms was not confirmed by the metaproteome.

Figure 4-4. Key genes from metabolic pathways of interest were searched in the metagenomic libraries from OL-KR13. The abundance of genes is shown relative to the beta subunit of the single-copy RNA polymerase (rpoB) present in all microorganisms. Genes with an asterisk (*) were also detected in the metaproteome. Full gene names and descriptions are provided in the text (section 3.2). Sample IDs denote the sampling months June (J), July (Ju), September (S2 and S3) and November (N1 and N2).


To assign metabolic pathways detected in the metagenome (Figure 4-4) to taxonomic groups, assembled contigs from OL-KR13_405 metagenomes were binned according to their composition and coverage (Figure 1-7). Incomplete bins (<75%) and/or bins with a high level of contamination (>10%) were discarded. The remaining bins were considered to be metagenome assembled genomes (MAGs). MAGs were then searched for metabolisms of interest.

55

4.2.1.1 Sulfate-reducing bacteria MAGs were searched for genes involved in reductive sulfur cycling. The genes for dissimilatory sulfate reduction (Figure 3-8) were found in eight MAGs recovered from OL-KR13_405 (Table 4-1).

Table 4-1. Sulfidogenic MAGs recovered from OL-KR13_405. Presence (+) or absence (-) of genes within a MAG are noted. TCA = tricarboxylic acid cycle.

Process SO42- reduction

S2O32- disp.



Genes sat aprAB dsrAB phsA acsB cooS

citrate synthase (TCA)

ACSS pta [NiFe] group

Desulfocapsa + + + + + + - + Desulfuri-vibrionaceae

+ + + + + + + +

Pseudo-desulfovibrio

+ + - + + + + +

Desulfarculus + + + + - + - + Desulfurivibrio + + + + + + + + Desulfobacterales + - + + + + - - Thermo-desulfovibrio 1

+ - + + - + - +

Thermo-desulfovibrio 2

+ - + + - + - +

Desulfocapsa shared greatest sequence identity (97%) with Desulfocapsa thiozymogenes strain Bra2 (1478 bp; NR_029306). This OTU was also found in OL-KR11_405 (Figure 3-6). Desulfocapsa thiozymogenes Bra2 is able to disproportionate thiosulfate, sulfite and elemental sulfur (equations 3–5) and couple the oxidation of alcohols to sulfate reduction (equation 13) (Janssen et al., 1996). 2CH3CH2OH + SO4

2- → 2CH3COO- + HS- + 2H2O + H+ Equation (13) The complete sulfate reduction pathway was found in the Desulfocapsa proteome, indicating that this genus is actively reducing sulfate in OL-KR13_405. Alcohols are oxidised by alcohol dehydrogenases. Three alcohol dehydrogenases were detected in the Desulfocapsa MAG, peptides from two of which were also detected in the proteome. This indicates that Desulfocapsa can reduce sulfate coupled to ethanol oxidation as shown in equation 13. Ethanol was detected in low concentrations in OL-KR13_405 (~3 µM). Peptides from thiosulfate reductase (phsA), tetrathionate reductase (ttrB) and rhodanese-related sulfurtransferase were detected in the Desulfocapsa proteome indicating that Desulfocapsa also disproportionates sulfur compounds. When Desulfocapsa thiozymogenes Bra2 grows by disproportionation of inorganic sulfur compounds, CO2 can act of the sole carbon source. The Desulfocapsa MAG from OL-KR13_405 encoded the WL pathway for carbon fixation (Table 4-1), which was also partially expressed in the proteome indicating that Desulfocapsa can use CO2 as a carbon source.

56

Hydrogen did not support the growth of Desulfocapsa thiozymogenes Bra2 (Janssen et al., 1996), however hydrogen oxidation genes (hyaAB) were found in the OL-KR13_405 Desulfocapsa MAG suggesting that it is capable of utilising hydrogen. Peptides from Desulfocapsa hydrogenases were not detected in the proteome. The 16S rRNA gene sequence recovered from the Desulfurivibrionaceae MAG (778 bp) shared low sequence identity (88%) with known sulfate reducers. Few proteins from this MAG were detected in the metaproteome, however those that were (sat, aprB and dsrA) indicate that Desulfurivibrionaceae is an active sulfate reducer. Desulfurivibrionaceae also encoded the Wood-Ljungdahl pathway for carbon fixation and genes for hydrogen oxidation (hyaAB) as well as genes for organic carbon oxidation (Table 4-1). The 16S rRNA gene sequence recovered from the Pseudodesulfovibrio MAG shared 99% sequence identity with Pseudodesulfovibrio aespoeensis strain Aspo-2 (1472 bp; NR_074871). The corresponding Pseudodesulfovibrio OTU was present in low abundance in 16S rRNA gene amplicon libraries (0.2 ± 0.1%). Pseudodesulfovibrio is a hydrogenotrophic (hydrogen oxidising) sulfate reducer isolated from 600 m depth at Äspö, which forms part of the Fennoscandian Shield in Sweden (Motamedi and Pedersen, 1998; Pedersen et al., 2014). This OTU was also detected in OL-KR46_570 (Table 5-1). The Pseudodesulfovibrio MAG harboured genes for hydrogen oxidation (Table 4-1). The hydrogenase subunits hydAB were encoded in the MAG and corresponding peptides from the alpha subunit were detected in the proteome. Few proteins from Pseudodesulfovibrio, were detected in its proteome, likely because of its low abundance in the groundwater. Nevertheless, peptides from the beta subunit of adenyl-sulfate reductase (aprB) were detected in the proteome. Although the key enzyme carbon monoxide dehydrogenase (coos) from the WL pathway was found in the Pseudodesulfovibrio MAG (Table 4-1), the WL pathway was incomplete. Pseudodesulfovibrio had no complete carbon fixation pathways. This is contrary to the Pseudodesulfovibrio MAG constructed from OL-KR46_570 groundwater which harboured the reductive pentose phosphate pathway for carbon fixation (Table 5-1). The OL-KR13_405 Pseudodesulfovibrio MAG encoded genes for acetate oxidation to acetyl-CoA, which could be further oxidised to CO2 using the citrate cycle (Table 4-1). A Deltaproteobacteria MAG contained a 241 bp 16S rRNA gene sequence of the V1 region. This did not match the region of the 16S rRNA gene sequenced in amplicon libraries (V4). The V1 fragment shared 83% identity with Desulfarculus baarsii strain DSM 2075 (accession number NR_074919). Desulfarculus baarsii is a sulfate-reducing bacteria that can grow using organic and inorganic carbon sources (Sun et al., 2010). The only protein detected in the Desulfarculus proteome was an ATPase, which catalyses the conversion of ATP into ADP to release energy. Desulfarculus possessed the sulfate reduction pathway, Wood-Ljungdahl pathway and TCA cycle. It also encoded the genes for hydrogen oxidation (hydAB) and the utilisation of fatty acids (butyrate, propionate, acetate). This suggests Desulfarculus is able to use both organic and inorganic compounds for growth with sulfate. Desulfurivibrio had the complete sulfate reduction pathway and corresponding peptides from Sat and AprB were also expressed in the proteome. Desulfurivibrio species are reported to grow chemolithoautotrophically by the disproportionation of elemental sulfur

57

(Melton et al., 2016) (equation 5). Thiosulfate reductase (phsA) and tetrathionate reductase (ttrB) were encoded in Desulfurivibrio and may catalyse disproportionation of sulfur compounds. Some genes from the WL pathway, including the key enzyme cooS were detected in the proteome. Hydrogen oxidation genes (hyaAB) were found in the genome. The alpha and beta subunits of adenylyl-sulfate reductase (AprAB) were detected in the proteome of Desulfobacterales indicating active sulfate reduction by this SRB. Desulfobacterales was the only SRB MAG that did not encode hydrogen oxidation genes (hyaAB or hydAB). The Wood-Ljungdahl pathway was detected, including the key enzymes cooS and acsB. The TCA was complete along with genes for the utilisation of fatty acids. Two sulfate-reducing MAGs from OL-KR13_405 were classified as the genus Thermodesulfovibrio from the phylum Nitrospirae (Table 4-1). Cultured Thermodesulfovibrio are thermophilic (growth above 40°C) and can use sulfur compounds and Fe(III) as electron acceptors (Sekiguchi et al., 2008). Both MAGs harboured the complete pathway for sulfate reduction, but few proteins were detected in their proteome. The alpha subunit of adenyl-sulfate reductase (aprA) was detected in the proteome of Thermodesulfovibrio 1. Both MAGs harboured the WL pathway, and acsB was detected in the proteome of Thermodesulfovibrio 2. Both encoded the hydrogen oxidation genes hyaAB.

4.2.1.2 Methane cycling archaea Methane coenzyme M reductase (McrA) which catalyses the production/consumption of methane (Figure 3-11) was found in the proteome of OL-KR13_405 (Figure 4-4). None of the MAGs retained contained the mcrA gene, or the pathway for methanogenesis. Low quality bins that had been discarded were therefore searched for genes encoding Mcr subunits that correspond to those detected in the metaproteome. Seven low quality bins of Candidatus Methanoperedens were recovered ranging from 42.9−94.4% completion and 2.6−34.1% contamination. The bin with the lowest contamination was selected for further analysis, this bin was 42.9% complete. Despite being incomplete, the bin encoded the full pathway for methane oxidation/methanogenesis, which was partially expressed in the proteome. Candidatus Methanoperedens is an ANME-2d methane oxidising archaeon (Haroon et al 2013). Candidatus Methanoperedens perform AOM using nitrate, nitrite, iron (III), sulfate or manganese as a terminal electron acceptor (equations 14−18) (Haroon et al., 2013; Cai et al., 2018; Ino et al., 2018). 5CH4 + 8NO3

- + 8H+ → 5CO2 + 4N2 + 14H2O Equation (14) 3CH4 + 8NO2

- + 8H+ → 3CO2 + 4N2 + 10H2O Equation (15) CH4 + 8Fe3+ + 2H2O → CO2 + 8Fe2+ + 8H+ Equation (16) CH4 + 4MnO2+ 7H+ → HCO3

- + 4Mn2+ + 5H2O Equation (17)

58

CH4 + SO42- → HCO3

- + HS- + H2O Equation (18) Candidatus Methanoperedens species reported to reduce nitrate encode an ammonium-forming cytochrome c nitrite reductase (nrfAH) and nitrate reductase (narGHI) (Berger et al., 2017). There were no nitrate-reducing genes in any of Candidatus Methanoperedens bins, although they were incomplete. Nitrate reducing genes were therefore sought from the metagenome, but none were related to Candidatus Methanoperedens. This suggests that Candidatus Methanoperedens from OL-KR13_405 does not have the metabolic ability to utilise nitrate as an electron acceptor for AOM. Candidatus Methanoperedens are not known to associate with SRB for sulfate reduction (as reported for other ANME archaea (Boetius et al., 2000; Orphan, 2001)). However, it has been proposed that Candidatus Methanoperedens may utilise assimilatory sulfate reduction genes to perform dissimilatory sulfate reduction (Ino et al., 2018). The OL-KR13_405 Candidatus Methanoperedens MAG encoded sulfate adenylyltransferase (sat), which can be used in both assimilatory and dissimilatory sulfate reduction. The MAG also encoded other assimilatory-type sulfate reduction genes. Candidatus Methanoperedens which utilise Fe(III) or Mn(IV) as an electron acceptor contain multiple multiheme c-type cytochromes involved in electron transfer during dissimilatory metal reduction. Putative multiheme c-type cytochromes were identified in the OL-KR13_405 Candidatus Methanoperedens MAG suggesting that is may be able to utilise Fe(III) or Mn(IV) as an electron acceptor for methane oxidation.

4.2.1.3 Acetogenic bacteria A MAG related to Acetobacterium was identified and presumed to represent an acetogen. Acetogens are obligately anaerobic bacteria that produce acetate. Acetate can be generated from CO2 (equation 19) or via the fermentation of sugars (equation 20). 4H2 + 2CO2 → CH3COO- + H+ + 2H2O Equation (19) C6H12O6 → 3CH3COO- + 3H+ Equation (20) Few peptides were detected in the proteome of Acetobacterium from OL-KR13_405. Peptides that were detected corresponded to genes from the WL pathway (including fhs; Figure 3-12) and a NADP-reducing hydrogenase subunit, suggesting Acetobacterium is actively producing acetate from inorganic carbon and hydrogen (equation 19). No acetate was detected in OL-KR13_405.

4.2.1.4 Fermentation An Actinobacteria MAG classified in Propionicimonas was detected in OL-KR13_405. This MAG was also detected in OL-KR11_411 (section 3.2.3.4). Peptides detected in the proteome of Propionicimonas included ABC transporters for sugars as well as the complete set of enzymes to oxidise pyruvate to acetate. This MAG is presumed to

59

represent a fermenter, capable of fermenting complex organic carbon and producing smaller organic compounds.

4.3 Summary of metabolism in OL-KR13_405 groundwater

The microbial metabolisms identified in OL-KR13_405 are summarised in Figure 4-5. Sulfate-reducing bacteria are abundant in OL-KR13_405 and marker genes from the sulfate reduction pathway were detected in the metaproteome confirming the hydrogeochemical data showing that sulfate reduction is an active process in OL-KR13_405. Hydrogen was often detected in OL-KR13_405, and the presence of respiratory hydrogenases in SRB MAGs suggests that they are capable of utilising H2 as an electron donor and CO2 as a carbon source during growth on sulfate (equation 1). Genes for the oxidation of organic compounds were also identified. Organic carbon can be used as both electron donor and carbon source. Methane has been proposed as a potential electron donor for sulfate reduction in groundwater at Olkiluoto (Posiva, 2013; Bomberg et al., 2015). The oxidation of methane would result in heavier isotopic methane, as the light isotope is preferentially removed by microbial methane oxidation. Over the sampling period the δ13CCH4 in OL-KR13_405 groundwater decreased from -42 ‰ to -44 ‰, indicating that lighter methane is contributing to the CH4 pool. However, no evidence for methanogenesis was found in OL-KR13_405; no known methanogens were detected in the 16S rRNA gene amplicon libraries, and no mcrA genes related to known methanogenic archaea were found in the metagenome. The change in isotopic signature of methane could instead result from the long-term pumping of this section. If methane is being oxidised (equations 14–18), the low δ13C content of CH4 would be transferred to DIC, resulting in more negative values for δ13CDIC. The isotopic signature of DIC in OL-KR13_405 is light (-28.7 ‰) suggesting that methane oxidation could contribute to a minor proportion of the DIC in OL-KR13_405. Abundant peptides detected in the OL-KR13_405 metaproteome were related to Candidatus Methanoperedens, a methane-oxidising archaeon. These included peptides from the marker gene mcrA, along with other genes from the methane oxidation/methanogenesis pathway. This indicates that this pathway is active in OL-KR13_405. Electron acceptors for AOM include nitrate, nitrite, sulfate, iron and manganese (equations 14–18). ANME related to Candidatus Methanoperedens nitroreducens have been shown to use nitrate, iron and manganese as electron acceptors for methane oxidation (Haroon et al., 2013; Ettwig et al., 2016). There was no metagenomic evidence that the OL-KR13_405 Candidatus Methanoperedens could perform nitrate dependent methane oxidation, as no genes for nitrate reduction were found related to this genus. This has been previously reported for an archaeon related to Candidatus Methanoperedens nitroreducens found in groundwater enriched in sulfate and methane from 300 metres depth in the granitic subsurface of Japan (Ino et al., 2018). In that study, short-term incubation experiments with 13C-labelled methane indicated that the ANME-2d archaeon coupled AOM with sulfate reduction. The authors proposed that

60

the archaeon uses the genes from the assimilatory pathway for dissimilatory sulfate reduction. Genes for assimilatory sulfate reduction were also detected in the OL-KR13_405 Candidatus Methanoperedens MAG, although no corresponding peptides were detected in the proteome. It is not possible to conclude if they could be used for dissimilatory sulfate reduction in this MAG as proposed by Ino et al., (2018).

Figure 4-5. Metabolic pathways identified in OL-KR13_405. Electron acceptors (blue) and electron donors (green) are indicated. If the electron acceptor is unknown, possibilities are indicated in grey. Carbon sources (yellow) are only marked for SRB. Processes identified in the proteome are indicated by a solid line, processes identified in the metagenome only are indicated by a dashed line. SCFA = short chain fatty acids.

61

5 RESULTS AND DISCUSSION OL-KR46


OL-KR46_570 is the deepest of the three drillholes investigated (528.2−531.5 mbsl) with the highest concentration of sulfide (>1 mM; Figure 5-1F). Sulfate is present in OL-KR46_570 as a result of mixing with shallower sulfate-rich groundwater that occurred when the drillhole was open over 6 years after drilling in 2007. Now that the fracture is isolated with packers, the groundwater is slowly recovering from the influx of sulfate-rich water and sulfate is depleted over time (Figure 5-1E). In July 2016, during the sampling period, there was a packer leakage which caused some dilution with water from the upper part of the drillhole. This event is reflected in the hydrogeochemical data for 01/08/2016 e.g. sulfate increases (Figure 5-1E) and chloride is decreases (Figure 5-1A). The δ34SSO4 VCDT was monitored over the sampling period (Figure 5-1H). δ34SSO4 varied between +25‰ and +31‰ over the time course but showed no increasing or decreasing trend overall. The isotopic ratio of +31‰ indicates that the light isotope (32S) has been depleted by microbial activity. Thiosulfate, a key intermediate in the sulfur cycle, was also abundant in OL-KR46_570 (Figure 5-1F). As the deepest sample in this study, methane is highest in OL-KR46_570 (Figure 5-2A). The δ13CCH4 of methane was ~-32‰ (Figure 5-2B). Methane is heavier in OL-KR46_570 than OL-KR11_411 and OL-KR13_405, indicative of a greater influence from thermally formed methane (Figure 3-4). DOC was high (Figure 5-2E; 4.7 ± 0.7 mM) and is contributed to by organic compounds; acetate (1.5 ± 0.3 mM), acetone (8.1 ± 3.6 μM), ethanol (28.6 ± 16.7 μM), methanol (65.4 ± 22.7 μM) and 2-butanol (3.7 μM ± 1.3 μM). These organic compounds account for ~65% of the total carbon mass of measured DOC. The remaining fraction can be accounted for by the presence of n-butylbenzenesulfonamide (30 mg/L; measured 05–07/09/2018). N-butylbenzenesulfonamide can leach into groundwater from polyamide tubing used in the drillhole. It is unlikely that this compound would be anaerobically degraded in situ. Hydrogen, formate, lactate, propionate, pyruvate, butyrate and glucose were measured but not detected. DIC was very low in OL-KR46_570 and was sometimes below detection limit (Figure 5-2C). Due to the low DIC concentration, it was not always possible to measure δ13CDIC (Figure 5-2D). Four measurements were taken (two in May and two in June). DIC increased in OL-KR46_570 after the packer leakage in July. Following the influx of DIC (Figure 5-2C), DIC is fully consumed and the concentration of acetate increases (Figure 5-2F).

62

Figure 5-1. OL-KR46_570 hydrogeochemical measurements. Error bars show standard deviation in (H), (n=2). Data in A–G were analysed by TVO. Data in (H) and thiosulfate data in (F) was analysed by EPFL. All data are provided in Supplementary Information (Table S5 & S6).

63

Figure 5-2. OL-KR46_570 hydrogeochemical measurements. DIC = dissolved inorganic carbon, DOC = dissolved organic carbon. Error bars show standard deviation; (A) methane, n=6, (B) δ13CCH4 ‰ VPDB, n=2, (D) δ13CDIC ‰ VPDB, n=2, (E) DOC, n=2. For DIC, 0 values are below detection limit (1.6 mg/L). Data in (C) was analysed by TVO. Data in A, B, D, E and F were analysed by EPFL. All data are provided in Supplementary Information (Table S5 & S6).

64

5.2 Microbial community analysis

Cell counts indicated an abundance of 1.9 ∙105 cells/mL in OL-KR46_570. The diversity of the microbial community in OL-KR46_570 was low. Sulfide was abundant in OL-KR46_570 groundwater (Figure 5-1F). Accordingly, sulfate-reducing bacteria were abundant in OL-KR46_570 groundwater (Figure 5-3). Sulfate-reducing bacteria (Desulfomicrobium, Candidatus Desulforudis, Desulfosporosinus, Dethiosulfatibacter, Desulfarculaceae and Desulfovibrio) represented ~67% of the 16S rRNA gene amplicon libraries (Figure 5-3). An acetogenic bacterium, Acetobacterium, was also detected in OL-KR46_570 (Figure 5-3) and could contribute to the abundance of acetate observed in OL-KR46_570 groundwater (Figure 5-2F). The relative abundance Acetobacterium was greatest in August, after a packer leak introduced inorganic carbon to the fracture (Figure 5-2C). A methanogenic archaeon, Methanobacteriaceae, was also relatively abundant in the groundwater (Figure 5-3).

65

Figure 5-3. Genera detected in OL-KR46_570 groundwater by 16S rRNA gene amplicon sequencing (V4 region). All samples were collected on a 0.2 µm filter. For clarity, only genera with an average relative abundance of ≥1% over the sampling period are shown. Three metagenomic libraries were generated from OL-KR46_570 (sampling months March, June and August) with one metaproteome (August). Key genes from metabolic pathways of interest were searched within the metagenomic libraries (Figure 5-4). Dissimilatory sulfite reductase (dsrAB), an indicator of dissimilatory sulfate reduction, was found in both the metagenome and the metaproteome, confirming that sulfate reduction is active in OL-KR46_570. The potential for oxidative sulfur cycling was indicated by the presence of the sulfur-oxidising protein soxB. This gene was detected in low relative abundance and was not detected in the metaproteome suggesting that sulfide oxidation may not be a significant metabolism in OL-KR46. [NiFe] hydrogenases (hyaB and hydB) were searched as a marker of hydrogenotrophic respiration. Both beta subunits genes were detected in the metaproteome indicating that hydrogen oxidising microorganisms are active in OL-KR46_570, despite hydrogen not being detected. This suggests microorganisms rapidly turnover hydrogen. The alpha subunit of methyl-coenzyme M reductase (mcrA), which catalyses both the production and consumption of methane was sought as an indicator of

66

methanogenic/methanotrophic activity. The mcrA gene was detected in both the metagenome and metaproteome of OL-KR46_570 indicating that methane cycling in ongoing. Nitrate reductases (napA and narG) and nitrous reductase (nosZ) were sought as indicators of reductive nitrogen cycling. The metagenome of OL-KR46_570 showed metabolic potential nitrate reduction (Figure 5-5), but this was not confirmed in the metaproteome.

Figure 5-4. Key genes from metabolic pathways of interest were searched in the metagenomic libraries from OL-KR46_570. The abundance of genes is shown relative to the beta subunit of the single-copy RNA polymerase (rpoB) present in all microorganisms. Genes with an asterisk (*) were also detected in the metaproteome. Full gene names and descriptions are provided in the text (section 3.2). Sample IDs denote the sampling months March (M), June (J) and August (A).


To assign metabolic pathways detected in the metagenome (Figure 4-4) to taxonomic groups, assembled contigs from OL-KR46_570 metagenomes were binned according to their composition and coverage (Figure 1-7). Incomplete bins (<75%) and/or bins with a high level of contamination (>10%) were discarded. One lower quality bin was retained (68% completeness and 0.7% contamination). This bin was retained as it had low contamination and shared the same taxonomy as a good quality MAG recovered from groundwater from OL-KR13_411. The good quality MAG from OL-KR13_405 was therefore used to check for genes that were missing in the lesser-complete OL-KR46_570 MAG. The remaining bins were considered to be metagenome assembled genomes (MAGs). MAGs were then searched for metabolisms of interest.

67

5.2.1.1 Sulfate-reducing bacteria Four MAGs containing genes for dissimilatory sulfate reduction (sat, aprAB, dsrAB) were recovered (Table 5-1). These were classified as Desulfomicrobium, Desulfarculaceae, Candidatus Desulforudis and Pseudodesulfovibrio. The Deltaproteobacteria MAGs (Desulfomicrobium, Desulfarculaceae and Pseudodesulfovibrio) all harboured thiosulfate reductase (phsA) which was also expressed in their proteome. Thiosulfate reductase catalyses the initial step in the disproportionation of thiosulfate and the gene is common among Deltaproteobacteria. The Firmicutes MAG (Candidatus Desulforudis) did not contain thiosulfate reductase. Table 5-1. MAGs of sulfate-reducing bacteria in OL-KR46

Process SO42- reduction

S2O32- disp.



Genes sat aprAB dsrAB phsA acsB cooS

citrate synthase (TCA)

ACSS pta [NiFe] group 1

Desulfomicrobium + + + + + + + + Desulfarculaceae + + + + - + + +

Candidatus Desulforudis + - + + - + - +

Pseudo-desulfovibrio + + - + + + + +

The WL pathway can be used for carbon fixation and the complete oxidation of organic carbon (Figure 3-12). All SRB expressed the key enzyme carbon-monoxide dehydrogenase (coos) in their proteome. Only Desulfarculus and Desulforudis harboured the complete WL pathway, including the second key enzyme, acetyl-CoA synthase (acsB) (Table 5-1). Pseudodesulfovibrio encoded an alternative pathway for carbon fixation via the rPP pathway. Genes from the rPP pathway were expressed in the proteome of Pseudodesulfovibrio, including the key enzyme phosphoribulose kinase (PRK). Acetyl-CoA produced via the WL or rPP pathway from CO2 can be incorporated into biomass. All SRB MAGs also encoded genes necessary for the reversible oxidation of acetate to acetyl-CoA (Table 5-1) which can be used as an electron donor (and carbon source) coupled to sulfate reduction (equation 2). All SRB MAGs encoded Group 1 [NiFe] hydrogenases in their proteome showing that hydrogen is actively utilised as an electron donor in OL-KR46_570. Candidatus Desulforudis is a member of the Firmicutes, the phylum to which all endospore-forming bacteria belong. Accordingly, the Candidatus Desulforudis MAG contained a number of proteins encoding genes for spore maturation, sporulation and germination. Sporulation can confer a survival advantage in oligotrophic environments, as spore formation can enable bacteria to survive long periods of nutrient limitation in a dormant state. Firmicutes were abundant in OL-KR46_570 (~45%) compared to shallower groundwaters OL-KR11_411 and OL-KR13_405 (<5%).

68

5.2.1.2 Acetogenic bacteria Acetogens are able to produce acetate from CO2 and H2 by the reactions of the WL pathway (equation 19). They can also produce acetate from the oxidation of organic substrates via glycolysis (EMP pathway) (equation 20). A MAG related to the acetogenic Acetobacterium was identified in OL-KR46. The genes fhs and acsB were expressed in the proteome of Acetobacterium, suggesting acetate production via the WL pathway is active. Acetobacterium also encoded a near complete glycolysis pathway, which was partially expressed in the proteome. This suggests Acetobacterium generates acetate from both inorganic and organic compounds (equations 19 & 20). This is consistent with the abundance of acetate in OL-KR46_570 (Figure 5-2F).

5.2.1.3 Methane cycling archaea The alpha subunit of methyl-coenzyme M reductase (mcrA) catalyses both the production of methane (methanogenesis) and the anaerobic oxidation of methane (AOM) (Figure 3-11). The mcrA gene had low relative abundance in the metagenome (Figure 5-4) but peptides from this subunit were detected in the metaproteome. All mcrA genes identified were related to the methanogen Methanobacterium, suggesting that the mcrA enzyme operates in the methane-producing direction in OL-KR46_570. One methanogenic MAG classified as Methanobacterium was recovered. The MAG contained a contig with the subunits mcrACGDB, which was expressed in the proteome indicating that methanogenesis is an active process in OL-KR46_570. Methanogenic archaea also use the WL pathway in the reductive direction for CO2 fixation, but they conserve energy by the conversion of H2 and CO2 to methane (autotrophic methanogenesis; equation 21) and/or acetate to methane (acetoclastic methanogenesis; equation 22). The Methanobacterium MAG had genes from both the autotrophic and acetoclastic pathway for methanogenesis, indicating it is able to utilise both CO2 and acetate. It also encoded a formate dehydrogenase indicating that is can also utilise formate (equation 23). 4H2 + CO2 → CH4 + 2H2O Equation (21) CH3COO- + H+ ⟶ CH4 + CO2 Equation (22) COOH- + H+ + H2O ⟶ CH4 + CO2 Equation (23)

5.2.1.4 Sulfide oxidising bacteria Two MAGs, classified as Hoeflea and Hydrogenophaga encoded genes for sulfide oxidation (Table 5-2). Hoeflea, a member of the Alphaproteobacteria, possessed the complete Sox system (SoxXA, SoxYZ, SoxB and SoxCD) for the oxidation of thiosulfate, sulfite, sulfur (Figure 3-9). Hoeflea also encoded the sulfide oxidising sqr and sulfite oxidising soeABC (Table 5-2). The soxC and sqr gene were detected in low abundance in its proteome, suggesting Hoeflea may be actively metabolising sulfide.

69

Table 5-2. OL-KR46_570 MAGs with genes involved in the oxidative sulfur cycle. Process HS- oxidation SO3-

oxidation Carbon fixation DNRA Denitri-

fication

Genes sat,

aprAB, dsrAB

Sox complex sqr soeABC PRK

nar, nap, nirK/S,

nrf

nar, nap, nirBD,

nor, nos

Hoeflea - soxXAYZB soxCD + + + narGHI

nirK

narGHI nirBD nosZ

Hydrogenophaga - soxXAYZB soxCD ++ - + narGHI

nirS

narGHI nirBD norBC nosZ

Hydrogenophaga also encoded the complete Sox system and two sqr genes (Table 5-2). Electron acceptors for sulfide oxidation are oxygen, nitrate or nitrite. Under anaerobic conditions nitrate or nitrite can be utilised (equations 8 and 9). Both Hoeflea and Hydrogenophaga encoded genes for denitrification, however nitrate and nitrite were not detected in OL-KR46_570 so this metabolism may be limited by lack of a sufficient electron acceptor, unless nitrate and nitrite are present at levels below the detection limit. Both Hoeflea and Hydrogenophaga encoded the rPP pathway for carbon fixation, indicating that they can grow chemolithoautotrophically, with sulfide, nitrate and inorganic carbon. Hoeflea also had genes for organic carbon oxidation via the glycolysis and ED pathway, which were partially present in its proteome. Genes for the utilisation/production of fatty acids and alcohols were also detected. Hoeflea encoded a complete TCA cycle and the key enzyme citrate synthase was detected in its proteome. Too few proteins were identified from Hydrogenophaga to identify active metabolic pathways, probably due to its low abundance in the community (0.3 ± 0.2% in 16S rRNA gene amplicon libraries). Hydrogenophaga encoded seven [NiFe] hydrogenases for hydrogen oxidation.

5.2.2 Fermentation

Two MAGs classified as Clostridiales and Erysipelothrix were identified as potential fermenters as neither harboured any respiratory pathways. Both MAGs harboured a complete glycolysis pathway, which was partially detected in their proteome. In addition, ABC transporters for the uptake of extracellular sugars and peptides were also detected. An almost complete pathway for lysine fermentation was found in the proteome of Clostridiales indicating it can utilise peptides as a source of carbon and nitrogen. Both MAGs harboured genes for acetate production from pyruvate and genes for the production of other small organic compounds.

70

5.3 Summary of metabolism in OL-KR46 groundwater

The concentration of sulfide was highest in OL-KR46_570 of the three groundwaters investigated. Unsurprisingly, SRB represented the most abundant metabolic group in the groundwater, corroborating hydrogeochemical and isotopic evidence that sulfate reduction is an active metabolic process. Proteomic data indicate that the oxidation of hydrogen and small organic compounds provides electrons for sulfate reduction. Hydrogen was not detected in the groundwater, but it can be difficult to measure in environments where hydrogen-utilising microorganisms maintain the concentration at levels below detection (Pedersen, 2014). The presence of respiratory hydrogenases in all SRB MAGs, however, indicates that hydrogen is actively utilised as an electron donor for sulfate reduction. Hydrogen may be sourced from geogenic gases or by fermentative hydrogen production (Figure 5-5). Hydrogenases were also identified in the proteomes of Acetobacterium and Methanobacterium suggesting hydrogen is an important electron donor for multiple taxa in this groundwater. This is consistent with previous models of a hydrogen driven subsurface biosphere (Kotelnikova and Pedersen, 1997; Pedersen, 1999). The isotopic signature of methane suggests that it is predominantly geogenic, but the presence of Methanobacterium peptides in the proteome suggest that biogenic methane also contributes. The absence of AOM OL-KR46_570 is supported by the microbiological data as ANME archaea, which catalyse AOM, were not detected and all mcrA genes were related to the methanogenic Methanobacterium. If ANME archaea are present in the sulfate-rich groundwater that is drawn-down into the deeper methane-rich groundwater during mixing then they should still be detected, but they were not found in 16S rRNA gene amplicon libraries. It is possible that the low abundance, slow growing ANME archaea were unable to establish in competition with faster growing SRB. In particular endospore-forming Firmicutes, which were abundant in OL-KR46_570 groundwater (e.g. Candidatus Desulforudis), are well-adapted to grow quickly in response to favourable environmental conditions and their frequent isolation from subsurface marine sediments has been attributed to their fast growth response (Parkes et al., 2014). In OL-KR46_570, the introduction of sulfate to deep groundwater containing electron donors for sulfate reduction (hydrogen and acetate) provides favourable environmental conditions for growth. Genes that encode enzymes for the oxidation of small organic compounds including formate, acetate, ethanol and lactate were also detected in SRB MAGs suggesting that organic carbon as well as hydrogen is driving sulfidogenesis. Organic compounds generated by primary production or produced by fermentative microorganisms can provide an electron donor for SRB (Figure 5-5). Acetate, acetone, ethanol and methanol were detected in the groundwater, however, there must also be more complex organic substrates present to provide an energy source for fermentative microorganisms Clostridiales and Erysipelothrix. Diverse heterotrophic populations have also been reported elsewhere in the Fennoscandian Shield (Parkes et al., 2014) despite complex organic carbon not typically being considered available in deep oligotrophic ecosystems with long residence times (>10,000 years) where organic compounds from the surface are not replenished. The degradation of recalcitrant natural organic matter (NOM) offers one possible source of complex organic carbon the deep subsurface.

71

The activity of bacteriophages detected in Olkiluoto groundwater may also provide a source of carbon (Purkamo et al., 2015, 2016; Wu et al., 2016) by inducing cell lysis and precipitating the associated release of small organic compounds and peptides into the carbon pool from cell necromass. Bacteriophages have been detected in groundwater from the Fennoscandian Shield (Pedersen, 2012) and one lytic to Pseudodesulfovibrio aespoeensis has been isolated from groundwater at Äspö (Kyle et al., 2008; Pedersen, 2013). Furthermore, bacteriophage-induced cell lysis has been postulated to control numbers of microorganisms in subsurface Fennoscandian Shield microbial communities (Eydal et al., 2009). Fermentative populations could degrade the released organic carbon and peptides, e.g., by the lysine fermentation pathway detected in the proteome of Clostridiales, and further produce small organic compounds that can be utilised by SRB. δ13CDIC values at Olkiluoto typically vary between -25‰ and -10‰, indicative of the degradation of organic carbon (Pedersen, 2012). Due to the low DIC content in the methane-rich groundwater at Olkiluoto, there are few comparative δ13CDIC values. One study of methane-rich, sulfate-depleted groundwater at Olkiluoto (863 m depth) reported a positive δ13CDIC value, +16.8 ‰ (Posiva, 2013). The positive δ13CDIC value was consistent with methanogens and acetogens being the dominant metabolic groups resulting in a greater proportion of carbon dioxide consumption and removal of the lighter 12C. The δ13CDIC values reported here indicate that carbon dioxide production (rather than consumption) is the dominant process contributing to the DIC pool. This supports that SRB, the most abundant metabolic group, could contribute to the DIC pool by metabolising small organic compounds and releasing CO2. SRB MAGs also expressed genes necessary for CO2 fixation, a CO2 consuming process. However, SRB that utilise the WL pathway can also reverse its action to generate metabolic energy by coupling the oxidation of acetate to the reduction of sulfate to sulfide (Posiva, 2013). The oxidation of other small organic compounds detected in the groundwater, such as ethanol, would also yield CO2. The MAGs Hoeflea and Hydrogenophaga had the metabolic potential for sulfide oxidation (Figure 5-7). Few proteins were detected from these MAGs highlighting a limitation of metaproteomic analysis, where proteins from lower abundance microorganisms may not be detected. Hoeflea and Hydrogenophaga had the complete pathway for denitrification suggesting that they couple sulfide oxidation to the reduction of nitrate. However, this process may be limited by lack of a suitable electron acceptor.

72

Figure 5-5. Metabolic pathways identified in OL-KR46_570. Electron acceptors (blue) and electron donors (green) are indicated. Carbon sources (yellow) are only marked for SRB. Processes identified in the proteome are indicated by a solid line, processes identified in the metagenome only are indicated by a dashed line.

73

6 INCUBATION EXPERIMENTS

Using a metagenomic and metaproteomic approach we have observed that the microbial processes responsible for sulfide production in OL-KR11_411, OL-KR13_405 and OL-KR46_570 are distinct. In OL-KR46_570, sulfate reduction is fuelled by H2 and acetate. Acetate is produced in situ by Acetobacterium and provides both an electron donor and carbon source for sulfate reduction (Figure 5-5). In OL-KR13_405, active AOM is evident in the proteome of an ANME archaea related to Candidatus Methanoperedens, however it is unclear from the genomic study if sulfate is the electron acceptor, or if AOM is coupled to an alternative electron acceptor (Figure 4-5). In OL-KR11, sulfate reduction was not apparent from hydrogeochemical or isotopic data, however, metaproteogenomics revealed a cryptic sulfur cycle ongoing in OL-KR11_411 whereby sulfide produced by SRB is oxidised back to sulfate with nitrate (Figure 3-15). The following experiments aim to provide direct chemical evidence for these processes. Groundwater was collected from OL-KR11_411, OL-KR13_405 and OL-KR46_570 and amended with stable isotope compounds containing 15N or 13C. Cells that incorporate the stable isotope compounds can then be identified by nanoSIMS, which can spatially resolve the measurement of isotopic ratios. This enables us to identify the uptake of specific elements that are labelled with stable isotopes (13C, 15N). For each experimental condition, a parallel incubation was prepared in the same way, but without the addition of isotopically labelled substrates i.e. no 15N or 13C is added. Unlabelled cells are also measured by nanoSIMS to determine the background isotopic signal of the cells.

6.1.1 Sulfide oxidation in OL-KR11_411

Sulfide oxidation coupled to denitrification was identified as a key metabolic process in OL-KR11_411 (Figure 3-15). To enrich sulfide-oxidising nitrate reducers, groundwater was amended with sulfide and stoichiometric amounts nitrate or nitrite in the laboratory under anoxic conditions (Table 6-1). 15N-ammonium was added as a nitrogen source, as ammonium is assimilated by actively growing cells. Cells that are actively growing in the incubations can then be identified by nanoSIMS as they incorporate the 15N label.

74

Table 6-1. OL-KR11_411 groundwater incubation experiments. Each incubation (A–F) was conducted in triplicate.

Incubation Electron acceptor

(mM)

Electron donor (mM)

Nitrogen source (mM)

Isotope label Figure

A NO3- (0.64) HS- (0.40) 15NH4

+ (1.00) 15N 6-1A

B NO3- (0.64) HS- (0.40) 14NH4

+ (1.00) None 6-1B

C Killed control

NO3- (0.64) HS- (0.40) 14NH4

+ (1.00) None 6-1C

D NO2- (0.80) HS- (0.40) 15NH4

+ (1.00) 15N 6-2A

E NO2- (0.80) HS- (0.40) 14NH4

+ (1.00) None 6-2B

F Killed control

NO2- (0.80) HS- (0.40) 14NH4

+ (1.00) None 6-2C

Within 10-days incubation, the amended sulfide was completely oxidised in incubations with nitrate (Figure 6-1A–B). Sulfide was not oxidised in killed controls (Figure 6-1C) confirming that the oxidation of sulfide was the result of microbial activity. To further enrich cell biomass, the incubations were spiked with sulfide and nitrate after 27- and 49-days incubation (Figure 6-1A–C). Sulfide was spiked a third time after 84 days incubation, but nitrate was not added as there was still ~1.5–2.0 mM remaining from previous additions. Nitrate and sulfate were also measured over the course of the incubation period (Figure 6-1A–C). Sulfate and nitrate concentrations were different in parallel incubations amended with nitrate (incubations A and B; Table 6-1). In incubation A, nitrate is not depleted during the first 49 days incubation, indicating that an alternative electron acceptor must have been used. The stoichiometry (sulfide consumed to sulfate produced) is consistent with sulfide oxidation coupled to oxygen reduction (equation 24). This suggests that oxygen was present in this set of incubations. OL-KR11_405 groundwater is anoxic, thus oxygen must have been introduced during preparation of the incubation experiments. HS- + 2O2 ⟶ SO4

2- + H+ Equation (24) After 49 days incubation, nitrate begins to be reduced (~0.46 mM; Figure 6-1A). A concomitant production of nitrite is observed indicating that some of the nitrate was reduced to nitrite (e.g. equation 25). However, the stoichiometry of equation 25 does not account for the total amount of sulfide oxidised in the incubation experiments. Sulfate continued to be produced as sulfide was depleted (0.54 mM sulfide oxidised; 0.45 mM sulfate produced) suggesting almost complete oxidation of sulfide to sulfate.

75

2HS- + 8NO3- ⟶ 2SO4

2- + 8NO2- + 2H+ Equation (25)

In the parallel incubation with no isotopic label (incubation D; Figure 6-1B) the concentration of nitrate decreased from the beginning of the incubation. After 49 days of incubation, a total of 1.01 ± 0.11 mM sulfide had been oxidised and 0.40 ± 0.07 mM nitrate reduced (Figure 6-1B). The stoichiometry indicates that sulfide is oxidised to sulfur when coupled to denitrification (equation 26). Sulfate did not accumulate in the enrichments, consistent with the oxidation of sulfide to an intermediate oxidation state species. Nitrite remained below the limit of quantification (0.5 mg/L = 0.01 mM). After 49 days incubation sulfide was no longer depleted, suggesting cell growth ceased (Figure 6-1B). 5HS- + 2NO3

- + 7H+ ⟶ 5S0 + N2 + 6H2O Equation (26)

76

Figure 6-1. OL-KR11_411 groundwater incubations. Groundwater was amended with sulfide, nitrate and ammonium. Sulfide, nitrate and sulfate were measured. Incubations were spiked with nitrate and sulfide after sulfide was consumed.

77

In incubations amended with nitrite (incubations D and E; Table 6-1), sulfide was oxidised coupled to the reduction of nitrite (Figure 6-2A and B). Sulfate did not accumulate suggesting that sulfide is oxidised to an intermediate oxidation state species. However, a greater concentration of nitrite was consumed than the requirement to oxidise sulfide to sulfur (equation 27), and it was neither consistent with the oxidation of sulfide to sulfite nor to thiosulfate. 3HS- + 2NO2 + 5H+ ⟶ 3S0 + N2 + 4H2O Equation (27) The concentration of nitrite continues to decrease after sulfide is depleted, suggesting a secondary process is occurring. If sulfide is oxidised to elemental sulfur, this may fuel further nitrite consumption via denitrification.

78

Figure 6-2. OL-KR11_411 groundwater incubations. Groundwater was amended with sulfide, nitrite and ammonium. Sulfide, nitrite and sulfate were measured. Incubations were spiked with nitrate and sulfide after sulfide was consumed.

79

Groundwater amended with sulfide and nitrite (incubation D; Table 6-1) was selected for CARD-FISH nanoSIMS analysis. The incubation was sampled after 16 days incubation. Three sets of CARD-FISH probes targeting (1) bacteria (2) archaea and (3) Epsilonproteobacteria were applied to the samples, however they all failed to hybridise with cells. This suggests that the CARD-FISH probes were unable to enter the cell. This can be due to the fixation strategy used as paraformaldehyde cross links proteins and can make the cell wall impenetrable. All cells were counter stained with DAPI which can pass through an intact cell membrane. All DAPI stained cells from incubation D were enriched in 15N (Figure 6-3). This shows that all cells were actively assimilating 15N-ammonium into biomass. As phylogenetic identification of cells via CARD-FISH was unsuccessful, DNA was extracted and sequenced (Figure 6-4).

Figure 6-3. DAPI (4′6-diamidino-2-phenylindole) stained cells from groundwater amended with sulfide, nitrite and 15N-ammonium (incubation D; Table 6-1) (left) and corresponding isotope ratio image inferred from secondary ion (15N12C/14N12C) nanoSIMS measurements (right). Scale bar = 2µm.

80

Figure 6-4. 16S rRNA gene amplicon libraries (V4 region) of OL-KR11_411 groundwater incubations amended with sulfide and nitrate or nitrite. Aug-18 shows the relative abundance of enriched genera in groundwater used to prepare the incubations sampled in August 2018. A, B, D and E refer to the incubation conditions in Table 6-1. DNA was extracted from all of the live groundwater incubations (A, B, D and E; Table 6-1). 16S rRNA gene amplicon sequencing revealed that Sulfurimonas dominated all anoxic incubations with nitrate and nitrite (incubations B, D and E; Figure 6-4), accounting for ~80% relative abundance. A Sulfurimonas MAG was recovered from OL-KR11_405 groundwater in 2016 (Table 3-2). The Sulfurimonas MAG recovered from OL-KR11_405 encoded sulfide:quinone oxidoreductases (sqr) which oxidises sulfide to sulfur. The MAG was lacking the Sox gene cluster (Table 3-2), suggesting that Sulfurimonas is not able to completely oxidise sulfide to sulfate. Experimental data show that in enrichments dominated by Sulfurimonas, sulfide is oxidized primarily to sulfur under nitrate- and nitrite-reducing conditions (Figure 6-1B, 6-2A–B). This is consistent with the genetic capacity of the Sulfurimonas MAG recovered from OL-KR11_405 in 2016. In incubation A (Table 6-1), the stoichiometry indicated that there was oxygen contamination in the laboratory incubations (Figure 6-1A). This is supported by the 16S rRNA gene amplicon sequencing (Figure 6-3). The microbial community enriched was different to parallel incubations (cf. A and B, Figure 6-3). Pseudomonadaceae, Hydrogenophaga, Rhizobiacae and Thiobacillus are the most abundant microorganisms in incubation A. These microorganisms are sulfide oxidisers and facultative anaerobes, with the metabolic capacity to use oxygen.

81

The concentration of sulfide and nitrate/nitrite added to these groundwater incubations exceeds the concentration that would be encountered in situ, nevertheless it demonstrates the metabolic capacity for active coupling of sulfur and nitrogen cycles in Olkiluoto groundwater.

6.1.2 Anaerobic oxidation of methane in OL-KR13_405

Methane oxidation was identified as a metabolic process in OL-KR13_405 (Figure 4-5), but the electron acceptor used for this metabolism is unclear. Possible electron acceptors include sulfate, iron(III) and nitrate. To enrich methane oxidising archaea from OL-KR13_405, groundwater was amended with methane and an electron acceptor. The headspace of incubations with labelled methane were amended with 0.6 mL 100% 13CH4 (Table 6-2; incubations 1, 4 and 6). The headspace of unlabelled incubations was amended with 60 mL of 1% 12CH4 (Table 6-2; incubations 2, 3, 5, 6, 8 and 9). Incubations with sulfate (1, 2 and 3) were not amended with additional sulfate, as sulfate is naturally present in OL-KR13_405 groundwater. Incubations with nitrate (4, 5 and 6) were amended with 0.8 mM nitrate. Incubations with Fe(III) were amended with 1.0 mM iron citrate. To identify growing cells via nanoSIMS, all incubations were also amended with 1.0 mM 15N-ammonium. Table 6-2. OL-KR13_405 groundwater incubation experiments. Each incubation (1–9) was conducted in triplicate.

Incubation Electron acceptor (mM)

Electron donor

Nitrogen source (mM)

Isotope label Figures

1 SO42- (0.6) 13CH4

15NH4+ (1.0)

13C/15N 6-5A, 6-6

2 SO42- (0.6) 12CH4

14NH4+ (1.0) None 6-5A,

6-6

3 Killed control SO4

2- (0.6) 12CH4 14NH4+ (1.0) None 6-5A,

6-6

4 NO3- (0.8) 13CH4 15NH4

+ (1.0) 13C/15N 6-5B,

6-7A

5 NO3- (0.8) 12CH4 14NH4

+ (1.0) None 6-5B, 6-7B

6 Killed control NO3

- (0.8) 12CH4 14NH4+ (1.0) None 6-5B,

6-7C

7 Fe(III) (1.0) 13CH4 15NH4+ (1.0)

13C/15N 6-5C, 6-8A

8 Fe(III) (1.0) 12CH4 14NH4+ (1.0) None 6-5C,

6-8B

9 Killed control Fe(III) (1.0) 12CH4 14NH4

+ (1.0) None 6-5C, 6-8C

The headspace methane concentration was monitored in OL-KR13_405 groundwater incubations over a ~5-month period. No methane oxidation was evident from the measurement of methane in incubations with sulfate (Figure 6-5A), nitrate (Figure 6-5B) or iron(III) (Figure 6-5C).

82

Figure 6-5. Methane concentration OL-KR13_405 groundwater incubations amended methane and sulfate (A), methane and nitrate (B) or methane and iron(III) (C).

83

Sulfate was also measured in groundwater incubations amended with methane (Figure 6-6). No sulfate reduction was detected; the concentration of sulfate in live incubations was the same as in the killed control (Figure 6-6). Thus, in these incubations, the presence of sulfate and methane did not stimulate the activity of anaerobic methane oxidising archaea.

Figure 6-6. Sulfate in OL-KR13_405 groundwater incubations amended with methane. In incubations with nitrate, nitrate was converted to nitrite (Figure 6-7A: 1.01 mM nitrate reduced, 1.17 mM nitrite produced; Figure 6-7B: 0.61 mM nitrate reduced 0.59 mM nitrite produced). Nitrate reduction to nitrite can be coupled to methane oxidation (equation 28). The observed reduction of nitrate (Figure 6-7A and B) would require 0.15–0.25 mM CH4 to be oxidised according to equation 27. It is possible that this concentration of methane was oxidised, however it is in the within the error range of the methane measurements (Figure 6-5B). CH4 + 4NO3

- → CO2 + 4NO2- + 2H2O Equation (28)

Sulfide is present in OL-KR13_405 groundwater (~0.45 mM) and could also act as an electron donor for nitrate reduction in these groundwater incubations. No sulfate was produced, so if sulfide was oxidised, it was not oxidised to sulfate. Sulfide could be oxidised to sulfur. Using the in-situ measurement of sulfide, there is insufficient sulfide present in the groundwater to reduce 1.01 mM (Figure 6-7A) and 0.61 mM (Figure 6-7B) mM of nitrate to nitrite (equation 29). 2NO3

- + 2HS- + 2H+ ⟶ 2NO2- + 2S0 + 2H2O Equation (29)

84

Figure 6-7. OL-KR13_405 groundwater incubations amended with methane and nitrate. Nitrate and nitrate were not measured at the start of incubation (0 days); grey symbols at 0 days incubation therefore indicate that the value is assumed based on the measured concentration of nitrate and nitrite in killed controls at 47- and 103-days incubation, assuming no biological activity in these samples.

85

Iron reduction was observed in anaerobic incubations amended with Fe(III) (Figure 6-8A and B). Upon preparation of the incubations, a black precipitate was immediately formed. 1 mM Fe(III) was added to the incubations, ~0.5 mM of which was immediately converted to Fe(II) (Figure 6-8A–C). The precipitation occurred in both live incubations (Figure 6-8A and B) and the killed control (Figure 6-8C) indicating that the process was abiotic. Fe(III) can react abiotically with sulfide to form Fe(II) and sulfur. The Fe(II) can then react with sulfide to form FeS (equation 30) (Johansson et al., 2019).

2HS− + 2Fe(III) → Fe2+ + S0 + FeS + 2H+ Equation (30)

Following the precipitation of FeS, 0.47 and 0.34 mM Fe(III) was reduced in live incubations (Figure 6-7A and B). According to equation 31 this would equate to the oxidation of 0.04–0.06 mM methane. This small change in methane concentration was within the error of the methane measurements, so could not be detected (Figure 6-5C). CH4 + 8Fe3+ + 2H2O → CO2 + 8Fe2+ + 8H+ Equation (31)

86

Figure 6-8. Iron reduction in OL-KR13 groundwater incubations amended with iron(III) and methane. Iron(III) is calculated by subtracting the measured Fe(II) concentration from total iron. The total iron analysis included the precipitated FeS, accounting for the small decrease in total iron in all incubations.

87

Incubations with iron were selected for further analysis, using CARD-FISH probes targeting Archaea to determine if methane oxidising archaea were enriched. Probes targeting Archaea failed to hybridise with cells in the sample. DNA was thus extracted and sequenced to determine if Archaea were present in the groundwater enrichment. The DNA yield from the water phase was very low. This could be due to cells attaching to the FeS precipitate in the enrichment. Therefore, DNA was also extracted from filters which captured the FeS precipitate. The DNA yield was again very low from the filter samples. Because the DNA yield was very low from OL-KR13_405 incubation, a greater number of PCR cycles (>30) were required to yield enough DNA for 16S rRNA gene amplicon sequencing. Increasing the number of PCR cycles on low biomass samples can lead to the amplification of contaminating DNA from the extraction procedure. To determine if sequences were from contamination, procedural extraction blanks were also sequenced and any sequences present in the blank were removed from the sequencing library (Figure 6-9). Although contaminating sequences were removed the results should still be treated with caution. Of the sequences recovered from OL-KR13_405 incubations (Figure 6-9), sulfur disproportionating species Desulfocapsa and Desulfurivibrio were the most abundant in DNA extracted from the filters, including the FeS precipitate. Desulfocapsa and Desulfurivibrio could utilise S0 produced by the abiotic reaction of sulfide and Fe(III) (equation 28). Marinilabiliaceae was most abundant in amplicon libraries from the liquid phase. Marinilabiliaceae are reported to be anaerobic fermenters (Zhao et al., 2012). No methane oxidizing archaea were identified by 16S rRNA gene amplicon sequencing. Futhermore, nanoSIMS samples showed no enrichment of 13CH4 in microbial biomass. The cell biomass on filters used for nanoSIMS was very low, and no 15N incorporation was detected either, suggesting the detected cells were not growing. Thus, it is possible that ANME were present, but it was not possible to detect them due to their low abundance. It is not possible to demonstrate that iron reduction is coupled to methane oxidation with the present data.

88

Figure 6-9. 16S rRNA gene amplicon libraries from OL-KR13_405 groundwater incubations amended with Fe(III) and CH4 (Table 6-2). DNA was extracted from a filter with FeS precipitate and from the liquid enrichment. Libraries from the filter are marked with a (P) and libraries from the liquid are marked with an (L). KR13Aug18 shows the in situ relative abundance of microorganisms from OL-KR13_405 groundwater collected in August 2019. The relative abundance of microorganisms does not equal 100% as contaminating sequences represented a large majority and were removed from amplicon libraries (see text for details).

89

6.1.3 Acetate oxidation in OL-KR46_570

Sulfate reduction is ongoing in OL-KR46_570 (Figure 5-5). Sulfate-reducing bacteria detected in OL-KR46_570 groundwater contained genes for hydrogen oxidation, acetate oxidation and carbon fixation (Table 5-1). To determine whether sulfate-reducing bacteria from OL-KR46_570 are autotrophic (assimilate inorganic carbon) or heterotrophic (assimilate organic carbon) incubations were prepared with 13C-acetate or and 13C-bicarbonate (Table 6-3). 15N-ammonium was also added to the groundwater incubations to indicate microbial activity. Table 6-3. OL-KR46 groundwater incubations. Each incubation (1–9) was conducted in triplicate.

Incubation Electron acceptor

Electron donor

Carbon source

Nitrogen source

Isotope label Figure

A SO42- H2 13CH3COO- 15NH4

+ 13C/15N 6-10A

B SO42- H2 12CH3COO- 14NH4

+ None 6-10B

C Killed

Control SO4

2- H2 12CH3COO- 14NH4+ None

6-10C

D SO42- H2 13HCO3

- 15NH4+ 13C/15N 6-11A

E SO42- H2 12HCO3

- 14NH4+ None 6-11B

F Killed

Control SO4

2- H2 12HCO3- 14NH4

+ None 6-11C

G SO42- CH3COO- 13CH3COO- 15NH4

+ 13C/15N 6-12A

H SO42- CH3COO- 12CH3COO- 14NH4

+ None 6-12B

I Killed

Control SO4

2- CH3COO- 12CH3COO- 14NH4+ None

6-12C

To enrich sulfate-reducing bacteria from OL-KR46_570, groundwater was amended with either (1) 1 mM bicarbonate and a 20% hydrogen headspace (Table 6-3; incubations A, B and C), (2) 1 mM acetate and a 20% hydrogen headspace (Table 6-3; incubations D, E and F) or (3) 1 mM acetate only (Table 6-3; incubations G, H and I). Acetate was present in all incubations (regardless of whether it was added) as it is naturally present in OL-KR46_570 groundwater. Incubations were initially not amended with additional sulfate as sulfate is naturally present in OL-KR46_570 groundwater. No sulfate reduction was observed during the first 50 days incubation of OL-KR46_570 groundwater under any condition (Figures 6-10, 6-11 and 6-12). Hydrogen was depleted during this time, suggesting other microorganisms present are able to oxidise hydrogen. To try and stimulate sulfate reduction, the incubations were spiked with 0.7 mM sulfate

90

after 49 days. This concentration was chosen to replicate the sulfate concentration in OL-KR46_570 during 2016, when sulfate reduction was observed (section 5). After the addition of sulfate, sulfate decreased in incubations amended with H2, whether the carbon source was acetate (Figure 6-10A and B) or bicarbonate (Figure 6-11 B). Sulfate reduction was limited in incubations amended with 13C (Figure 6-10A and 6-11A) relative to 12C (Figure 6-10B and 6-11B), suggesting a preference for growth with 12C. When hydrogen was omitted and only acetate was added, no sulfate reduction was observed (Figure 6-12C). No sulfate reduction was detected in control incubations (Figure 6-10C, 6-11C, 6-12C).

91

Figure 6-10. OL-KR46_570 groundwater incubations amended with 13C-acetate and hydrogen (A) or 12C-acetate and hydrogen (B), and 12C-acetate and hydrogen (killed; C). The red arrows indicate the sulfate addition after 49 days incubation.

92

Figure 6-11: OL-KR46_570 groundwater incubations amended with 13C-bicarbonate and hydrogen (A) or 12C-bicarbonate and hydrogen (B), and 12C-bicarbonate and hydrogen (killed; C). The red arrows indicate the sulfate addition after 49 days incubation.

93

Figure 6-12: OL-KR46 groundwater incubations amended with 13C-acetate (A), 12C-acetate (B), and 12C-acetate (killed; C). The red arrows indicate the sulfate addition after 49 days incubation.

94

Incubations amended with acetate and hydrogen (Figure 6-10A) or bicarbonate and hydrogen (6-10B) were selected for nanoSIMS analysis. In both instances, not all cells detected by DAPI staining were detected by nanoSIMS (Figure 6-13 & 6-14). This shows that not all cells present assimilated ammonium and carbon and may not be actively growing. However, all cells that incorporated 15N-ammonium also incorporated 13C-carbon (Figure 6-13 & 6-14). This shows that microorganisms from OL-KR46_570 groundwater enriched under the experimental conditions with hydrogen (Table 6-3) are able to utilise both bicarbonate and acetate as a carbon source to produce biomass. CARD-FISH probes targeting Bacteria, Archaea and Deltaproteobacteria were tested but again were unsuccessful. Cells from all incubations (OL-KR11, OL-KR13 and OL-KR46) were fixed in the same way. As CARD-FISH was unsuccessful in all instances, this indicates that the probes were unable to enter the cell. This is likely due to the fixation strategy used as paraformaldehyde crosslinks proteins and can make the cell wall impenetrable. Without CARD-FISH to identify taxonomy, it was not possible to conclude which microorganism(s) were able to incorporate acetate (Figure 6-13) and bicarbonate (Figure 6-13).

Figure 6-13. DAPI stained cells from groundwater amended with sulfate, 13C-acetate and 15N-ammonium (left) with corresponding isotope ratio images inferred from secondary ion nanoSIMS measurements; (13C12C/13C12C) (middle) (15N12C/14N12C) (right). Images are 25 x 25 µm, scale bar = 2µm.

Figure 6-14. DAPI stained cells from groundwater amended with sulfate, 13C-bicarbonate and 15N-ammonium (left) with corresponding isotope ratio images inferred from secondary ion nanoSIMS measurements; (13C12C/13C12C) (middle) (15N12C/14N12C) (right). Images are 40 x 40 µm, scale bar = 2µm. 16S rRNA gene amplicon libraries were generated from DNA extracted from OL-KR46_570 groundwater incubations to determine which microorganisms are present. The

95

greatest proportion of sulfate-reducing bacteria were detected in enrichments with acetate and hydrogen (Figure 6-15). Dethiosulfatibacter (thiosulfate-reducing) and Desulfomicrobium (sulfate-reducing) represented ~60% of the microbial community. In these enrichments sulfate reduction appears to be coupled to hydrogen oxidation (Figure 6-10A and 6-10B). Acetate was not measurably oxidised (Figure 6-10A and 6-10B) but was incorporated into cell biomass (Figure 6-14). Desulfomicrobium were also detected (~30%) in incubations with bicarbonate and hydrogen. Desulfomicrobium had a greater relative abundance in incubations with 12C-bicarbonate than 13C-bicarbonate (Figure 6-15) which may explain why comparatively less sulfate reduction was observed in 13C-bicarbonate experiments (Figure 6-13A). The sulfate reducer Desulfovibrio was only detected in incubations with bicarbonate (Figure 6-15) suggesting that Desulfovibrio requires bicarbonate for growth. Acetobacterium were also enriched in incubations with hydrogen and bicarbonate, consistent with the metabolism determined for this genus by metaproteogenomics (section 5.2.1.2). In incubations amended with acetate only, no sulfate reduction was observed (Figure 6-12). The microbial community composition was similar to that observed for in situ groundwater (Figure 6-15). A methanogen identified as Methanolobus was enriched in incubations amended with 13C acetate, but this was not observed for parallel incubations with 12C acetate.

96

Figure 6-15: 16S rRNA gene amplicon sequencing (V4 region) from OL-KR46 groundwater amended with acetate, acetate and hydrogen, or bicarbonate and hydrogen (Table 6-3). The in situ relative abundance of microorganisms detected in OL-KR46 groundwater in August 2018 is also shown.

97

7 SYNTHESIS

Key metabolic processes uncovered at Olkiluoto are highlighted in Figure 7-1. The key electron acceptor in deep sulfate-free saline groundwater is carbon dioxide. With geogenic hydrogen as an electron donor, carbon dioxide can be reduced to methane by methanogens such as Methanobacterium, or to acetate by acetogens such as Acetobacterium. When sulfate-rich brackish groundwater mixes with sulfate-free saline groundwater, this introduces an abundant electron acceptor. Geogenic hydrogen and acetate produced by acetogenesis provide both electron donor and carbon source, stimulating sulfate reduction by SRB including Desulfomicrobium and Candidatus Desulforudis. The concentration of DIC in deep saline groundwaters, like OL-KR46_570, is generally too low to measure. Calcite dissolution may be a source of DIC, activated by carbonate consumption and dissolution to maintain equilibrium in the groundwater. DIC can be introduced to the groundwater during mixing, which was evident in OL-KR46_570 after the packer leakage in July 2016 (Figure 5-1 and 5-2) which also resulted in increased relative abundance of Acetobacterium in following month (Figure 5-3), which uses CO2 to produce acetate. Acetobacterium was also enriched in groundwater incubations amended with bicarbonate and hydrogen (Figure 6-15). Sulfate-reducing bacteria that are capable of completely oxidising acetate will produce carbon dioxide, recycling the carbon to provide an electron acceptor for acetogens and methanogens. Desulfomicrobium had the greatest relative abundance in groundwater incubations amended with acetate and hydrogen (Figure 6-15). As the concentration of DIC in saline groundwaters is low, it is often not possible to measure the isotopic ratio of DIC. The processes occurring in OL-KR46_570 makes the 13C evolution complex. Two C-isotope samples were taken from OL-KR46_570. When acetogens utilise CO2, 13C would increase in the residual CO2, thus isotopically light 13C is significantly needed to produce the measured values -14 to -17‰. The signature may be representative of the DIC which is being introduced, or may indicate that DIC primarily comes from the anaerobic mineralisation of organic compounds and production of CO2. Of the three drillholes studied, the concentration of sulfide was greatest in the OL-KR46_570. This may be explained by the greater relative abundance of sulfate-reducing bacteria in this groundwater, fuelled by hydrogen. It also suggests that sulfide-oxidising processes are limited at this depth. The metabolic potential for sulfide oxidation was detected in OL-KR46_570 MAGs, but the metaproteome did not confirm that the process was active. This may be due to limitation of a suitable electron acceptor (e.g. nitrate). Additionally, Fe(III) and Fe(II) appears to be limited in deep saline groundwaters, which could provide a sink for sulfide either via direct or indirect precipitation. Some framboidal pyrite was observed, which requires elemental sulfur, indicating intermediate oxidation state compounds may also be produced during S metabolism. In shallower brackish-SO4

2- groundwaters, which have also undergone some mixing, sulfate is abundant, originating from ancient Littorina Sea input. Sulfate reduction is clearly apparent from hydrogeochemical measurements when sulfide accumulates in the groundwater and the isotopic signature is significantly enriched in the heavy isotope (34S)

98

compared to seawater value, as in OL-KR13_405. Abundant sulfate reducers identified in brackish-SO4

2- are able to disproportionate sulfur compounds (thiosulfate, sulfite and sulfide) in addition to reducing sulfate, with hydrogen and organic matter. Sulfate reduction is not always clear from the hydrogeochemical measurements alone. In OL-KR11_411, no sulfide was detected and the isotopic signature of sulfate was not enriched relative to Olkiluoto groundwater with the same sulfate source. In this case sulfide is removed from the groundwater by the activity of sulfide oxidising bacteria. Sulfide does not accumulate in the groundwater when sulfide oxidising bacteria are active. Proteomic data indicates that the electron acceptor for sulfide oxidation is nitrate. However, nitrate is not detected in the water and could limit this process. Nitrate reducers were facultative anaerobes and could utilise trace concentrations of oxygen in the groundwater. However, the groundwater was reducing (~200 mV) so oxygen would be extremely limited/absent. The source of nitrate in the groundwater is unknown. Active methane oxidation by archaea related to ANME-2d Candidatus Methanoperedens in brackish-SO4

2- was evident in the proteome. However, the isotopic signature of DIC, suggests that only a minor proportion of CO2 could be contributed from the oxidation of methane. The electron acceptor for methane oxidation is unclear. Genomic data indicates that the Candidatus Methanoperedens detected in OL-KR13_405 does not have the genetic capacity for nitrate-dependant methane oxidation but groundwater incubations were unable to resolve the coupling to iron or sulfate.

99

Figure 7-1. Microbial metabolic processes occurring in Olkiluoto groundwater.

101

8 REFERENCES

Adhikari, R. R., Glombitza, C., Nickel, J. C., Anderson, C. H., Dunlea, A. G., Spivack, A. J., et al. (2016). Hydrogen utilization potential in subsurface sediments. Front. Microbiol. 7, 8. doi:10.3389/fmicb.2016.00008.

Alneberg, J., Bjarnason, B. S., De Bruijn, I., Schirmer, M., Quick, J., Ijaz, U. Z., et al. (2014). Binning metagenomic contigs by coverage and composition. Nat. Methods 11, 1144–1146. doi:10.1038/nmeth.3103.

Altschul, S. F., Gish, W., Miller, W., Myers, E. W., and Lipman, D. J. (1990). Basic local alignment search tool. J Mol Biol 215, 403–410. doi:10.1016/s0022-2836(05)80360-2.

Bagnoud, A., Chourey, K., Hettich, R. L., De Bruijn, I., Andersson, A. F., Leupin, O. X., et al. (2016). Reconstructing a hydrogen-driven microbial metabolic network in Opalinus Clay rock. in Nature Communications, 12770. doi:10.1038/ncomms12770.

Bengtsson-Palme, J., Hartmann, M., Eriksson, K. M., Pal, C., Thorell, K., Larsson, D. G. J., et al. (2015). metaxa2: Improved identification and taxonomic classification of small and large subunit rRNA in metagenomic data. Mol. Ecol. Resour. 15, 1403–1414. doi:10.1111/1755-0998.12399.

Berger, S., Frank, J., Martins, P. D., Jetten, M. S. M., and Welte, C. U. (2017). High-quality draft genome sequence of “Candidatus Methanoperedens sp.” strain BLZ2, a nitratereducing anaerobic methane-oxidizing archaeon enriched in an anoxic bioreactor. Genome Announc. 5, e01159-17. doi:10.1128/genomeA.01159-17.

Boetius, A., Ravenschlag, K., Schubert, C. J., Rickert, D., Widdel, F., Gleseke, A., et al. (2000). A marine microbial consortium apparently mediating anaerobic oxidation methane. Nature 407, 623–626. doi:10.1038/35036572.

Bomberg, M., Nyyssönen, M., Pitkänen, P., Lehtinen, A., and Itävaara, M. (2015). Active Microbial Communities Inhabit Sulphate-Methane Interphase in Deep Bedrock Fracture Fluids in Olkiluoto, Finland. Biomed Res. Int. 2015, 1–17. doi:10.1155/2015/979530.

Bray, N. L., Pimentel, H., Melsted, P., and Pachter, L. (2016). Near-optimal probabilistic RNA-seq quantification. Nat. Biotechnol. 34, 525–527. doi:10.1038/nbt.3519.

Cai, C., Leu, A. O., Xie, G.-J. J., Guo, J., Feng, Y., Zhao, J.-X. X., et al. (2018). A methanotrophic archaeon couples anaerobic oxidation of methane to Fe(III) reduction. ISME J. 12, 1929–1939. doi:10.1038/s41396-018-0109-x.

Canfield, D. E., Thamdrup, B., and Fleischer, S. (1998). Isotope fractionation and sulfur metabolism by pure and enrichment cultures of elemental sulfur-disproportionating bacteria. Limnol. Oceanogr. 43, 253–264. doi:10.4319/lo.1998.43.2.0253.

Caporaso, J. G., Kuczynski, J., Stombaugh, J., Bittinger, K., Bushman, F. D., Costello, E. K., et al. (2010). QIIME allows analysis of high-throughput community sequencing data. Nat. Methods 7, 335–336. doi:10.1038/nmeth.f.303.

Caporaso, J. G., Lauber, C. L., Walters, W. A., Berg-Lyons, D., Lozupone, C. A.,

102

Turnbaugh, P. J., et al. (2011). Global patterns of 16S rRNA diversity at a depth of millions of sequences per sample. Proc. Natl. Acad. Sci. 108, 4516–4522. doi:10.1073/pnas.1000080107.

Caspi, R., Billington, R., Fulcher, C. A., Keseler, I. M., Kothari, A., Krummenacker, M., et al. (2018). The MetaCyc database of metabolic pathways and enzymes. Nucleic Acids Res. 46, D633–D639. doi:10.1093/nar/gkx935.

Chourey, K., Jansson, J., VerBerkmoes, N., Shah, M., Chavarria, K. L., Tom, L. M., et al. (2010). Direct Cellular Lysis/Protein Extraction Protocol for Soil Metaproteomics. J. Proteome Res. 9, 6615–6622. doi:10.1021/pr100787q.

Chourey, K., Nissen, S., Vishnivetskaya, T., Shah, M., Pfiffner, S., Hettich, R. L., et al. (2013). Environmental proteomics reveals early microbial community responses to biostimulation at a uranium- and nitrate-contaminated site. Proteomics 13, n/a-n/a. doi:10.1002/pmic.201300155.

Cline, J. D. (1969). Spectrophotometric determination of hydrogen sulfide in natural waters. Limnol. Oceanogr. 14, 454–458. doi:10.4319/lo.1969.14.3.0454.

Cole, J. R., Wang, Q., Cardenas, E., Fish, J., Chai, B., Farris, R. J., et al. (2009). The Ribosomal Database Project: improved alignments and new tools for rRNA analysis. Nucleic Acids Res. 37, D141–D145. doi:10.1093/nar/gkn879.

Dahl, C., Friedrich, C., and Kletzin, A. (2008a). “Sulfur Oxidation in Prokaryotes,” in Encyclopedia of Life Sciences (Chichester, UK: John Wiley & Sons, Ltd). doi:10.1002/9780470015902.a0021155.

Dahl, C., Schulte, A., Stockdreher, Y., Hong, C., Grimm, F., Sander, J., et al. (2008b). Structural and Molecular Genetic Insight into a Widespread Sulfur Oxidation Pathway. J. Mol. Biol. 384, 1287–1300. doi:10.1016/J.JMB.2008.10.016.

De Zwart, J., Sluis, J., Kuenen, J. G., Quentmeier, A., and Fischer, J. (1997). Competition for Dimethyl Sulfide and Hydrogen Sulfide by Methylophaga sulfidovorans and Thiobacillus thioparus T5 in Continuous Cultures. Appl. Environ. Microbiol. 63, 3318–22. doi:10.1128/aem.67.7.2873-2882.2001.

Edgar, R. C. (2010). Search and clustering orders of magnitude faster than BLAST. Bioinformatics 26, 2460–2461. doi:10.1093/bioinformatics/btq461.

Edgar, R. C. (2016). UNOISE2: improved error-correction for Illumina 16S and ITS amplicon sequencing. Cold Spring Harbor Laboratory doi:10.1101/081257.

Ettwig, K. F., Zhu, B., Speth, D., Keltjens, J. T., Jetten, M. S. M., and Kartal, B. (2016). Archaea catalyze iron-dependent anaerobic oxidation of methane. Proc. Natl. Acad. Sci. 113, 12792–12796. doi:10.1073/pnas.1609534113.

Eydal, H. S. C., Jägevall, S., Hermansson, M., and Pedersen, K. (2009). Bacteriophage lytic to Desulfovibrio aespoeensis isolated from deep groundwater. ISME J. 3, 1139–1147. doi:10.1038/ismej.2009.66.

Finneran, K. T., Johnsen, C. V, and Lovley, D. R. (2003). Rhodoferax ferrireducens sp. nov., a psychrotolerant, facultatively anaerobic bacterium that oxidizes acetate with the reduction of Fe(III). Int. J. Syst. Evol. Microbiol. 53, 669–673. doi:10.1099/ijs.0.02298-0.

103

Finster, K. (2008). Microbiological disproportionation of inorganic sulfur compounds. J. Sulfur Chem. 29, 281–292. doi:10.1080/17415990802105770.

Finster, K. W., Kjeldsen, K. U., Kube, M., Reinhardt, R., Mussmann, M., Amann, R., et al. (2013). Complete genome sequence of Desulfocapsa sulfexigens, a marine deltaproteobacterium specialized in disproportionating inorganic sulfur compounds. Stand. Genomic Sci. 8, 58–68. doi:10.4056/sigs.3777412.

Frederiksen, T.-M., and Finster, K. (2003). Sulfite-oxido-reductase is involved in the oxidation of sulfite in Desulfocapsa sulfoexigens during disproportionation of thiosulfate and elemental sulfur. Biodegradation 14, 189–98. Available at: http://www.ncbi.nlm.nih.gov/pubmed/12889609 [Accessed December 13, 2017].

Friedrich, C. G., Bardischewsky, F., Rother, D., Quentmeier, A., and Fischer, J. (2005). Prokaryotic sulfur oxidation. Curr. Opin. Microbiol. 8, 253–259. Available at: http://www.scopus.com/inward/record.url?eid=2-s2.0-20444496515&partnerID=40&md5=8f6be3761d7f3b16cf177e9b201949cb.

Friedrich, C. G., Rother, D., Bardischewsky, F., Quentmeier, A., and Fischer, J. (2001). Oxidation of reduced inorganic sulfur compounds by bacteria: emergence of a common mechanism? Appl. Environ. Microbiol. 67, 2873–82. doi:10.1128/AEM.67.7.2873-2882.2001.

Griesbeck, C., Schütz, M., Schödl, T., Bathe, S., Nausch, L., Mederer, N., et al. (2002). Mechanism of sulfide-quinone reductase investigated using site-directed mutagenesis and sulfur analysis. Biochemistry 41, 11552–11565. doi:10.1021/bi026032b.

Habicht, K. S., Canfield, D. E., and Rethmeier, J. (1998). Sulfur isotope fractionation during bacterial reduction and disproportionation of thiosulfate and sulfite. Geochim. Cosmochim. Acta 62, 2585–2595. doi:10.1016/S0016-7037(98)00167-7.

Han, Y., and Perner, M. (2016). Sulfide consumption in Sulfurimonas denitrificans and heterologous expression of its three sulfide-quinone reductase homologs. J. Bacteriol. 198, 1260–1267. doi:10.1128/JB.01021-15.

Haroon, M. F., Hu, S., Shi, Y., Imelfort, M., Keller, J., Hugenholtz, P., et al. (2013). Anaerobic oxidation of methane coupled to nitrate reduction in a novel archaeal lineage. Nature 500, 567–570. doi:10.1038/nature12375.

Hoehler, T. M., Alperin, M. J., Albert, D. B., and Martens, C. S. (2001). Apparent minimum free energy requirements for methanogenic Archaea and sulfate-reducing bacteria in an anoxic marine sediment. FEMS Microbiol. Ecol. 38, 33–41. doi:10.1016/S0168-6496(01)00175-1.

Hyatt, D., Chen, G.-L., Locascio, P. F., Land, M. L., Larimer, F. W., and Hauser, L. J. (2010). Prodigal: prokaryotic gene recognition and translation initiation site identification. BMC Bioinformatics 11, 119. doi:10.1186/1471-2105-11-119.

Ino, K., Hernsdorf, A. W., Konno, U., Kouduka, M., Yanagawa, K., Kato, S., et al. (2018). Ecological and genomic profiling of anaerobic methane-oxidizing archaea in a deep granitic environment. ISME J. 12, 31–47. doi:10.1038/ismej.2017.140.

Janssen, P. H., Schuhmann, A., Bak, F., and Liesack, W. (1996). Disproportionation of inorganic sulfur compounds by the sulfate-reducing bacterium Desulfocapsa

104

thiozymogenes gen, nov., sp. nov. Arch. Microbiol. 166, 184–192. doi:10.1007/s002030050374.

Johansson, J., Stahlén, J., Taborowski, T., Pedersen, K., Lukkari, S., and Marmo, J. (2019). Working Report 2018-30, Microbial release of iron from Olkiluoto rock minerals. Posiva Oy.

Kanehisa, M., Goto, S., Sato, Y., Furumichi, M., and Tanabe, M. (2012). KEGG for integration and interpretation of large-scale molecular data sets. Nucleic Acids Res. 40, D109-14. doi:10.1093/nar/gkr988.

Kanehisa, M., Sato, Y., Furumichi, M., Morishima, K., and Tanabe, M. (2019). New approach for understanding genome variations in KEGG. Nucleic Acids Res. 47, D590–D595. doi:10.1093/nar/gky962.

Kanehisa, M., Sato, Y., and Morishima, K. (2016). BlastKOALA and GhostKOALA: KEGG Tools for Functional Characterization of Genome and Metagenome Sequences. J. Mol. Biol. 428, 726–731. doi:10.1016/j.jmb.2015.11.006.

Kotelnikova, S., and Pedersen, K. (1997). Evidence for methanogenic Archaea and homoacetogenic Bacteria in deep granitic rock aquifers. in FEMS Microbiology Reviews, 339–349. doi:10.1016/S0168-6445(97)00016-8.

Kyle, J. E., Eydal, H. S. C., Ferris, F. G., and Pedersen, K. (2008). Viruses in granitic groundwater from 69 to 450 m depth of the Äspö hard rock laboratory, Sweden. ISME J. 2, 571–574. doi:10.1038/ismej.2008.18.

Li, D., Liu, C. M., Luo, R., Sadakane, K., and Lam, T. W. (2015). MEGAHIT: An ultra-fast single-node solution for large and complex metagenomics assembly via succinct de Bruijn graph. Bioinformatics 31, 1674–1676. doi:10.1093/bioinformatics/btv033.

Loy, A., Duller, S., Baranyi, C., Mussmann, M., Ott, J., Sharon, I., et al. (2009). Reverse dissimilatory sulfite reductase as phylogenetic marker for a subgroup of sulfur-oxidizing prokaryotes. Environ. Microbiol. 11, 289–99. doi:10.1111/j.1462-2920.2008.01760.x.

Luef, B., Frischkorn, K. R., Wrighton, K. C., Holman, H.-Y. N., Birarda, G., Thomas, B. C., et al. (2015). Diverse uncultivated ultra-small bacterial cells in groundwater. Nat. Commun. 6, 6372. doi:10.1038/ncomms7372.

Marcia, M., Ermler, U., Peng, G., and Michel, H. (2009). The structure of Aquifex aeolicus sulfide:quinone oxidoreductase, a basis to understand sulfide detoxification and respiration. Proc. Natl. Acad. Sci. U. S. A. 106, 9625–30. doi:10.1073/pnas.0904165106.

Marietou, A. (2016). Nitrate reduction in sulfate-reducing bacteria. FEMS Microbiol. Lett. 363, fnw155. doi:10.1093/femsle/fnw155.

Melton, E. D., Sorokin, D. Y., Overmars, L., Chertkov, O., Clum, A., Pillay, M., et al. (2016). Complete genome sequence of Desulfurivibrio alkaliphilus strain AHT2T, a haloalkaliphilic sulfidogen from Egyptian hypersaline alkaline lakes. Stand. Genomic Sci. 11, 67. doi:10.1186/s40793-016-0184-4.

105

Menzel, P., Ng, K. L., and Krogh, A. (2016). Fast and sensitive taxonomic classification for metagenomics with Kaiju. Nat. Commun. 7, 11257. doi:10.1038/ncomms11257.

Mills, J. V., Antler, G., and Turchyn, A. V. (2016). Geochemical evidence for cryptic sulfur cycling in salt marsh sediments. Earth Planet. Sci. Lett. 453, 23–32. doi:10.1016/j.epsl.2016.08.001.

Milucka, J., Ferdelman, T. G., Polerecky, L., Franzke, D., Wegener, G., Schmid, M., et al. (2012). Zero-valent sulphur is a key intermediate in marine methane oxidation. Nature 491, 541–546. doi:10.1038/nature11656.

Motamedi, M., and Pedersen, K. (1998). Note: Desulfovibrio aespoeensis sp. nov., a mesophilic sulfate-reducing bacterium from deep groundwater at aspo hard rock laboratory, Sweden. Int. J. Syst. Bacteriol. 48, 311–315. doi:10.1099/00207713-48-1-311.

Nealson, K. H., Inagaki, F., and Takai, K. (2005). Hydrogen-driven subsurface lithoautotrophic microbial ecosystems (SLiMEs): Do they exist and why should we care? Trends Microbiol. 13, 405–410. doi:10.1016/j.tim.2005.07.010.

Orphan, V. J. (2001). Methane-Consuming Archaea Revealed by Directly Coupled Isotopic and Phylogenetic Analysis. Science (80-. ). 293, 484–487. doi:10.1126/science.1061338.

Parkes, R. J., Cragg, B., Roussel, E., Webster, G., Weightman, A., and Sass, H. (2014). A review of prokaryotic populations and processes in sub-seafloor sediments, including biosphere: Geosphere interactions. Mar. Geol. 352, 409–425. doi:10.1016/j.margeo.2014.02.009.

Parks, D. H., Imelfort, M., Skennerton, C. T., Hugenholtz, P., and Tyson, G. W. (2015). CheckM: Assessing the quality of microbial genomes recovered from isolates, single cells, and metagenomes. Genome Res. 25, 1043–1055. doi:10.1101/gr.186072.114.

Pedersen, K. (1999). Evidence for a Hydrogen-driven, Intra-terrestrial Biosphere in Deep Granitic Rock Aquifers. Atl. Canada Soc. Microb. Ecol. Available at: http://plato.acadiau.ca/isme/Symposium11/pedersen.PDF [Accessed January 30, 2018].

Pedersen, K. (2012). Subterranean microbial populations metabolize hydrogen and acetate under in situ conditions in granitic groundwater at 450 m depth in the ??sp?? Hard Rock Laboratory, Sweden. FEMS Microbiol. Ecol. 81, 217–229. doi:10.1111/j.1574-6941.2012.01370.x.

Pedersen, K. (2013). Metabolic activity of subterranean microbial communities in deep granitic groundwater supplemented with methane and H 2. ISME J. 7, 839–849. doi:10.1038/ismej.2012.144.

Pedersen, K. (2014). “Microbial life in terrestrial hard rock environments,” in Microbial life of deep biosphere, eds. J. Kallmeyer and D. Wagner (Walter de Grutyer), 63–81. doi:10.1515/9783110300130.

106

Pedersen, K., Arlinger, J., Bengtsson, A., Edlund, J., Eriksson, L., Hallbeck, L., et al. (2013). Sulphate Reduction Experiment: SURE-1 Workreport 2013-57. Available at: http://www.posiva.fi/en/databank/workreports/sulphate_reduction_experiment_sure-1.1874.xhtml [Accessed November 24, 2017].

Pedersen, K., Bengtsson, A., Edlund, J., Rabe, L., Hazen, T., Chakraborty, R., et al. (2014). Complete Genome Sequence of the Subsurface, Mesophilic Sulfate-Reducing Bacterium Desulfovibrio aespoeensis Aspo-2. Genome Announc. 2, 3–4. doi:10.1128/genomeA.00509-14.

Poser, A., Lohmayer, R., Vogt, C., Knoeller, K., Planer-Friedrich, B., Sorokin, D., et al. (2013). Disproportionation of elemental sulfur by haloalkaliphilic bacteria from soda lakes. Extremophiles 17, 1003–1012. doi:10.1007/s00792-013-0582-0.

Poser, A., Vogt, C., Knöller, K., Ahlheim, J., Weiss, H., Kleinsteuber, S., et al. (2014). Stable Sulfur and Oxygen Isotope Fractionation of Anoxic Sulfide Oxidation by Two Different Enzymatic Pathways. Environ. Sci. Technol. 48, 9094–9102. doi:10.1021/es404808r.

Posiva (2013). Olkiluoto Site description, Posiva report 2011-02. Available at: http://www.posiva.fi/en/databank/posiva_reports/olkiluoto_site_description_2011.1871.xhtml [Accessed November 24, 2017].

Purkamo, L., Bomberg, M., Kietäväinen, R., Salavirta, H., Nyyssönen, M., Nuppunen-Puputti, M., et al. (2016). Microbial co-occurrence patterns in deep Precambrian bedrock fracture fluids. Biogeosciences 13, 3091–3108. doi:10.5194/bg-13-3091-2016.

Purkamo, L., Bomberg, M., Nyyssönen, M., Kukkonen, I., Ahonen, L., and Itävaara, M. (2015). Heterotrophic Communities Supplied by Ancient Organic Carbon Predominate in Deep Fennoscandian Bedrock Fluids. Microb. Ecol. 69, 319–332. doi:10.1007/s00248-014-0490-6.

Quast, C., Pruesse, E., Yilmaz, P., Gerken, J., Schweer, T., Yarza, P., et al. (2013). The SILVA ribosomal RNA gene database project: Improved data processing and web-based tools. Nucleic Acids Res. 41, D590–D596. doi:10.1093/nar/gks1219.

Ragsdale, S. W., and Pierce, E. (2008). Acetogenesis and the Wood-Ljungdahl pathway of CO(2) fixation. Biochim. Biophys. Acta 1784, 1873–98. doi:10.1016/j.bbapap.2008.08.012.

Sekiguchi, Y., Muramatsu, M., Imachi, H., Narihiro, T., Ohashi, A., Harada, H., et al. (2008). Thermodesulfovibrio aggregans sp. nov. and Thermodesulfovibrio thiophilus sp. nov., anaerobic, thermophilic, sulfate-reducing bacteria isolated from thermophilic methanogenic sludge, and emended description of the genus Thermodesulfovibrio. Int. J. Syst. Evol. Microbiol. 58, 2541–2548. doi:10.1099/ijs.0.2008/000893-0.

Søndergaard, D., Pedersen, C. N. S., and Greening, C. (2016). HydDB: A web tool for hydrogenase classification and analysis. Sci. Rep. 6, 34212. doi:10.1038/srep34212.

107

Sorokin, D. Y., Tourova, T. P., Mußmann, M., and Muyzer, G. (2008). Dethiobacter alkaliphilus gen. nov. sp. nov., and Desulfurivibrio alkaliphilus gen. nov. sp. nov.: two novel representatives of reductive sulfur cycle from soda lakes. Extremophiles 12, 431–439. doi:10.1007/s00792-008-0148-8.

Spötl, C., and Vennemann, T. W. (2003). Continuous-flow isotope ratio mass spectrometric analysis of carbonate minerals. Rapid Commun. Mass Spectrom. 17, 1004–1006. doi:10.1002/rcm.1010.

Sun, H., Spring, S., Lapidus, A., Davenport, K., Del Rio, T. G., Tice, H., et al. (2010). Complete genome sequence of Desulfarculus baarsii type strain (2st14). Stand. Genomic Sci. 3, 276–84. doi:10.4056/sigs.1243258.

Thorup, C., Schramm, A., Findlay, A. J., Finster, K. W., and Schreiber, L. (2017). Disguised as a Sulfate Reducer: Growth of the Deltaproteobacterium Desulfurivibrio alkaliphilus by Sulfide Oxidation with Nitrate. MBio 8, e00671-17. doi:10.1128/mBio.00671-17.

Vaittinen, T., Ahokas, H., Nummela, J., and Paulamäki, S. (2011). Hydrogeological Structure Model of the Olkiluoto Site – Update in 2010, Posiva Report 2011-65. Available at: http://www.posiva.fi/tietopankki/tyoraportit/hydrogeological_structure_model_of_the_olkiluoto_site_update_in_2010.1769.xhtml?xm_freetext=structure&xm_col_type=5&cd_order=col_report_number&cd_offset=#.W27__5MzbOR.

Wang, Z., Kang, D. D., Froula, J., Egan, R., and Wang, Z. (2015). MetaBAT, an efficient tool for accurately reconstructing single genomes from complex microbial communities. PeerJ 3, e1165. doi:10.7717/peerj.1165.

Watanabe, T., Kojima, H., and Fukui, M. (2014). Complete genomes of freshwater sulfur oxidizers Sulfuricella denitrificans skB26 and Sulfuritalea hydrogenivorans sk43H: Genetic insights into the sulfur oxidation pathway of betaproteobacteria. Syst. Appl. Microbiol. 37, 387–395. doi:10.1016/j.syapm.2014.05.010.

Wu, X., Holmfeldt, K., Hubalek, V., Lundin, D., Åström, M., Bertilsson, S., et al. (2016). Microbial metagenomes from three aquifers in the Fennoscandian shield terrestrial deep biosphere reveal metabolic partitioning among populations. ISME J. 10, 1192–1203. doi:10.1038/ismej.2015.185.

Xia, Y., Lü, C., Hou, N., Xin, Y., Liu, J., Liu, H., et al. (2017). Sulfide production and oxidation by heterotrophic bacteria under aerobic conditions. ISME J. 11, 2754–2766. doi:10.1038/ismej.2017.125.

Xu, H., Luo, X., Qian, J., Pang, X., Song, J., Qian, G., et al. (2012). FastUniq: A Fast De Novo Duplicates Removal Tool for Paired Short Reads. PLoS One 7, e52249. doi:10.1371/journal.pone.0052249.

Zhao, C., Gao, Z., Qin, Q., and Ruan, L. (2012). Mangroviflexus xiamenensis gen. nov., sp. nov., a member of the family Marinilabiliaceae isolated from mangrove sediment. Int. J. Syst. Evol. Microbiol. 62, 1819–1824. doi:10.1099/ijs.0.036137-0.

109

9 SUPPLEMENTARY INFORMATION

Table S1. TVO hydrogeochemical methods. PARAMETER APPARATUS AND

METHOD DETECTION LIMIT

UNCERTAINTY OF THE MEASUREMENT

Laboratory

pH pH meter, ISO-10532 0.05 pH units TVO

Conductivity Conductivity analyser SFS-EN-27888

5 µS/cm 5% TVO

Sodium fluorescein

Fluorometry 0.7 µg/L 6% at level 15 µg/L 5% at level 200 µg/L 1% at level 275 µg/L

TVO

Alkalinity Titration SFS 3005 and SFS-EN ISO 9963-1 to the appropriate extent

0.03 mmol/L

22% at level 0.03−1 mmol/L 2% at level 1−12 mmol/L

TVO

Acidity Titration SFS 3005 and SFS-EN ISO 9963-1 to the appropriate extent

0.05 mmol/L 20% at level 0.05−0.1 mmol/L 16% at level 0.1−0.5 mmol/L 11% at level >0.5 mmol/L

TVO

DIC TOC-LCSH SFS-EN 1484 to the appropriate extent

0.4 mg/L 23% at level 0.4−5 mg/L 4% at level 5−100 mg/L

TVO

NPOC TOC-LCSH SFS-EN 1484 to the appropriate extent


TVO

Al Ca Fetot K Mg Mn Na Si Sr

ICP-OES 2 µg/L 0.2 mg/L 0.005 mg/L 0.1 mg/L 0.005 mg/L 0.003 mg/L 0.2 mg/L 0.01 mg/L 0.002 mg/L

18% at level 2−10 µg/L 10% at level 10−100 µg/L 16% 32% at level 0.005−0.05 mg/L 5% at level 0.05−10 mg/L 12% 12% 11% at level 0.003−0.05 mg/L 10% at level 0.05−2 mg/L 10% at level 2−20 mg/L 12% at level 20−500 mg/L 15% at level 0.01−0.1 mg/L 6% at level 0.1−20 mg/L 30% at level 0.002−0.05 mg/L 6% at level 0.05−20 mg/L

TVO

Fe2+ Spectrophotometry ASTM E1615-08 to the appropriate extent

0.01 mg/L 8% at level 0.01−0.3 mg/L 4% at level 0.3−0.7 mg/L

TVO

Co, Pb, Btot, Ba, Cd, Cu, As, Ni, Zn, U

ICP-MS (high resolution)

0.5 µg/L 2 µg/L 5 µg/L 0.2 µg/L

10% Near detection limit 30%

VTT

Ac IC, conductivity detector


TVO

110

PARAMETER APPARATUS AND METHOD

DETECTION LIMIT

UNCERTAINTY OF THE MEASUREMENT

Laboratory

Cl

Titration SFS 3006 to the appropriate extent

50 mg/L 6.5% at level 50−30000 mg/L TVO

Cl IC, conductivity detector


TVO

Br IC, conductivity detector


TVO

F ISE /Metrohm 905, Titrando

0.05 mg/L 3.3% at level 0.1− 5 mg/L TVO

IC, conductivity detector

0.05 mg/L 8% at level 0.05−5 mg/L

PO4 IC, conductivity detector

0.1 mg/L 8% at level 0.1−5 mg/L TVO

S2- Spectrophotometer SFS 3038 to the appropriate extent

0.02 mg/L 24% at level 0.02−0.1 mg/L 11.5% at level 0.1−0.6 mg/L

TVO

SO4 IC, conductivity detector

0.02 21% at level 0.2−1 mg/L 2% at level 1−50 mg/L

TVO

Stot H2O2 oxidation + IC 0.2 mg/L 20% at level 1 mg/L 6.8% at level 3 mg/L

TVO

NH4 Spectrophotometer SFS 3032 to the appropriate extent

0.02 mg/L 9% at level 0.02−0.15 mg/L 6% at level 0.15−1 mg/L

TVO

Total nitrogen, Ntot

TOC-LCSH +TNM-L, SFS-EN 12260 to the appropriate extent


TVO

Nitrate nitrogen, NO3-N



TVO

Nitrite nitrogen, NO2-N


0.1 mg/L 10% at level 0.1−20 mg/L TVP

18O MS

< 0.1‰ GTK 18O (SO4) MS

0.5‰ Uni

Waterloo 3H Fluid scintillation

spectrometry (LSC) after electrolytic enrichment. measured in Tritium units (TU)

0.2 TU ~ 0.3-1.0 TU Hydroisotop

2H MS

1‰ GTK 13C (DIC) MS

Precision is ~ 0.1‰ Uni

Groningen 14C (DIC) AMS

Precision is ~ 0.5% Uni

Groningen 87Sr/86Sr MS

0.003‰ GTK

34S (SO4) MS 0.1 mBq/L 0.2‰ Uni Waterloo

111

Table S2: DNA extraction and sequencing information. Drill hole Sample Sample Month Date collected Total volume (L) Total DNA (ng) Sequencing / Facility OL-KR11 KR11 M March 07/03/16 1 27.25 16S / RTL OL-KR11 KR11_May May 10/05/16 2 44 16S / UNIL

11/05/16 OL-KR11 KR11 Ju1 June 08/06/16 2 6.67 16S / UNIL

11/06/16 OL-KR11 KR11 J June 09/06/16 10.7 38.92 MetaG / CoDL

10/06/16 OL-KR11 KR11 J2 July 07/07/16 10 4.08 16S / UNIL

08/07/16 OL-KR11 KR11_A2 August 10/08/16 1 3.12 16S / UNIL

OL-KR11 KR11_S1 September 27/09/16 12 75.99 MetaG / JGI

28/09/16 16S / UNIL OL-KR11 KR11 S2 September 27/09/16 10 288.6 MetaG / RTL

29/09/16 16S / UNIL OL-KR11 KR11 S3 September 04/10/16 5.5 39.52 MetaG / JGI

16S / UNIL OL-KR11 KR11_N1 November 08/11/16 3.5 21.36 16S / UNIL

OL-KR11 KR11_N2 November 09/11/16 10 49.2 MetaG / RTL

16S / UNIL OL-KR11 KR11 N3 November 15/11/16 7.3 26.8 MetaG / RTL

16/11/16 OL-KR11 KR11 0.1 November 09/11/16 2 22.77 MetaG / JGI

16S / UNIL OL-KR13 KR13 M March 07/03/16 1 19.57 16S / RTL

OL-KR13 KR13_May May 07/05/16 2 21.56 16S / UNIL

07/05/16 OL-KR13 KR13_Ju1 June 12/06/16 2 7.08 16S / UNIL

12/06/16 OL-KR13 KR13 J June 13/06/16 9.3 25.98 MetaG / CoDL

13/06/16 OL-KR13 KR13 Ju July 11/07/16 5 31.39 MetaG / JGI

112

Drill hole Sample Sample Month Date collected Total volume (L) Total DNA (ng) Sequencing / Facility 16S / UNIL

OL-KR13 KR13 A August 13/08/16 1 7.37 16S / UNIL

OL-KR13 KR13_S1 September 27/09/16 12.5 89.92 MetaG / JGI 28/09/16 16S / UNIL

OL-KR13 KR13_S2 September 28/09/16 8.5 202.8 MetaG / RTL 29/09/16 16S / UNIL

OL-KR13 KR13 S3 September 29/09/16 8.5 131.82 MetaG / JGI 30/09/16 16S / UNIL

OL-KR13 KR13 N1 November 08/11/16 4.2 82.8 MetaG / RTL 16S / UNIL

OL-KR13 KR13 N2 November 09/11/16 4.3 47.6 MetaG / RTL 16S / UNIL

OL-KR13 KR13_0.1 November 12/11/16 3 1.64 16S / UNIL

OL-KR46 KR46 M March 08/03/16 1 31.92 16S / RTL MetaG / JGI

OL-KR46 KR46 May May 05/05/16 2 8.44 16S / UNIL 06/05/16

OL-KR46 KR46 Ju1 June 08/06/16 2 2.10 16S / UNIL 09/06/16

OL-KR46 KR46_J June 09/06/16 - 6 38.73 MetaG / CoDL 14/06/16

OL-KR46 KR46_A August 11/08/16 1 138.32 16S / UNIL

OL-KR46 KR46 S September 29/09/2016 - 10.1 2.31 MetaG / JGI 04/10/16 2.35 16S / UNIL

OL-KR46 KR46 N November 11/11/2016 - 9.6 10.35 16S / UNIL 17/11/17

OL-KR46 KR46 0.1 November 09/11/16 2 3.46 16S / UNIL

113

Table S3. Metagenomes sequenced and Assembly statistics. Assembly stats

Drill hole Sample ID Reads Gbp #Contigs Mbp Min length Max length Avg length (bp) N50

OL-KR11 KR11_J 67,121,122 10 489,141 314.59 200 235,557 643 646

OL-KR11 KR11_S1 90,291,426 14 24,850 23.37 200 730,453 940 1,352

OL-KR11 KR11_S2 209,940,244 32 518,822 390.71 200 1,563,057 753 840

OL-KR11 KR11_S3 117,448,984 18 146,314 107.71 200 1,144,411 736 764

OL-KR11 KR11_N2 68,037,490 10 217,180 165.49 200 1,562,800 762 897

OL-KR11 KR11_N3 209,531,600 32 315,209 252.24 200 1,144,748 800 972

OL-KR11 KR11_0.1 164,816,320 25 1,940,677 1,240.64 200 1,575,296 639 646

OL-KR13 KR13_J 59,792,366 9 472,518 336.93 200 243,128 713 769

OL-KR13 KR13_Ju 116,968,106 18 85,320 63.46 200 681,364 744 826

OL-KR13 KR13_S1 117,249,610 18 108,134 85.16 200 394,177 788 879

OL-KR13 KR13_S2 156,980,992 24 766,188 677.81 200 566,588 885 1,243

OL-KR13 KR13_S3 139,725,824 21 179,910 141.35 200 249,505 786 919

OL-KR13 KR13_N1 99,023,450 15 345,707 329.78 200 565,246 954 1,560

OL-KR13 KR13_N2 125,749,732 19 540,587 515.46 200 565,246 954 1,506

OL-KR46 KR46_J 75,957,050 11 150,681 130.19 200 511,721 864 1,080

OL-KR46 KR46_M 106,726,824 16 46,106 50.41 200 789,579 1'093 1,948

OL-KR46 KR46_S 91,718,374 14 34,939 47.13 200 789,578 1'349 6,619

114

Table S4. Samples collected for metaproteomic analysis Drill hole Date Sampled Volume filtered (L) Technical Replicates

OL-KR11 11th August 2016 20.7 1

OL-KR11 30th Sept - 03rd Oct 2016 4.5 3

OL-KR11 11th Nov - 14th Nov 2016 5 3

OL-KR13 14th - 15th August 2016 21 1

OL-KR13 30th Sept - 03rd Oct 2016 2.8 3

OL-KR13 10th Nov - 17th Nov 2016 5.5 3

OL-KR46 18th August 2016 7 3

115

Table S5. Groundwater chemistry results for OL-KR11, OL-KR13 and OL-KR46 analysed by TVO. Empty cell = not analysed DIC = Dissolved inorganic carbon NPOC = non purgeable organic carbon TDS = total dissolved solids -1 value for acetate = low amount of acetate was observed in chromatogram, but quantitative value was not able to determined.

TARGET DateSection,

StartSection,

endz-depth,

startz-depth,

endNH4

(mg/L)Ba

(µg/L)HCO3

(mg/L)Br

(mg/L)Ca

(mg/L)

Carbonate alkalinity (mmol/L)

Cl (mg/L)

Conductivity (mS/cm)

DIC (mg/L)

F (mg/L)

OL-KR11 8.2.2016 411 430 -366.7 -383.5 0.07 146 11 360 < 0.05 2810 9.02 28 0.6OL-KR11 14.3.2016 411 430 -366.7 -383.5 0.08 153 11 320 < 0.05 2710 8.73 29 0.6OL-KR11 26.4.2016 411 430 -366.7 -383.5 0.07 165 9.6 300 < 0.05 2560 8.4 31 0.6OL-KR11 16.5.2016 411 430 -366.7 -383.5 165 < 0.05 2550 8.33OL-KR11 30.5.2016 411 430 -366.7 -383.5 0.07 159 9.6 270 < 0.05 2560 8.33 31 0.6OL-KR11 27.6.2016 411 430 -366.7 -383.5 0.08 171 9.1 320 < 0.05 2530 8.24 31 0.6OL-KR11 19.7.2016 411 430 -366.7 -383.5 195 < 0.05 2350 7.68OL-KR11 1.8.2016 411 430 -366.7 -383.5 0.08 171 9.8 320 < 0.05 2580 8.15 32 0.6OL-KR11 22.8.2016 411 430 -366.7 -383.5 171 < 0.05 2520 8.1OL-KR11 19.9.2016 411 430 -366.7 -383.5 0.07 171 9.4 300 < 0.05 2480 8.14 33 0.6OL-KR11 11.10.2016 411 430 -366.7 -383.5 189 < 0.05 2390 7.84OL-KR11 31.10.2016 411 430 -366.7 -383.5 0.08 171 11 310 < 0.05 2530 8.24 32 0.6OL-KR11 21.11.2016 411 430 -366.7 -383.5 177 < 0.05 2550 8.08OL-KR11 11.12.2017 411 430 -366.7 -383.5 0.09 189 8 290 < 0.05 2310 7.64 36OL-KR11 7.3.2018 411 430 -366.7 -383.5 0.09 183 8.8 280 < 0.05 2390 7.76 35 0.6OL-KR11 16.7.2018 411 430 -366.7 -383.5 0.09 189 9.3 280 < 0.05 2340 7.69 36 0.7OL-KR11 27.8.2018 411 430 -366.7 -383.5 0.09 195 8.4 290 < 0.05 2250 7.6 38 0.6

116

TARGET DateFe (total)

(mg/L)Fe2+

(mg/L)Mg

(mg/L)Mn

(mg/L)NO3

(mg/L)NO2

(mg/L)N(Total)

(mg/L)NPOC

(mg/L) pHPO4

(mg/L)K

(mg/L)SiO2

(mg/L)

Sodium fluorescein

(µg/L)Na

(mg/L)OL-KR11 8.2.2016 0.22 0.21 106 0.36 < 0.4 < 0.2 18 7.7 < 0.1 9.6 12 2.3 1400OL-KR11 14.3.2016 0.58 0.57 98 0.33 < 0.2 < 0.1 16 7.8 < 0.1 8.9 12 1.9 1350OL-KR11 26.4.2016 0.16 0.15 93 0.31 < 0.2 < 0.1 15 7.8 < 0.1 9.7 13 1.7 1330OL-KR11 16.5.2016 0.17 7.7OL-KR11 30.5.2016 0.13 0.11 87 0.29 < 0.2 < 0.1 12 7.7 < 0.1 8.7 11 < 1 1330OL-KR11 27.6.2016 0.23 0.06 89 0.32 < 0.2 < 0.1 12 7.7 < 0.1 8.5 13 1 1290OL-KR11 19.7.2016 0.11 7.8OL-KR11 1.8.2016 0.082 0.06 87 0.3 < 0.2 < 0.1 12 7.6 < 0.1 9 13 1.3 1320OL-KR11 22.8.2016 0.06 7.6OL-KR11 19.9.2016 0.094 0.11 88 0.3 < 0.2 < 0.1 11 7.8 < 0.1 9.1 13 < 1 1290OL-KR11 11.10.2016 0.08 7.8OL-KR11 31.10.2016 0.075 0.08 86 0.3 < 0.2 < 0.1 11 7.8 < 0.1 9.3 13 < 1 1300OL-KR11 21.11.2016 0.1 7.8OL-KR11 11.12.2017 0.053 86 < 0.2 < 0.1 7 7.8 8.5 13 < 1 1180OL-KR11 7.3.2018 0.047 0.05 84 0.27 < 0.2 < 0.1 9.6 7.7 < 0.1 9.1 12 < 1 1210OL-KR11 16.7.2018 0.046 0.05 75 0.27 < 0.2 < 0.1 0.72 8.5 7.7 < 0.1 8.1 12 < 1 1240OL-KR11 27.8.2018 0.038 0.03 81 0.27 < 0.2 < 0.1 0.74 7.8 7.9 < 0.1 8.3 13 < 1 1250

117

TARGET DateSr

(mg/L)SO4

(mg/L)S2 -

(mg/L)S (total)

(mg/L)

Total acidity

(mmol/L)

Total alkalinity (mmol/L)

TDS (mg/L)

Acetate (mg/L)

O-18 (‰ VSMOW)

H-2 (‰ VSMOW)

O-18 (SO4) (‰VSMOW)

S-34 (SO4) (‰ VCDT)

H-3 (TU)

OL-KR11 8.2.2016 3.8 340 < 0.02 110 0.12 2.4 5200 < 0.3 -9.87 -73.3 14.04 26.59 0.6OL-KR11 14.3.2016 3.9 325 < 0.02 110 0.13 2.5 4994 -1 -9.7 -72.9 14.04 25.61 0.6OL-KR11 26.4.2016 3.7 300 0.03 98 0.12 2.7 4785 < 0.3 -10.07 -73.2 14.08 26.78 0.9OL-KR11 16.5.2016 299 0.14 97 2.7 < 3OL-KR11 30.5.2016 3.2 294 0.06 93 0.11 2.6 4733 < 0.3 -9.88 -73.2 13.78 25.86 0.8OL-KR11 27.6.2016 3.4 283 0.04 93 0.12 2.8 4719 < 0.3OL-KR11 19.7.2016 292 < 0.02 94 3.2 < 1.5 -9.92 -73.4 14.72 26.14 0.9OL-KR11 1.8.2016 3.3 279 0.13 92 0.14 2.8 4794 < 0.3OL-KR11 22.8.2016 280 0.06 91 2.8 -1OL-KR11 19.9.2016 3.4 277 0.14 92 0.13 2.8 4642 < 0.3 -9.88 -73.2 14.52 25.84 0.6OL-KR11 11.10.2016 293 0.06 96 3.1 < 0.6OL-KR11 31.10.2016 3.3 282 0.03 91 0.1 2.8 4717 < 0.3 -10.02 -73.1 13.67 26.9 0.7OL-KR11 21.11.2016 284 0.08 93 2.9 < 0.3OL-KR11 11.12.2017 3 295 < 0.02 96 0.12 3.1 4383 < 0.3 -9.88 -72.6 14.51 27.49 0.8OL-KR11 7.3.2018 3 273 0.09 92 0.1 3 4454 < 0.3 -9.84 -72.6 14.22 26.68 0.8OL-KR11 16.7.2018 2.9 262 0.17 87 0.14 3.1 4419 < 0.3 -9.98 -73.1 12.97 26.83 0.8OL-KR11 27.8.2018 3.1 264 0.29 89 0.12 3.2 4364 < 0.3 -9.95 -73.2 13.74 27.22 1

118

TARGET DateSection,

StartSection,

endz-depth,

startz-depth,

endNH4

(mg/L)Ba

(µg/L)HCO3

(mg/L)Br

(mg/L)Ca

(mg/L)


Cl (mg/L)


DIC (mg/L)

F (mg/L)

OL-KR13 22.2.2016 405.5 414.4 -330.5 -337.9 0.06 79 23 750 < 0.05 4270 12.59 11 1.1OL-KR13 14.3.2016 405.5 414.4 -330.5 -337.9 < 0.02 73 25 740 < 0.05 4310 12.71 10 1OL-KR13 25.4.2016 405.5 414.4 -330.5 -337.9 0.02 73 24 760 < 0.05 4350 12.94 10 1.1OL-KR13 16.5.2016 405.5 414.4 -330.5 -337.9 73 < 0.05 4380 12.95OL-KR13 1.6.2016 405.5 414.4 -330.5 -337.9 < 0.02 67 25 830 < 0.05 4410 13.01 9.1 1.1OL-KR13 27.6.2016 405.5 414.5 -330.5 -338.0 0.03 73 24 840 < 0.05 4500 13.15 9.4 1.1OL-KR13 19.7.2016 405.5 414.5 -330.5 -338.0 67 < 0.05 4550 13.16OL-KR13 28.7.2016 405.5 -330.5 0.04 67 25 810 < 0.05 4480 13.2 8.9 1.1OL-KR13 22.8.2016 405.5 414.5 -330.5 -338.0 67 < 0.05 4540 13.22OL-KR13 19.9.2016 405.5 414.4 -330.5 -337.9 < 0.02 61 25 830 < 0.05 4550 13.35 9.7 1.2OL-KR13 5.10.2016 405.5 414.4 -330.5 -337.9 67 < 0.05 4550 13.28OL-KR13 31.10.2016 405.5 414.4 -330.5 -337.9 0.12 59 27 820 < 0.05 4530 13.42 9.3 1.1OL-KR13 21.11.2016 405.5 414.4 -330.5 -337.9 61 < 0.05 4520 13.13OL-KR13 8.12.2016 405.5 414.4 -330.5 -337.9 25 840 4510 13.16OL-KR13 8.12.2016 405.5 414.4 -330.5 -337.9 25 830 4520 13.1OL-KR13 11.12.2017 405.5 414.4 -330.5 -337.9 0.05 67 26 830 < 0.05 4450 13.29 8.8OL-KR13 7.3.2018 405.5 414.4 -330.5 -337.9 0.03 67 27 840 < 0.05 4620 13.5 9.2 1.3OL-KR13 16.7.2018 405.5 414.4 -330.5 -337.9 0.02 61 27 850 < 0.05 4680 13.59 8.2 1.3OL-KR13 27.8.2018 405.5 414.4 -330.5 -337.9 0.05 61 26 880 < 0.05 4730 13.7 8.1 1.2

119


(mg/L)Fe2+

(mg/L)Mg

(mg/L)Mn

(mg/L)NO3

(mg/L)NO2

(mg/L)N(Total)

(mg/L)NPOC

(mg/L) pHPO4

(mg/L)K

(mg/L)SiO2

(mg/L)

Sodium fluorescein

(µg/L)Na

(mg/L)OL-KR13 22.2.2016 0.024 0.02 35 0.16 < 0.4 < 0.2 5.2 7.8 < 0.1 9.8 10 < 1 1820OL-KR13 14.3.2016 < 0.02 < 0.01 30 0.15 < 0.2 < 0.1 < 1.5 7.7 < 0.1 10 10 < 1 1840OL-KR13 25.4.2016 < 0.02 0.02 35 0.15 < 0.2 < 0.1 4.7 7.7 < 0.1 9.7 10 < 1 1880OL-KR13 16.5.2016 0.02 7.6OL-KR13 1.6.2016 0.16 0.17 32 0.18 < 0.4 < 0.2 3.8 7.5 < 0.1 9.5 11 < 1 1930OL-KR13 27.6.2016 0.19 0.22 35 0.16 < 0.2 < 0.1 4.2 7.5 < 0.1 8.8 10 < 1 1890OL-KR13 19.7.2016 0.1 7.4OL-KR13 28.7.2016 < 0.02 0.01 36 0.15 < 0.4 < 0.2 4.2 7.6 < 0.2 8.9 9.9 < 1 1890OL-KR13 22.8.2016 0.03 7.8OL-KR13 19.9.2016 < 0.02 0.02 36 0.15 < 0.4 < 0.2 4 7.6 < 0.2 9.6 10 < 1 1960OL-KR13 5.10.2016 < 0.01 7.4OL-KR13 31.10.2016 < 0.02 0.05 36 0.16 < 0.2 < 0.1 4.5 7.9 < 0.1 9.6 10 < 1 1940OL-KR13 21.11.2016 0.52 8OL-KR13 8.12.2016 0.034 35 0.15 < 0.2 < 0.1 < 1.5 7.7 < 0.1 8.6 10 1900OL-KR13 8.12.2016 0.18 35 0.15 < 0.2 < 0.1 4.4 8.1 < 0.1 8.7 10 1910OL-KR13 11.12.2017 < 0.02 35 0.3 < 0.1 < 1.2 7.6 9.2 10 < 1 1940OL-KR13 7.3.2018 < 0.02 0.02 36 0.15 < 0.2 < 0.1 3.8 7.5 < 0.1 9.5 9.5 < 1 1970OL-KR13 16.7.2018 < 0.02 0.05 34 0.14 < 0.2 < 0.1 0.57 5.2 7.7 < 0.1 9 9.6 < 1 2020OL-KR13 27.8.2018 0.023 0.02 35 0.15 < 0.2 < 0.1 0.66 4.5 8.2 < 0.1 8.9 10 < 1 2020

120

TARGET DateSr

(mg/L)SO4

(mg/L)S2 -

(mg/L)S (total)

(mg/L)

Total acidity

(mmol/L)

Total alkalinity (mmol/L)

TDS (mg/L)

Acetate (mg/L)

O-18 (‰ VSMOW)

H-2 (‰ VSMOW)


S-34 (SO4) (‰ VCDT)

H-3 (TU)

OL-KR13 22.2.2016 6.7 50 14 29 0.13 1.3 7069 < 0.3 -11.86 -84.7 11.8 39.13 0.7OL-KR13 14.3.2016 7.2 70 14 31 0.13 1.2 7130 -1 -11.76 -84.1 15.98 48.51 0.6OL-KR13 25.4.2016 7.1 55 11 30 0.13 1.2 7216 < 0.3 -11.98 -83.3 12.94 46.96 0.6OL-KR13 16.5.2016 66 14 32 1.2 < 0.6OL-KR13 1.6.2016 7.7 68 13 30 0.09 1.1 7405 < 0.6 -11.79 -84.5 15.79 49.23 0.6OL-KR13 27.6.2016 7.4 65 12 30 0.12 1.2 7467 < 0.3 -11.87 -84.9 15.57 47.58 0.6OL-KR13 19.7.2016 67 13 31 1.1 < 0.6OL-KR13 28.7.2016 7.3 70 12 31 0.1 1.1 7417 < 0.6 -11.76 -83.9 15.33 43.74 0.6OL-KR13 22.8.2016 60 12 30 1.1 < 1.5OL-KR13 19.9.2016 8 70 13 30 0.08 1 7574 < 0.6 -11.85 -84.5 15.42 43.22 0.6OL-KR13 5.10.2016 64 13 30 1.1 < 1.2OL-KR13 31.10.2016 7.5 56 15 29 < 0.05 0.97 7511 < 0.3OL-KR13 21.11.2016 59 12 31 1 < 0.3 -11.84 -84.6 13.85 43.05 0.8OL-KR13 8.12.2016 7.1 54 9.9 29 < 0.3OL-KR13 8.12.2016 7.2 54 8.8 28 < 0.3OL-KR13 11.12.2017 7.2 71 14 30 0.11 1.1 7459 < 0.3 -11.82 -84.7 0.6OL-KR13 7.3.2018 7.5 63 11 30 0.11 1.1 7662 -1 -11.86 -84.9 15.43 47.13 1.2OL-KR13 16.7.2018 7.7 54 12 26 0.11 1 7765 -1 -12.01 -85.2 5.79 34.36 1.1OL-KR13 27.8.2018 7.9 53 13 24 0.11 1 7846 -1 -11.96 -85.6 15.24 48.56 0.9

121

TARGET DateSection,

StartSection,

endz-depth,

startz-depth,

endNH4

(mg/L)Ba

(µg/L)HCO3

(mg/L)Br

(mg/L)Ca

(mg/L)


Cl (mg/L)


DIC (mg/L)

F (mg/L)

OL-KR46 28.1.2016 570.5 573.5 -528.7 -531.5 < 0.02 146 56 1820 0.14 9140 25.4 4.6 0.3OL-KR46 14.3.2016 570.5 573.5 -528.7 -531.5 < 0.02 140 59 1900 0.14 9560 26.2 2.1 0.3OL-KR46 25.4.2016 570.5 573.5 -528.7 -531.5 < 0.02 140 60 1950 0.17 9700 26.7 1.6 0.3OL-KR46 16.5.2016 570.5 573.5 -528.7 -531.5 140 0.13 9960OL-KR46 27.5.2016 570.5 573.5 -528.7 -531.5 < 0.02 134 63 2060 0.1 9760 26.8 1.7 0.3OL-KR46 27.6.2016 570.5 573.5 -528.7 -531.5 0.03 140 64 1870 0.14 10200 27.1 < 1.6 0.3OL-KR46 1.8.2016 570.5 573.5 -528.7 -531.5 < 0.02 79 39 1180 0.1 7650 21.5 3.3 0.2OL-KR46 22.8.2016 570.5 573.5 -528.7 -531.5 140 0.13 9910 26.9OL-KR46 16.9.2016 570.5 573.5 -528.7 -531.5 < 0.02 134 65 2140 0.13 10200 27.8 < 1.6OL-KR46 11.10.2016 570.5 573.5 -528.7 -531.5 134 0.16 10200OL-KR46 28.10.2016 570.5 573.5 -528.7 -531.5 0.03 128 72 2160 0.15 10300 28 < 2 0.3OL-KR46 21.11.2016 570.5 573.5 -528.7 -531.5 128 0.16 10300OL-KR46 7.3.2018 570.5 575.5 -528.7 -533.4 < 0.02 153 68 2140 0.16 10200 27.6 3.4 < 0.4OL-KR46 16.7.2018 570.5 573.5 -528.7 -531.5 0.02 570 122 74 2180 < 0.05 10500 28.4 1.5 0.3OL-KR46 24.8.2018 570.5 573.5 -528.7 -531.5 < 0.02 122 71 2310 0.09 10700 28.6 2.3 0.3OL-KR46 5.8.2019 566.0 575.0 -524.4 -532.9 0.03 600 128 65 2320 0.14 10500 28.6 < 2 0.3

122


(mg/L)Fe2+

(mg/L)Mg

(mg/L)Mn

(mg/L)NO3

(mg/L)NO2

(mg/L)N(Total)

(mg/L)NPOC

(mg/L) pHPO4

(mg/L)K

(mg/L)SiO2

(mg/L)

Sodium fluorescein (µg/L)

Na (mg/L)

OL-KR46 28.1.2016 < 0.025 0.03 91 0.12 < 0.8 < 0.4 53 9.1 < 0.2 14 9.6 1.1 3670OL-KR46 14.3.2016 < 0.025 < 0.01 78 0.092 < 0.8 < 0.4 58 8.7 < 0.4 14 8.5 < 1 3770OL-KR46 25.4.2016 < 0.025 < 0.01 72 0.1 < 0.8 < 0.4 60 9 < 0.4 14 7.4 < 1 3950OL-KR46 16.5.2016 < 0.01OL-KR46 27.5.2016 < 0.025 0.02 67 0.083 < 0.8 < 0.4 56 8.6 < 0.4 12 7.1 < 1 3900OL-KR46 27.6.2016 < 0.04 0.01 59 0.086 < 0.4 < 0.2 59 9.5 < 0.2 11 6.4 1.2 3770OL-KR46 1.8.2016 < 0.025 0.08 237 0.097 < 0.4 < 0.2 32 8.8 < 0.2 19 11 < 1 3170OL-KR46 22.8.2016 0.03 8.7OL-KR46 16.9.2016 < 0.025 < 0.01 54 < 0.015 < 0.4 < 0.2 59 8.3 < 0.2 12 7 < 1 3940OL-KR46 11.10.2016 < 0.01OL-KR46 28.10.2016 < 0.04 0.09 53 < 0.024 < 0.5 < 0.25 63 8.4 < 0.25 12 7.3 < 1 4000OL-KR46 21.11.2016 0.09OL-KR46 7.3.2018 < 0.04 < 0.01 52 < 0.024 < 0.8 < 0.4 65 8.8 < 0.4 13 5.3 < 1 4010OL-KR46 16.7.2018 < 0.04 0.01 27 0.044 < 0.5 < 0.25 3.5 76 7.3 < 0.25 10 6 < 1 4200OL-KR46 24.8.2018 < 0.04 < 0.01 25 0.05 < 0.5 < 0.25 2.9 73 8.7 < 0.25 10 6.2 < 1 4220OL-KR46 5.8.2019 < 0.04 0.03 32 0.061 < 0.4 < 0.2 3.2 83 8.6 < 0.8 9.9 6.3 < 1 4200

123

TARGET DateSr

(mg/L)SO4

(mg/L)S2 -

(mg/L)S (total)

(mg/L)

Total acidity

(mmol/L)

Total alkalinity (mmol/L) TDS (mg/L)

Acetate (mg/L)

O-18 (‰ VSMOW)

H-2 (‰ VSMOW)


S-34 (SO4) (‰ VCDT)

H-3 (TU)

OL-KR46 28.1.2016 18 123 44 79 2.4 15130 84 -9.59 -63.6 12.05 28.94 0.6OL-KR46 14.3.2016 18 103 34 69 2.3 15690 94 -9.95 -64.8 11.35 27.66 0.6OL-KR46 25.4.2016 19 87 36 62 2.3 16040 96 -10.12 -65.2 11.03 26.72 0.6OL-KR46 16.5.2016 86 34 62 2.3 96OL-KR46 27.5.2016 19 67 33 56 2.2 16120 98 -9.99 -65.4 11.73 25.55 0.6OL-KR46 27.6.2016 18 61 34 47 2.3 16230 97 -10.11 -65.8 11.66 27.65 0.6OL-KR46 1.8.2016 12 380 17 140 1.3 12790 48 -8.51 -59.3 10.24 24.26 0.6OL-KR46 22.8.2016 91 40 62 2.3 98OL-KR46 16.9.2016 20 63 34 45 2.2 16670 110 -10.1 -65.6 12.04 25.9 0.6OL-KR46 11.10.2016 65 31 50 2.2 120OL-KR46 28.10.2016 20 63 31 44 2.1 16850 120OL-KR46 21.11.2016 55 30 49 2.1 110 -10.22 -65.6 10.91 27.62 0.6OL-KR46 7.3.2018 19 30 35 44 2.5 16730 110 -10.19 -66 13.62 25.43 0.6OL-KR46 16.7.2018 20 7.4 25 28 2 17170 120 -10.66 -67.3 15.65 16.11 0.6OL-KR46 24.8.2018 20 4.4 22 24 2 17510 120 -10.56 -67.3 12.17 17.58 0.6OL-KR46 5.8.2019 21 20 28 30 2.1 17330 110

124

Table S6. Chemistry, isotope and gas data analysed by EPFL.

OL-KR11 Sulfur species

Sample Date sulfate (mM) thiosulfate (mM) Sample Date sulfide (mM) sulfide (µM) Sample Date ∂34S VCDT3.3.2016 3.35 0.00 9.3.2016 0.00 0.94 7.3.2016 25.507.3.2016 3.36 0.00 9.3.2016 0.00 0.38 7.3.2016 25.899.5.2016 3.15 0.00 7.3.2016 0.00 0.85 7.3.2016 25.671.5.2016 3.04 0.00 7.3.2016 0.00 0.75 7.3.2016 25.788.6.2016 3.06 0.00 11.6.2016 0.00 2.94 9.3.2016 25.72

11.6.2016 3.06 0.00 11.6.2016 0.01 7.82 9.3.2016 26.316.7.2016 3.02 0.00 11.6.2016 0.01 8.32 9.3.2016 25.416.7.2016 3.00 0.00 11.6.2016 0.01 12.28 9.3.2016 25.76

10.8.2016 2.93 0.00 6.7.2016 0.01 7.72 6.5.2016 25.9710.8.2016 2.83 0.00 6.7.2016 0.00 2.64 9.5.2016 26.0028.9.2016 2.94 0.00 6.7.2016 0.01 8.22 11.6.2016 24.7528.9.2016 2.96 0.00 6.7.2016 0.01 7.11 11.6.2016 17.949.11.2016 2.89 0.00 10.8.2016 0.01 11.57 6.7.2016 23.849.11.2016 2.94 0.00 10.8.2016 0.01 10.45 6.7.2016 23.62

n=14 10.8.2016 0.01 12.13 10.8.2016 24.42Average 3.04 0.00 10.8.2016 0.01 14.61 10.8.2016 24.40Std Dev 0.16 0.00 28.9.2016 0.01 13.82 28.9.2016 24.40

28.9.2016 0.01 13.48 9.11.2016 25.62Sample Date sulfate (mM) 28.9.2016 0.01 14.16 n=18

16.2.2018 3.01 9.11.2016 0.01 10.84 Average 24.8316.2.2018 3.06 9.11.2016 0.01 10.32 Std Dev 1.9016.2.2018 3.09 n=2221.8.2018 2.96 Average 0.01 8.16 Sample Date ∂34S VCDT Std Dev (n=3)21.8.2018 2.88 Std Dev 0.00 4.90 16.2.2018 26.3 0.321.8.2018 2.86 16.2.2018 25.6 0.3

Sample Date sulfide (mM) sulfide (µM) 16.2.2018 25.5 0.621.8.2018 0.02 19.4421.8.2018 0.02 22.2221.8.2018 0.02 22.92

125


Sample Date sulfate (mM) thiosulfate (mM) Sample Date sulfide (mM) Sample Date ∂34S VCDT2.3.2016 0.54 0.09 2.3.2016 0.55 7.3.2016 47.937.3.2016 0.81 0.00 2.3.2016 0.46 7.5.2016 47.157.5.2016 0.73 0.11 7.3.2016 0.59 12.6.2016 41.117.5.2016 0.62 0.14 7.3.2016 0.58 12.6.2016 44.36

12.6.2016 0.80 0.10 7.3.2016 0.57 9.7.2016 43.4512.6.2016 0.79 0.11 7.3.2016 0.56 9.7.2016 44.279.7.2016 0.79 0.10 12.6.2016 0.62 13.8.2016 45.889.7.2016 0.80 0.10 12.6.2016 0.62 13.8.2016 44.83

13.8.2016 0.76 0.07 12.6.2016 0.55 4.10.2016 43.9513.8.2016 0.65 0.07 12.6.2016 0.71 4.10.2016 45.474.10.2016 0.79 0.09 9.7.2016 0.43 n=104.10.2016 0.80 0.09 9.7.2016 0.50 Average 44.8411.11.2016 0.89 0.07 9.7.2016 0.43 Std Dev 1.9311.11.2016 0.72 0.06 9.7.2016 0.52 Std Err 0.61

n=14 13.8.2016 0.31Average 0.75 0.09 13.8.2016 0.45 Sample Date ∂34S VCDTStd Dev 0.09 0.03 13.8.2016 0.39 13.2.2018 42.5

13.8.2016 0.44Sample Date sulfate (mM) 4.10.2016 0.29

13.2.2018 0.68 4.10.2016 0.3713.2.2018 0.67 4.10.2016 0.4213.2.2018 0.65 4.10.2016 0.3221.8.2018 0.59 11.11.2016 0.0621.8.2018 0.59 11.11.2016 0.4621.8.2018 0.59 n=24

Average 0.47Std Dev 0.14

Sample Date sulfide (mM)22.8.2018 0.4422.8.2018 0.4522.8.2018 0.45

126


Sample Date sulfate (mM) thiosulfate (mM) Sample Date sulfide (mM) Sample Date ∂34S VCDT5.3.2016 1.19 0.42 4.3.2016 1.34 4.3.2016 29.848.3.2016 1.06 0.61 9.3.2016 1.55 9.3.2016 31.025.5.2016 0.87 0.32 9.3.2016 1.57 4.5.2016 28.596.5.2016 1.16 0.35 9.3.2016 1.56 6.5.2016 28.918.6.2016 1.12 0.35 7.6.2016 1.74 7.6.2016 25.459.6.2016 0.79 0.50 8.6.2016 1.43 8.6.2016 24.6611.8.2016 1.68 0.46 7.6.2016 2.17 9.8.2016 24.7011.8.2016 1.65 0.68 8.6.2016 1.42 9.8.2016 24.5427.9.2016 0.83 0.25 9.8.2016 0.83 27.9.2016 27.3627.9.2016 0.73 0.24 9.8.2016 0.95 27.9.2016 27.548.11.2016 0.88 0.19 9.8.2016 0.95 n=108.11.2016 0.78 0.17 9.8.2016 0.94 Average 27.26

n=12 27.9.2016 0.95 Std Dev 2.34Average 1.06 0.38 27.9.2016 0.90Std Dev 0.32 0.16 27.9.2016 1.35 Sample Date ∂34S VCDT

27.9.2016 1.38 13.2.2018 sulfate concentration too lowSample Date sulfate (mM) 8.11.2016 1.12

13.2.2018 0.65 n=1713.2.2018 0.67 Average 1.3024.8.2018 0.14 Std Dev 0.3624.8.2018 0.1524.8.2018 0.12 Sample Date sulfide (mM)

23.8.2018 0.7623.8.2018 0.7523.8.2018 0.91

127

OL-KR11 Carbon species

Sample Date mM CH4 Sample Date ∂13C CH4 VPBD Sample Date ∂13C DIC VPDB Sample Date mg/L DOC mM DOC Sample Date mg/L TOC mM TOC3.3.2016 0.11 3.3.2016 -41.040 6.5.2016 -14.81 3.3.2016 18.45 1.53 3.3.2016 25.75 2.143.3.2016 0.12 7.3.2016 -40.790 9.5.2016 -14.85 7.3.2016 16.11 1.34 7.3.2016 44.50 3.713.3.2016 0.10 6.5.2016 -42.009 8.6.2016 -15.02 6.5.2016 13.53 1.13 6.5.2016 46.02 3.837.3.2016 0.12 9.5.2016 -42.031 11.6.2016 -14.87 9.5.2016 14.10 1.17 9.5.2016 38.20 3.187.3.2016 0.11 8.6.2016 -42.232 6.7.2016 -14.81 8.6.2016 12.64 1.05 8.6.2016 39.04 3.257.3.2016 0.13 11.6.2016 -41.967 6.7.2016 -14.80 11.6.2016 12.64 1.05 11.6.2016 40.19 3.356.5.2016 0.19 6.7.2016 -42.478 28.9.2016 -15.38 6.7.2016 15.11 1.26 6.7.2016 11.33 0.946.5.2016 0.21 6.7.2016 -42.827 28.9.2016 -15.35 6.7.2016 15.91 1.32 6.7.2016 15.98 1.336.5.2016 0.17 11.8.2016 -43.005 9.11.2016 -15.03 11.8.2016 21.04 1.75 11.8.2016 31.32 2.619.5.2016 0.29 10.8.2016 -42.548 n= 9 10.8.2016 21.83 1.82 10.8.2016 22.65 1.899.5.2016 0.16 28.9.2016 -43.290 Average -14.99 28.9.2016 22.96 1.91 28.9.2016 22.56 1.889.5.2016 0.17 28.9.2016 -43.324 Std Dev 0.23 28.9.2016 23.13 1.92 28.9.2016 27.46 2.298.6.2016 0.20 9.11.2016 -43.695 9.11.2016 11.35 0.94 9.11.2016 14.04 1.178.6.2016 0.20 n= 13 Sample Date ∂13C DIC VPDB 17.11.2016 11.70 0.97 17.11.2016 14.79 1.238.6.2016 0.22 Average -42.40 16.2.2018 -14.92 n= 14 n= 1411.6.2016 0.20 Std Dev 0.86 16.2.2018 -15.13 Average 16.46 1.37 Average 28.13 2.3411.6.2016 0.19 16.2.2018 -15.01 Std Dev 4.25 0.35 Std Dev 11.89 0.9911.6.2016 0.206.7.2016 0.24 Sample Date mg/L DOC mM DOC Sample Date mg/L TOC mM TOC6.7.2016 0.24 13.2.2018 7.35 0.61 13.2.2018 9.04 0.756.7.2016 0.24 21.8.2018 10.91 0.91 21.8.2018 10.35 0.866.7.2016 0.24 21.8.2018 10.56 0.88 21.8.2018 10.56 0.886.7.2016 0.24 21.8.2018 10.50 0.87 21.8.2018 9.94 0.836.7.2016 0.2310.8.2016 0.2610.8.2016 0.2810.8.2016 0.27 Sample Date Acetone µM Ethanol µM11.8.2016 0.19 3.3.2016 12.21 176.2011.8.2016 0.27 7.3.2016 18.05 0.0011.8.2016 0.26 6.5.2016 5.70 20.4228.9.2016 0.24 9.5.2016 8.22 70.1928.9.2016 0.27 8.6.2016 5.03 1.9628.9.2016 0.54 11.6.2016 6.99 1.1328.9.2016 0.27 6.7.2016 13.67 3.5228.9.2016 0.23 6.7.2016 15.04 4.4828.9.2016 0.26 11.8.2016 5.22 0.009.11.2016 0.28 10.8.2016 4.76 0.009.11.2016 0.29 28.9.2016 5.40 1.319.11.2016 0.29 28.9.2016 9.72 4.61

n=39 9.11.2016 0 0Average 0.21 n= 13 12Std Dev 0.08 Average 8.46 9.78

Std Dev 5.04 20.86Sample Date mM CH4

16.2.2018 0.0716.2.2018 0.0816.2.2018 0.07

128


Sample Date mM CH4 Sample Date ∂13C CH4 VPBD Sample Date ∂13C DIC VPDB Sample Date mg/L DOC mM DOC Sample Date mg/L TOC mM TOC02.03.16 0.88 02.03.16 -42.000 07.05.16 -26.70 02.03.16 10.29 0.86 02.03.16 9.03 0.7502.03.16 0.85 07.03.16 -41.680 07.05.16 -27.41 07.03.16 4.15 0.35 07.03.16 8.76 0.7302.03.16 0.74 07.05.16 -42.525 12.06.16 -28.10 07.05.16 11.01 0.92 07.05.16 11.58 0.9607.03.16 0.92 07.05.16 -42.584 12.06.16 -28.20 07.05.16 4.76 0.40 07.05.16 13.39 1.1107.03.16 0.89 12.06.16 -42.852 09.07.16 -27.97 12.06.16 4.20 0.35 12.06.16 13.54 1.1307.03.16 1.10 12.06.16 -43.044 09.07.16 -28.21 12.06.16 4.07 0.34 12.06.16 13.68 1.1407.05.16 0.90 09.07.16 -42.823 04.10.16 -28.67 09.07.16 4.30 0.36 09.07.16 3.43 0.2907.05.16 0.98 09.07.16 -43.209 04.10.16 -28.67 09.07.16 3.28 0.27 09.07.16 4.06 0.3407.05.16 0.89 13.08.16 -43.246 11.11.16 -28.11 13.08.16 6.10 0.51 13.08.16 8.11 0.6707.05.16 0.92 13.08.16 -43.238 n= 9 13.08.16 7.81 0.65 13.08.16 6.52 0.5407.05.16 0.87 04.10.16 -43.169 Average -28.00 27.09.16 6.55 0.55 27.09.16 6.68 0.5607.05.16 0.73 04.10.16 -43.402 Std Dev 0.62 27.09.16 6.58 0.55 27.09.16 6.62 0.5512.06.16 1.04 11.11.16 -43.717 11.11.16 4.75 0.40 11.11.16 8.38 0.7012.06.16 1.10 n= 13 Sample Date ∂13C DIC VPDB 17.11.16 4.48 0.37 17.11.16 8.14 0.6812.06.16 0.92 Average -42.88 14.2.2018 -28.46 n= 14 n= 1412.06.16 1.02 Std Dev 0.57 14.2.2018 -28.52 Average 5.88 0.49 Average 8.71 0.7312.06.16 1.02 14.2.2018 -28.67 Std Dev 2.38 0.20 Std Dev 3.30 0.2812.06.16 1.6909.07.16 1.12 Sample Date mg/L DOC mM DOC Sample Date mg/L TOC mM TOC09.07.16 1.03 13.2.2018 3.42 0.28 13.2.2018 3.47 0.2909.07.16 0.90 22.8.2018 5.30 0.44 21.8.2018 2.33 0.1909.07.16 1.02 22.8.2018 4.50 0.37 21.8.2018 3.40 0.2809.07.16 1.04 22.8.2018 4.62 0.38 22.8.2018 4.20 0.3509.07.16 1.1013.08.16 0.9913.08.16 0.9313.08.16 0.9013.08.16 0.91 Sample Date µM Acetone µM Ethanol13.08.16 0.84 02.03.16 8.95 013.08.16 0.88 07.03.16 9.14 004.10.16 0.94 07.05.16 4.87 004.10.16 0.79 07.05.16 1.67 004.10.16 0.83 12.06.16 3.66 1.8004.10.16 0.98 12.06.16 9.59 2.7804.10.16 0.87 09.07.16 3.57 2.5604.10.16 0.83 09.07.16 4.39 4.7011.11.16 0.71 13.08.16 5.82 12.7611.11.16 0.70 13.08.16 6.12 3.3711.11.16 0.69 04.10.16 10.99 6.12

n=39 04.10.16 8.16 2.83Average 0.95 11.11.16 0 0Std Dev 0.17 n= 13 13

Average 5.92 2.84Sample Date mM CH4 Std Dev 3.30 3.59

14.02.18 1.0814.02.18 1.1314.02.18 1.05

129


Sample Date mM CH4 Sample Date ∂13C CH4 VPBD Sample Date ∂13C VPDB Sample Date mg/L DOC mM DOC Sample Date mg/L TOC mM TOC04.03.16 1.57 4.3.2016 -32.84 4.5.2016 -14.19 5.3.2016 52.45 4.37 5.3.2016 59.24 4.9304.03.16 1.58 9.3.2016 -32.33 5.5.2016 -14.85 8.3.2016 72.97 6.08 8.3.2016 60.11 5.0004.03.16 1.51 4.5.2016 -31.662 7.6.2016 -16.76 5.5.2016 56.45 4.70 5.5.2016 65.04 5.4209.03.16 1.34 6.5.2016 -31.675 8.6.2016 -16.93 6.5.2016 56.08 4.67 6.5.2016 64.58 5.3809.03.16 2.32 7.6.2016 -31.73 27.9.2019 8.6.2016 56.12 4.67 8.6.2016 65.17 5.4309.03.16 1.95 8.6.2016 -31.603 27.9.2016 9.6.2016 57.42 4.78 9.6.2016 65.62 5.4604.05.16 1.18 9.8.2016 -32.341 8.11.2016 9.8.2016 43.14 3.59 9.8.2016 53.64 4.4704.05.16 1.14 9.8.2016 -31.988 n=4 9.8.2016 44.42 3.70 9.8.2016 49.66 4.1304.05.16 1.18 2.10.2016 -32.398 Average -15.68 27.9.2016 60.19 5.01 27.9.2016 55.40 4.6106.05.16 1.53 12.10.2016 -32.891 Std Dev 1.37 27.9.2016 58.61 4.88 27.9.2016 49.60 4.1306.05.16 15.68 8.11.2016 -32.691 8.11.2016 61.18 5.09 8.11.2016 61.64 5.1306.05.16 9.46 n=11 n=11 n=1107.06.16 1.21 Average -32.20 Average 56.28 4.69 Average 59.06 4.9207.06.16 1.25 Std Dev 0.49 Std Dev 8.08 0.67 Std Dev 6.12 0.5107.06.16 1.1908.06.16 1.69 Sample Date mg/L DOC mM DOC Sample Date mg/L TOC mM TOC08.06.16 2.16 24.8.2018 141.80 11.81 13.2.2018 58.75 4.8908.06.16 2.25 24.8.2018 141.00 11.74 23.8.2018 72.67 6.0509.08.16 0.76 24.8.2019 141.70 11.80 22.8.2018 71.60 5.9609.08.16 0.91 22.8.2018 71.83 5.9809.08.16 0.8411.08.16 1.2411.08.16 1.09 Sample Date µM Acetone µM Methanol µM Ethanol µM 2-butanol11.08.16 1.49 5.3.2016 14.540 67.018 32.184 3.18504.10.16 1.47 8.3.2016 11.645 53.125 3.61104.10.16 1.75 4.5.2016 8.480 122.679 33.157 4.03304.10.16 1.50 4.5.2016 7.456 46.401 0.000 4.14104.10.16 1.70 8.6.2016 7.197 59.367 27.638 3.79704.10.16 2.02 8.6.2016 8.097 56.620 23.407 3.75104.10.16 5.02 9.8.2016 3.739 63.270 19.081 2.12608.11.16 1.36 9.8.2016 5.782 40.798 16.893 2.25808.11.16 1.43 27.9.2016 11.575 68.613 51.897 6.69308.11.16 4.12 27.9.2016 2.865 76.151 52.758 3.512

n=29 n=10Average 1.47 Average 8.14 65.40 28.56 3.71Std Dev 0.40 Std Dev 3.64 22.72 16.71 1.25

Too small to measure

130

Hydrogen concentrationOL-KR11 OL-KR13 OL-KR46Sample Date µM H2 Sample Date µM H2 Sample Date µM H2

3.3.2016 0.00 2.3.2016 32.67 4.3.2016 0.003.3.2016 0.00 2.3.2016 29.84 4.3.2016 0.003.3.2016 0.00 0.00 2.3.2016 26.17 29.56 4.3.2016 0.00 0.007.3.2016 0.00 7.3.2016 33.95 9.3.2016 0.007.3.2016 0.00 7.3.2016 42.76 9.3.2016 0.007.3.2016 0.00 0.00 7.3.2016 81.01 52.57 9.3.2016 0.00 0.006.5.2016 0.00 7.5.2016 31.22 4.5.2016 0.006.5.2016 0.00 7.5.2016 34.85 4.5.2016 0.006.5.2016 0.00 0.00 7.5.2016 33.43 4.5.2016 0.00 0.009.5.2016 0.00 7.5.2016 34.10 6.5.2016 0.009.5.2016 0.00 7.5.2016 26.49 6.5.2016 0.009.5.2016 0.00 0.00 7.5.2016 28.01 31.35 6.5.2016 0.00 0.008.6.2016 0.00 12.6.2016 31.89 7.6.2016 0.008.6.2016 0.00 12.6.2016 39.58 7.6.2016 0.008.6.2016 0.00 0.00 12.6.2016 21.58 7.6.2016 0.00 0.00

11.6.2016 0.00 12.6.2016 28.27 8.6.2016 0.0011.6.2016 0.00 12.6.2016 32.69 8.6.2016 0.0011.6.2016 0.00 0.00 12.6.2016 150.58 50.77 8.6.2016 0.00 0.00

6.7.2016 0.00 9.7.2016 51.37 9.8.2016 0.006.7.2016 0.00 9.7.2016 31.95 9.8.2016 0.006.7.2016 0.00 9.7.2016 18.68 9.8.2016 0.00 0.006.7.2016 0.00 9.7.2016 33.73 11.8.2016 0.006.7.2016 0.00 9.7.2016 34.19 11.8.2016 0.006.7.2016 0.00 0.00 9.7.2016 52.09 37.00 11.8.2016 0.00 0.00

10.8.2016 0.00 13.8.2016 29.67 4.10.2016 0.0010.8.2016 0.00 13.8.2016 24.19 4.10.2016 0.0010.8.2016 0.00 0.00 13.8.2016 19.44 4.10.2016 0.0011.8.2016 2.2 13.8.2016 23.48 4.10.2016 0.0011.8.2016 0.00 13.8.2016 25.05 4.10.2016 0.0011.8.2016 0.00 0.73 13.8.2016 28.27 25.02 4.10.2016 0.00 0.0028.9.2016 0.00 4.10.2016 20.04 8.11.2016 0.0028.9.2016 0.00 4.10.2016 7.80 8.11.2016 0.0028.9.2016 12.75 4.10.2016 9.15 8.11.2016 0.00 0.0028.9.2016 0.00 4.10.2016 19.7028.9.2016 0.00 4.10.2016 10.8928.9.2016 0.00 2.13 4.10.2016 8.349.11.2016 0.00 4.10.2016 9.28 12.179.11.2016 0.00 11.11.2016 0.009.11.2016 0.00 0.00 11.11.2016 0.00

11.11.2016 0.00 0.00

Microbial Metabolism Resulting from the Mixing of Sulfate ... · WL Wood-Ljundahl Pathway TCA...

Documents

Transcript of Microbial Metabolism Resulting from the Mixing of Sulfate ... · WL Wood-Ljundahl Pathway TCA...