Impact of Oxygen and Pesticides on Microbial Cellulose Degradation in Aerated Agricultural Soils: A Microscaled Analysis of Processes and Prokaryotic Populations
Full text
Impact of Oxygen and Pesticides on Microbial Cellulose Degradation in Aerated Agricultural Soils: A Microscaled Analysis of Processes and Prokaryotic Populations Dissertation To obtain the Academic Degree Doctor rerum naturalium (Dr. rer. nat.) Submitted to the Faculty of Biology, Chemistry and Earth Sciences of the University of Bayreuth by Stefanie Schellenberger Bayreuth, November 2011
The present study was prepared between November 2007 and November 2011 at the Department for Ecological Microbiology at the University of Bayreuth under the direction of Prof. Harold L. Drake. The work was financially supported by the Deutsche Forschungsgemeinschaft (DFG Ko2912/3-1) and the University of Bayreuth. Vollständiger Abdruck der von der Fakultät für Biologie, Chemie und Geowissenschaften der Universität Bayreuth genehmigten Dissertation zur Erlangung des akademischen Grades eines Doktors der Naturwissenschaften (Dr. rer. nat). Dissertation eingereicht: 11.11.2011 Zulassung durch die Prüfungskommission: 28.02.2012 Wissenschaftliches Kolloquium: 24.04.2012 Amtierender Dekan: Prof. Dr. Beate Lohnert Prüfungsausschuss: Prof. Harold L. Drake (Erstgutachter) Prof. Dr. Ortwin Meyer (Zweitgutachter) Prof. Dr. Angelika Mustroph (Vorsitz) Prof. Dr. Heike Feldhaar Prof. Dr. Egbert Matzner
T ABLE O F C ONTENTS TABLES .................................................................................................. I FIGURES .............................................................................................. III EQUATIONS .......................................................................................... V ABBREVIATIONS ................................................................................. VI 1. INTRODUCTION ............................................................................. 1 1.1. Carbon Flow through Terrestrial Ecosystems ........................................... 1 1.2. Cellulose – a Major Polysaccharide in Soils .............................................. 2 1.2.1. Chemical Structure of Cellulose and its Enzymatic Hydrolysis ....... 3 1.2.1.1. Non-Complexed Enzyme Systems ................................................ 5 1.2.1.2. Complexed Enzyme Systems (Cellulosomes) ............................... 6 1.2.1.3. Extracellular Cellulose Disrupting Complex in Gram-negative Bacteria .......................................................................................... 6 1.2.2. Biological Degradation of Cellulose in Soils ..................................... 6 1.3. Diversity of Cellulose-Degrading Organisms ............................................. 9 1.3.1. Eukaryotes ............................................................................................ 9 1.3.2. Prokaroytes ........................................................................................ 10 1.4. Detection of Cellulose-Degrading Organisms ......................................... 12 1.5. Cellulose-Degrading Microorganisms in Soil Have Similar Ecological Niches ....................................................................................... 12 1.6. Environmental and Anthropogenic Factors Influence Bacterial Cellulose Degradation in Agricultural Soil ............................................... 14 1.6.1. Changing Availability of O 2 Induces Redox Changes ..................... 14 1.6.2. Pesticides Influence Biological Processes ...................................... 15 1.7. Hypotheses and Objectives ....................................................................... 16 2. MATERIAL AND METHODS ........................................................ 17 2.1. Chemicals, Gases, and Media.................................................................... 17 2.2. Sampling Site and Soil Characteristics .................................................... 18
T ABLE O F C ONTENTS 2.3. Soil Microcosms ......................................................................................... 19 2.3.1. Treatments for Stable Isotope Probing (SIP) ................................... 20 2.3.1.1. Supplementation of Cellulose ...................................................... 20 2.3.1.2. Supplementation of Cellobiose and Glucose ............................... 21 2.3.2. Treatments in Self-Constructed Incubation Chambers .................. 21 2.3.2.1. Design of Incubation Chambers .................................................. 21 2.3.2.2. Incubation Conditions .................................................................. 22 2.3.3. Treatments to Resolve Impact of Pesticides on the Degradation of Saccharides ............................................................. 23 2.3.3.1. Pesticides .................................................................................... 23 2.3.3.2. Impact of Bentazon, MCPA, and Nonylphenol on the Degradation of Cellobiose ........................................................... 26 2.3.3.3. Impact of Bentazon, MCPA, and Nonylphenol on the Degradation of Cellulosic Paper Sheets ...................................... 26 2.3.3.4. Quantification of Pesticide Effects on Processes Linked to the Degradation of Cellulose and Cellobiose .................................... 27 2.4. Analytical Methods .................................................................................... 27 2.4.1. Dry Weight and Gravimetric Water Content of Soil ........................ 27 2.4.2. Total Carbon and Nitrogen Content ................................................. 27 2.4.3. Soil Texture and Soil Type ................................................................ 28 2.4.4. pH ........................................................................................................ 28 2.4.5. Redox Potential .................................................................................. 28 2.4.6. Gases .................................................................................................. 28 2.4.7. Soluble Organic Compounds ........................................................... 31 2.4.8. Cellulose ............................................................................................. 32 2.4.9. Pesticides ........................................................................................... 32 2.4.10. Inorganic Compounds ....................................................................... 33 2.4.10.1. Nitrate (NO 3- ) and Total Amounts of Ammonium (NH 4+ ), Iron, Manganese, and Sulphate .......................................................... 33 2.4.10.2. Ferrous Iron (Fe 2+ ) ...................................................................... 33
T ABLE O F C ONTENTS 2.5. Molecular Methods ..................................................................................... 33 2.5.1. Extraction of Nucleic Acids ............................................................... 33 2.5.2. Separation of DNA and RNA ............................................................. 34 2.5.2.1. Solid Phase Columns................................................................... 34 2.5.2.2. Enzymatic Separation .................................................................. 34 2.5.3. Purification and Precipitation of Nucleic Acids ............................... 34 2.5.3.1. Isopropanol/Sodium Chloride Precipitation .................................. 34 2.5.3.2. Gel Extraction .............................................................................. 34 2.5.3.3. Filter Plates .................................................................................. 35 2.5.4. Quality Control and Quantification of Nucleic Acids ...................... 35 2.5.4.1. Spectrophotometry ...................................................................... 35 2.5.4.2. Pico-/RiboGreen-Based Quantification ........................................ 35 2.5.5. Agarose Gel Electrophoresis ............................................................ 36 2.5.6. rRNA Stable Isotope Probing (RNA SIP) .......................................... 36 2.5.6.1. Density Gradient Centrifugation ................................................... 37 2.5.6.2. Fractionation of Gradients ............................................................ 38 2.5.6.3. Measurement of the Density of the Fractions .............................. 38 2.5.6.4. RNA Precipitation ........................................................................ 39 2.5.7. Reverse Transcription of RNA into cDNA ........................................ 39 2.5.8. In Vitro Transcription of DNA into RNA ............................................ 39 2.5.9. Polymerase Chain Reaction (PCR) ................................................... 39 2.5.9.1. Primers and Thermal Protocols ................................................... 40 2.5.9.2. 16S/18S rRNA Genes .................................................................. 42 2.5.9.3. Clone Insert Sequences ............................................................... 43 2.5.10. Quantitative PCR (qPCR) ................................................................... 44 2.5.10.1. QPCR Primers and Thermal Protocols ........................................ 45 2.5.10.2. Preparation of Quantitative DNA Standards ................................ 48 2.5.10.3. Evaluation of the Specificity of Assays in Soil Samples ............... 49
T ABLE O F C ONTENTS 2.5.10.4. Correction of PCR Inhibition and Calculation of Transcript Numbers ...................................................................................... 49 2.5.11. Terminal Restriction Fragment Length Polymorphism (tRFLP) Analysis .............................................................................................. 51 2.5.11.1. Mung Bean Endonuclease Digestion .......................................... 52 2.5.11.2. Restriction Digestion ................................................................... 52 2.5.11.3. Denaturing Polyacrylamide Gelelectrophoresis (PAGE) ............. 52 2.5.11.4. Analysis of tRFLP Profiles ........................................................... 53 2.5.12. Construction of Gene Libraries ........................................................ 53 2.5.12.1. Ligation ........................................................................................ 53 2.5.12.2. Transformation ............................................................................ 54 2.5.12.3. Blue/white Screening of Clones .................................................. 55 2.5.13. Sequencing ........................................................................................ 55 2.5.14. Gene Sequence Analysis .................................................................. 56 2.5.14.1. Processing of Sequences Derived from Stable Isotope Probing Experiments ................................................................... 56 2.5.14.2. Rarefaction Analysis .................................................................... 57 2.5.14.3. Calculation of Phylogenetic Trees ............................................... 57 2.5.14.4. Deposition of Sequences in Official Databases ........................... 57 2.6. Statistics ..................................................................................................... 58 2.6.1. Mean Value, Standard Deviation, and Error Propagation .............. 58 2.6.2. Test of Significance (t-Test) .............................................................. 59 3. RESULTS ...................................................................................... 60 3.1. Degradation of [ 12 C]- and [ 13 C]-Cellulose, -Cellobiose, and -Glucose under Oxic and Anoxic Conditions .......................................................... 60 3.2. Identification of Active Prokaryotes in [ 13 C]-Cellulose, -Cellobiose, and -Glucose Supplemented Treatments by RNA Stable Isotope Probing ....................................................................................................... 63 3.2.1. Distribution of RNA in Gradient Fractions and Selection of ’Light’ and ’Heavy’ Fractions for Molecular Analyses .................... 63 3.2.2. TRFLP Profiles of Archaeal 16S rRNA cDNA Sequences .............. 65
T ABLE O F C ONTENTS 3.2.3. TRFLP Profiles of Bacterial 16S rRNA cDNA Sequences ............... 67 3.2.3.1. Comparison of ‘Heavy’ and ‘Light’ Fractions in Oxic and Anoxic [ 12 C]- or [ 13 C]-Treatments ................................................. 67 3.2.3.2. Identification of Labeled tRFs by Comparison of ‘Heavy’ Fractions between [ 12 C]- and [ 13 C]-Treatments, and Phylogenetic Affiliation ................................................................. 71 3.2.3.2.1. Bacteria that Incorporated [ 13 C]-Carbon in Oxic Treatments .... 71 3.2.3.2.2. Bacteria that Incorporated [ 13 C]-Carbon in Anoxic Treatments ................................................................................ 76 3.2.4. Bacterial 16S rRNA cDNA Gene Libraries of ‘Heavy’ Fractions ..... 81 3.2.4.1. Rarefaction and Coverage ........................................................... 81 3.2.4.2. Grouping of Sequences into Family-level OTUs and their Phylogenetic Affiliation ................................................................. 84 3.2.4.2.1. Labeled Taxa ............................................................................ 84 3.2.4.2.2. Non-labeled Taxa ..................................................................... 88 3.3. Labeled Eukaryotes .................................................................................... 90 3.4. Design of Family-level qPCR Assays ........................................................ 92 3.5. Effect of Pesticides on the Degradation of Cellulose and Cellobiose ... 96 3.5.1. In Situ-relevant (‘low’) Concentrations of Pesticides ...................... 97 3.5.2. Elevated (‘high’) Concentrations of Pesticides ............................... 98 3.5.3. 16S rRNA Content of Bacterial Taxa ............................................... 103 3.6. Effect of Fluctuating Availabilities of O 2 on the Oxic and Anoxic Degradation of CMC and Cellobiose, and Evaluation of Metabolic Responses of Saccharide-utilizing Prokaryotes .................................... 106 3.6.1. Effect of the Degradation of CMC and Cellobiose on Redox Potentials, and the Formation of Degradation Products .............. 106 3.6.2. 16S rRNA Content of Prokaryotic Taxa .......................................... 109 4. DISCUSSION ...............................................................................114 4.1. Aerobic and Anaerobic Degradation of Saccharides in An Aerated Agricultural Soil ........................................................................................ 114 4.2. Active Saccharide-Degrading Prokaryotic Communities ...................... 118 4.2.1. Key Taxa that Degraded Saccharides under Oxic Conditions ..... 119
T ABLE O F C ONTENTS 4.2.2. Key Taxa that Degraded Saccharides under Anoxic Conditions 121 4.2.3. Labeled Taxa that are of Minor Importance for the Degradation of Saccharides ................................................................................. 123 4.2.4. A Large Uncultured Sub-Community is Selectively Activated during the Aerobic and Anaerobic Degradation of Cellulose ...... 125 4.2.5. Degradation of Cellulose is a Stable Community Function in the Investigated Soil ........................................................................ 126 4.3. Diversity of Metabolic Active but not [ 13 C]-Labeled Bacteria ............... 127 4.4. Diversity of Active Eukaryotes ................................................................ 128 4.5. Pesticides have Minor Effects on the Metabolism of Cellulose- Degrading Communities .......................................................................... 130 4.6. Conclusions and Model ........................................................................... 132 4.7. Limitation of the Applied Methods ......................................................... 134 4.8. Future Perspectives ................................................................................. 136 5. SUMMARY .................................................................................. 137 6. ZUSAMMENFASSUNG .............................................................. 139 7. REFERENCES ............................................................................ 141 8. ACKNOWLEDGEMENTS ........................................................... 181 9. PUBLICATIONS AND PRESENTATIONS .................................. 182 9.1. Publications and Manuscripts ................................................................ 182 9.2. Presentions with Abstracts ..................................................................... 182 9.3. Presentations without Abstracts ............................................................ 183 10. DECLARATION .......................................................................... 184 APPENDICES ...................................................................................... A.I
T ABLES I TABLES Table 1. List of bacterial phyla including cultured aerobic and anaerobic cellulolytic species. ............................................................................................................. 11 Table 2. LB medium. ....................................................................................................... 17 Table 3. SOC medium ..................................................................................................... 18 Table 4. Soil characteristics ............................................................................................. 19 Table 5. Pesticides that were tested. ............................................................................... 24 Table 6. Concentration of pesticides in treatments. ......................................................... 25 Table 7. Parameters of GC measurements. .................................................................... 29 Table 8. Bunsen solubility coefficients of carbon dioxide, methane, and molecular hydrogen at different temperatures. ................................................................... 30 Table 9. Parameters of HPLC system for soluble organic compounds. ........................... 31 Table 10. Parameters of pesticide measurements by HPLC. ............................................. 32 Table 11. Primers for amplification of 16S rRNA genes of Bacteria and Archaea, and of sequences inserted into the pGEM-T vector of clones. ...................................... 40 Table 12. Chemical composition of PCR reactions ............................................................ 41 Table 13. Thermal protocols for amplification of 16S rRNA genes and clone insert sequences. ........................................................................................................ 42 Table 14. Primers for amplification of 18S rRNA genes of Eukarya. .................................. 43 Table 15. Primers for quantification of 16S rRNA genes of Bacteria, Archaea and family-level taxa, and for inhibition correction by qPCR. .................................... 45 Table 16. Chemical composition of qPCR reactions .......................................................... 47 Table 17. Thermal protocols for amplification of 16S rRNA genes and clone insert sequences. ........................................................................................................ 48 Table 18. Ligation reaction mix. ......................................................................................... 54 Table 19. Identification and occurrence of labeled tRFs in oxic [ 13 C]-treatments. ............... 75 Table 20. Identification and occurrence of labeled tRFs in anoxic [ 13 C]-treatments. ........... 80 Table 21. Characteristics of 16S rRNA cDNA gene libraries of ’heavy’ fractions of oxic and anoxic [ 13 C]-cellulose, [ 13 C]-cellobiose, and [ 13 C]-glucose treatments. ......... 82 Table 22. Relative abundances of labeled OTUs obtained from 16S rRNA cDNA gene libraries from ‘heavy’ fractions of [ 13 C]-treatments and their phylogenetic affiliation. ........................................................................................................... 85 Table 23. Non-labeled family-level taxa obtained from 16S rRNA cDNA gene libraries from ‘heavy’ fractions of [ 13 C]-treatments, their relative abundance over all 16S rRNA cDNA sequences, and cellulolytic isolates. ....................................... 89 Table 24. Identification and occurrence of labeled eukaryotic tRFs in [ 13 C]-cellulose treatments ......................................................................................................... 91 Table 25. Experimentally established parameters for specific qPCR measurements of family-level assays............................................................................................. 93 Table 26. Specificity of qPCR assays ................................................................................ 95
A BBREVIATIONS VIII N 1. nitrogen; 2. total number of analysed sequences n 1. amount of substance, 2. number of replicates/considered values n.a. non-applicable n.i. 1. not identified 2. no inhibition N A Avogadro constant NaCl sodium chloride NaOH sodium hydroxide N bases length of amplicon NCBI National Center for Biotechnology Information n gas amount of gas in the gaseous phase n gel,p amount of physically dissolved gas n ges total amount of gas n ges,c amount of chemically dissolved gas nm nanometer nM nanomolar; nanomole per liter no. number OD 660 optical density at 660 nm OTU operational taxonomic unit p 1 standardized air pressure PAGE Polyacrylamide gel p akt actual air pressure PCR polymerase chain reaction PCR-H 2 O particle-free autoclaved water PEG polyethylene glycol Pg petagramm pg picogramm pH the negative decimal logarithm of the hydrogen ion activity in a solution pheA gene encoding a phenol hydrolyase pk a acid dissociation constant PLFA phospholipid fatty acids pmoA gene encoding the particulate methane monooxygenase ppm parts per million p ü overpressure in incubation flasks qPCR quantitative polymerase chain reaction R reverse primer R 2 stability index RDP Ribosomal Database Project RID refractive index detector RNA ribonucleic acid RNAse ribonuclease rpm rounds per minute rRNA ribosomal ribonucleic acid RT-PCR reverse transcription PCR S standard deviation Seq. sequence S i number of expected OTUs SIP stable isotope probing
A BBREVIATIONS IX SOC super optimal broth; medium sp. species SQ starting quantity t 1. ton (=1,000 kg); 2. value that implies statistical significance T 1 standardized temperature TAE tris-aceate-EDTA; buffer T akt actual temperature Taq thermostable DNA polymerase isolated from Thermus aquaticus TaqI Restriction endonuclease isolated from Thermus aquaticus TBE tris-borat-EDTA; buffer TCD thermal conductivity detector TEMED N,N,N,N-tetramethylethylendiame temp. temperature T m basic melting temperature tRF terminal restriction fragment tRFLP terminal restriction fragment length polymorphism Tris tris(hydroxymethyl)-aminomethane U unit UFZ Helmholtz-Center for Environmental Research UV ultraviolet ν variance V 1. volume; 2. volt v/v volume per volume V 1 standardized molar gas volume V akt,mol molar volume of gas under actual conditions V fl volume of liquid phase V gas volume of the gas phase VWD variable wavelength detector w/v weight per volume W g gravimetric water content X-Gal 5-Bromo-4-chloro-3-indolyl-β-D-galactopyranoside
I NTRODUCTION 1 1. I NTRODUCTION 1.1. Carbon Flow through Terrestrial Ecosystems Conversion of biomass to carbon dioxide (CO 2 ) and methane (CH 4 ) in terrestrial ecosystems is a substantial part of global carbon cycling (Amundson 2001; Bernstein et al. 2007). The global carbon cycle is a long-term balance between atmospheric carbon, and terrestrial and aquatic carbon sinks and carbon sources (Dumonceaux 2005; Falkowski et al. 2000). Atmospheric carbon occurs primarily as carbon dioxide and its concentration increased during the last two centuries to 385 ppm (Lal 2008). The total amount of carbon in the atmosphere is about 780 Pg (Figure 1) and increases about 4.1 Pg year -1 due to anthropogenic activities (Amundson 2001; Bernstein et al. 2007; Lal 2008). Figure 1. Schematic scheme of the global carbon flow. Global carbon pools and exchange of carbon with the atmosphere are displayed. Numbers represent estimated amounts of carbon in Pg year -1 . Negative numbers indicate carbon consumed by the ecosystem type. Positive numbers indicate carbon that is released to the atmosphere. Values are according to Lal 2008. Anthropogenic activities like fossil fuel burning (7.5 Pg carbon year -1 ) and intensive land use (~1.6 Pg carbon year -1 ) emit more than 9 Pg carbon year -1 (Figure 1; Lal 2008). Oceans that take up 2.5 Pg of anthropogenic carbon dioxide and terrestrial ecosystems are important sinks for atmospheric carbon dioxide (Schimel 1995). It is estimated that ~120 Pg of carbon year -1 are fixed by plants in the form of atmospheric carbon dioxide and stored in plant biomass (Dumonceaux 2005). However, 2.6 Pg more carbon is fixed by unknown terrestrial sinks (Lal 2008). Autotrophic respiration by plants (~60 Pg year -1 ) and disturbances, e.g., wild fires,
I NTRODUCTION 2 contribute to the release of large amounts of carbon dioxide back into the atmosphere (Falkowski et al. 2000; Lal 2008). Another 60 Pg carbon year -1 is returned to the atmosphere by biological decomposition of plant biomass. Soil organisms (e.g., bacteria, fungi, protists) aerobically and anaerobically metabolize plant-derived carbon, produce carbon dioxide, and close the global carbon cycle (Dumonceaux 2005; Lal 2008). An important characteristic of soil is that it has the capacity to store large amounts of atmospheric carbon dioxide. Atmospheric carbon dioxide is photosynthetically fixed and put into the soil as recalcitrant plant-derived compounds (Amundson 2001; Lal 2008). Soils take up 1.7 Pg of anthropogenic carbon year -1 (Figure 1), making them to a key ecosystem concerning the global carbon budget. The amount of carbon that is stored in soil is determined by a lot of diverse factors, e.g., climate, topographic position, temperature, potential soil biota, or human activities (Amundson 2001). In contrast, a significant part of carbon dioxide that is returned back into the atmosphere primarily depends on the activity of soil organisms that decompose plant matter and plant-derived carbon sources (Falkowski et al. 2000). Conversion of natural forests, grasslands, and wetlands to agricultural ecosystems (i) decreases the amount of organic carbon that is stored in soil and (ii) increases the amount of carbon dioxide that is released into the atmosphere (Schimel 1995; Schlesinger and Andrews 2000). Especially tillage by crop farming facilitates carbon loss from soil by improving conditions for microbial decomposition of organic carbon. Crop farming leads to better soil aeration and higher soil temperatures, and simultaneously reduces the input of plant material into soil compared to the native vegetation (Amundson 2001; Schlesinger and Andrews 2000). Fertilization or use of pesticides in agriculture influences carbon flow in agricultural soils (Lal 2008; Schlesinger and Andrews 2000) and might also impact on soil organisms that are involved in the decomposition of plant-derived organic matter. Hence, soil organisms that decompose plants are key players in the global carbon cycle since they drive a substantial part of carbon flow through terrestrial ecosystems. 1.2. Cellulose – a Major Polysaccharide in Soils Biological decomposition of plant material to carbon dioxide is a complex process that starts at structural polymers. The lignocellulose complex forms the cell wall of plants and is composed of lignin, cellulose, hemicelluloses, and other polymers (Chang 2007; Kumar et al. 2008a; Malherbe and Cloete 2002). Lignocellulose is the most abundant and renewable energy source on earth. Terrestrial plants produce as much biomass as equivalent to two-thirds of the world’s energy requirement per year (Demain et al. 2005). The polysaccharide cellulose is the major component of lignocellulose and constitutes 35 – 50% of the dry weight of plants or litter from deciduous and coniferous trees (Kögel-Knabner 2002; Lynd et al. 2002). Dependent on the stage of growth, cotton can be composed of up to 95% of cellulose (Abidi et al. 2010). However, minor amounts of cellulose are found in cell wall polymers of certain bacteria, fungi, slime molds, and amoebae (Coughlan 1985; Lynd et al. 2002; Tomme et al. 1995). It is estimated that cellulose is produced on earth by photosynthesis at least at a rate of 10 9 – 10 10 t year -1 and that a similar amount is degraded by cellulolytic, i.e., cellulose-degrading, and saccharolytic organisms, i.e., organisms that metabolize products of cellulose hydrolysis like cellobiose or glucose (Coughlan 1985; Henrissat 1994). Structure and hydrolysis of cellulose has been extensively studied due to the increasing interests of biotechnologists and others for industrial use of this biopolymer (e.g., Bayer et al.
I NTRODUCTION 3 2008; Beguin and Aubert 1994; Berghem et al. 1975; Dubos 1928; Gardner and Blackwell 1974a; Gardner and Blackwell 1974b). However, relatively little is known about the diversity and function of organisms that might participate in the degradation of cellulose in soil. 1.2.1. Chemical Structure of Cellulose and its Enzymatic Hydrolysis Cellulose is a polysaccharide composed of β-D-glucose units that are linked by β-1,4-glycosidic bonds (Gardner and Blackwell 1974b; Kumar et al. 2008a). These linear cellulose chains have a variable degree of polymerization of up to 15,000 glucose units in cotton (Coughlan 1985). Cellulose chains are linked by intermolecular hydrogen bonds and form thereby crystalline cellulose (microfibrils) (Figure 2; O’Sullivan 1997). The degree of crystallinity is variable (Teeri 1997). The polysaccharide cellulose is insoluble and cannot be directly assimilated by cellulolytic organisms (Lynd et al. 2002). Cellulolytic organisms excrete hydrolytic enzymes, the so called cellulases, outside their cell wall, because they are unable to transport cellulose across the cell membrane (Wilson 2008). The tight structure of crystalline cellulose is broken up these cellulases what releases soluble saccharides (e.g., cellodextrins, cellobiose, and glucose) that are transported into the cell and further metabolized (Desvaux 2005b; Wilson 2008). Fungal and bacterial cellulases are well studied and have been classified based on structural properties and their specific enzymatic activities (Henrissat et al. 1998; Rabinovich et al. 2002). Cellulases attack crystalline cellulose in amorphous regions, i.e., regions that are not strongly stabilized by intramolecular hydrogen bonds (Beguin and Aubert 1994; Lynd et al. 2002; O’Sullivan 1997). Cellulases may be released as non-complexed (1.2.1.1) or as complexed enzyme systems (1.2.1.2) (Bayer and Lamed 1992; Bayer et al. 1998b; Desvaux 2005b). Non-complexed cellulases are primarily synthesized by aerobes, whereas complexed cellulase systems are synthesised by cellulolytic anaerobes (e.g., by Clostridium thermocellum; Bayer et al. 1998a; Demain et al. 2005; Wilson 2009a). In general, a tight attachment of cellulolytic organisms to the substrate is required to efficiently hydrolyze cellulose and optimize the consumption of released saccharides (Beguin and Aubert 1994). Three major types of cellulases are known (according to Lynd et al. 2002): (i) endoglucanases (or 1,4-β-D-4-glucan glucanohydrolases; EC 3.2.1.4), (ii) exoglucanases, including cellobiohydrolases (or 1,4-β-D glucan-4-glucan cellobiohydrolases; EC 3.2.1.91) and cellodextrinases (or 1,4-β-D-glucan glucanohydrolases; EC 3.2.1.74), and (iii) β-glucosidases (or β-glucoside glucohydrolases; EC 3.2.1.21). β-glucosidases act as extracellular and intracellular enzymes (Bedino et al. 1985; Inglin et al. 1980; Meyer and Canevascini 1981; Mihoc and Kluepfel 1990). Cellulases are distinguished from other glycoside hydrolases by their ability to (i) hydrolyze β-1,4-glucosidic bonds between glucosyl residues and (ii) to act not exclusively at chain ends (Bayer et al. 1998a; Lynd et al. 2002).
I NTRODUCTION 4 Figure 2. Enzymatic hydrolysis of crystalline cellulose . Composite scheme based on Bayer et al. 1998a; Bisaria and Ghose 1981; Lynd et al. 2002; Teeri 1997. Enzyme groups and the step of hydrolysis at which they attack the compounds are indicated by numbers: 1, Endoglucanases (EC 3.2.1.4); 2, Exoglucanases, including cellobiohydrolases (EC 3.2.1.91) and cellodextrinase (EC 3.2.1.74); 3, β-glucosidases (EC 3.2.1.21). Dotted lines in crystalline cellulose indicate molecular hydrogen bonds. Crystalline cellulose is attacked by endoglucanases ( 1 ) at internal amorphous regions and randomly cut into linear cellulose chains of various lengths, leading to new chain ends (Figure 2). Exoglucanases ( 2 ) can also act on crystalline cellulose fibres, presumably peeling off cellulose chains (Figure 2). Exoglucanases ( 2 ) attack cellulose chains in a processive manner on reducing and non-reducing ends yielding cellodextrins, cellobiose, and glucose; with cellobiose and glucose being the major products (Figure 2). Soluble cellodextrins and cellobiose are cut by β-glucosidases ( 3 ) into β-D-glucose (Figure 2). All types of enzymes act synergistically (e.g., Beguin and Aubert 1994; Beguin 1990; Henrissat 1994; Mansfield et al. 1999; Rabinovich et al. 2002). The saccharides cellobiose and glucose are metabolized by a broad n 3-6 + Cellulose (crystalline) Cellulose (linear) β ββ β-D-Cellobiose Cellodextrines β ββ β-D-Glucose Aerobic/Anaerobic Metabolism 1 2 3 2 2 n 3-6 n 3-6 + Cellulose (crystalline) Cellulose (linear)Cellulose (linear) β ββ β-D-Cellobiose Cellodextrines β ββ β-D-Glucose Aerobic/Anaerobic Metabolism 1 2 3 2 2
I NTRODUCTION 5 diversity of aerobic and anaerobic cellulolytic and saccharolytic organisms (Bayer et al. 2006; Lynd et al. 2002; Figure 3 and Figure 4). Cellulase enzyme systems of bacteria and fungi function similar and follow the same complex induction-repression mechanisms (Beguin and Aubert 1994; Coughlan 1991; Kumar et al. 2008a; Lynd et al. 2002). Cellulases and β-glucosidases are expressed constitutively at low levels, and excessive production is regulated by further mechanisms (Beguin and Aubert 1994; Busto et al. 1995; Lynd et al. 2002). In many organisms, cellulase biosynthesis is induced in the presence of cellulose or soluble hydrolysis products like cellobiose or cellodextrins. All known cellulase systems are additionally repressed in the presence of low molecular weight carbon compounds such as glucose that are more easily metabolized than cellulose (Beguin and Aubert 1994; Stewart and Leatherwood 1976). Cellobiose is the main intermediate of the entire cellulose hydrolysis process (Corazza et al. 2005). It is not only an inducer of cellulase expression, but also a regulator for endo- and exoglucanase activity. Both types of glucanases are inhibited by cellobiose upon a certain concentration (e.g., at 0.6 mM cellobiose in cultures of C. thermocellum; Beguin and Aubert 1994; Berghem et al. 1975; Gong et al. 1977; Zhang and Lynd 2005). β-glucosidases hydrolyze cellobiose into two glucose molecules to prevent product inhibition of glucanases. Hence, the cleavage of cellobiose is often the ratelimiting step in cellulose hydrolysis (Bhatia et al. 2002; Corazza et al. 2005). β-glucosidase activity in some cellulolytic fungi is regulated via substrate and product inhibition (e.g. in Trichoderma viridae; Bhatia et al. 2002; Hong et al. 1981). In contrast, the activity of β-glucosidases in the cellulolytic bacterium C. thermocellum is not influenced by glucose (Katayeva et al. 1992). Several other substrate- and enzyme-related factors, such as degree of substrate polymerisation/crystallinity or accessible surface area, are also determinants of enzymatic cellulose degradation (Kumar et al. 2008a; Mansfield et al. 1999). 1.2.1.1. Non-Complexed Enzyme Systems Non-complexed cellulase systems are mainly found in aerobic organisms and are well studied in fungi (e.g., Trichoderma sp.; Ilmen et al. 1997; Mandels and Reese 1957; Yang et al. 2004; Zhang and Lynd 2006) and bacteria (e.g., Cellulomonas sp. and Thermobifida sp.; Chaudhary et al. 1997, Irwin et al. 1993; Irwin et al. 1998; Lamed et al. 1987; Warren 1996). Each organism that uses noncomplexed enzyme systems secretes a set of individual cellulases into the environment. The fungus Trichoderma reesei excretes a set of five endoglucanases, two exoglucanases, and two β-glucosidases, whereas species of Cellulomonas excrete at least six endoglucanases and one exoglucanase (Lynd et al. 2002). A general feature of most cellulases is a modular structure including non-catalytic cellulose-binding domains (also termed as carbohydrate-binding modules) and catalytic domains (Bayer et al. 1998a; Doi 2008; Lynd et al. 2002; Wilson 2009a). The carbohydrate-binding modules mediate attachment of the enzyme to the cellulose surface, presumably to facilitate cellulose hydrolysis by bringing the catalytic domain in close proximity to the substrate, and to initiate processing of exoglucanases (Lynd et al. 2002; Teeri et al. 1998).
I NTRODUCTION 6 1.2.1.2. Complexed Enzyme Systems (Cellulosomes) Complexed enzyme systems are typically known from organisms living in anaerobic environments (Bayer et al. 2004; Lynd et al. 2002). Cellulosomes are large multienzyme complexes with numerous subunits tightly linked (Bayer and Lamed 1986; Mayer et al. 1987). All components form a huge protein complex, also called protubozyme (Bayer et al. 1994). Cellulosomes are flexible enough to bind to the cell wall of the cellulolytic organism and simultaneously to the surface of cellulose. They mediate cellular adhesion and form a corridor between the cell and the substrate by changing their conformation after binding (Bayer et al. 1994; Lynd et al. 2002). The component that distinguishes cellulosomes from free non-complex enzyme systems is a non-catalytic high molecular weight scaffolding protein that replaces the carbohydrate-binding module (Bayer et al. 1994; Bayer et al. 2004; Doi 2008; Wilson 2009a). The cellulosome allows optimal synergism between the excreted cellulases and also minimizes the distance over which hydrolysis products must diffuse. This facilitates efficient uptake of enzymatically released saccharides by the organism (Bayer et al. 1994; Lynd et al. 2002; Schwarz 2001). Cellulosomes are well studied in anaerobic bacteria (e.g., in Clostridia sp.; Desvaux 2005a; Petitdemange et al. 1984; Schwarz 2001; Zhang and Lynd 2005), but cellulosome-like structures are also known for some anaerobic fungi that are found in the digestion tract of ruminants (Doi 2008; Lynd et al. 2002; Orpin 1975; Rabinovich et al. 2002). 1.2.1.3. Extracellular Cellulose Disrupting Complex in Gram-negative Bacteria Recently, a third mechanism for cellulose hydrolysis was proposed for two gram-negative bacteria (Wilson 2008; Wilson 2009b). The aerobe Cytophaga hutchinsonii and the anaerobe Fibrobacter succinogenes neither use complexed nor non-complexed enzyme systems for the degradation of cellulose. Both bacteria do not encode processive endo- and exocellulases as are typically for other cellulolytic organisms (Wilson 2008). C. hutchinsonii exclusively encodes for non-processive endoglucanases and lacks carbohydrate-binding modules as well as scaffolding proteins (Xie et al. 2007). F. succinogenes does also not encode for known processive cellulases or scaffolding proteins (Jun et al. 2007; Malburg et al. 1997). However, both organisms can grow efficiently on cellulose (Fields et al. 2000; Nakagawa and Yamasato 1996). It is suggested that both organisms use a similar, but not identical extracellular cellulose disrupting complex that is bound to the outer cell membrane. This complex removes linear cellulose chains from crystalline cellulose and transports them through the outer membrane into the periplasmic space. In the periplasm these chains are hydrolysed by endoglucanases and released saccharides are transported through the plasma membrane into the cell for further metabolization (Wilson 2009a). 1.2.2. Biological Degradation of Cellulose in Soils Cellulose is a major source of carbon for soil microbial communities, and is degraded under oxic and anoxic conditions (Lynd et al. 2002; Bayer et al. 2006). Decomposition of complex plant biomass is catalyzed by various aerobic and anaerobic organisms (i.e., microorganisms capable of anaerobiosis, which includes obligate anaerobes and facultative aerobes) of the domains Bacteria and Eukarya (Baldrian and Valaskova 2008; Lynd et al. 2002). For decades, it has been hypothesized that degradation of cellulose is exclusively carried out by fungi and
I NTRODUCTION 7 bacteria, but it appears that some animals produce their own cellulases (1.3.1), which differ substantially from those of their indigenous microbial community (Watanabe and Tokuda 2001; Watanabe and Tokuda 2010). Cellulolytic and saccharolytic microorganisms inhabit similar ecological niches, and compete for carbon and energy sources. However, it was shown that aerobic degradation of cellulose by pure cultures of Cellulomonas flavigena is less efficient than degradation by mixed cultures of cellulolytic and saccharolytic species (Patel and Vaughn 1973). Species of Clostridium degrade cellulose much more efficiently in the presence of other non-cellulolytic bacteria than in pure culture (Enebo 1949; Kato et al. 2004). The saccharolytic partners improve cellulose degradation by efficient removal of potentially inhibiting hydrolysis products, e.g. cellobiose and glucose. These observations indicate that interactions of taxa with different substrate spectra might be important for the efficient degradation of cellulose (1.2.1). More than 90% of the global amount of cellulose is degraded in well aerated agricultural, grassland or forest soils (Bastian et al. 2009; Kurka 2001; Leschine 1995; Lynd et al. 2002; Vinten et al. 2002). Aerobic fungi and bacteria are the main degraders of cellulosic biomass in soil (Bastian et al. 2009; de Boer et al. 2005), and hydrolysis of cellulose mainly yields cellobiose and glucose (1.2.1). These saccharides are taken up for energy conservation and carbon assimilation, which allows for cell growth and formation of biomass (Desvaux 2006). The main end products of cellulose decomposition under oxic conditions are carbon dioxide and water (H 2 O; Bayer et al. 2006; Beguin and Aubert 1994; Schmidt and Ruschmeyer 1958). Anoxic environments in which anaerobic degradation of cellulose occurs are mainly found in the digestive tract of animals (Leschine 1995; Watanabe and Tokuda 2001), deep soil sediments (Leschine 1995), composts (de Bertoldi et al. 1983), and water-saturated environments such as freshwater, marine and estuarine sediments, flooded soils, or wetlands (Conrad 1996; Glissmann and Conrad 2000; Leschine 1995). Agricultural soils (except for flooded rice fields) are usually water-unsaturated and aerated. Nonetheless, anoxic microzones with low concentrations of oxygen (O 2 ) and low redox potential occur (1.6) in which anaerobic microbial activities can take place (Küsel and Drake 1995; Pett-Ridge and Firestone 2005; Picek et al. 2000; Totsche et al. 2010). Less studies investigated the degradation of crystalline cellulose and plant-derived carbon under anoxic conditions in aerated soils (Leschine 1995; Lynd et al. 2002), although anaerobic cellulolytic Bacteria are frequently isolated from various types of aerated soil (e.g., An et al. 2005; Ohmiya et al. 2005; Ozkan et al. 2001; Rampersad et al. 1998). Aerobic fungi are also important degraders of cellulosic biomass under oxic conditions, but fungal species are apparently not important cellulose degraders in soil under anoxic conditions (Baldrian and Valaskova 2008; de Boer et al. 2005). Thus, prokaryotes, i.e., Bacteria, represent the majority of known microorganisms that are primarily responsible for the degradation of cellulose under anoxic conditions (Hu and vanBruggen 1997) In anoxic microzones, cellulose and cellulose-derived saccharides are decomposed by obligate anaerobic and facultative aerobic bacteria via an intermediary ecosystem metabolism (Drake et al. 2009; Lynd et al. 2002; Wüst et al. 2009). In permanently flodded or wetland soils, fermentation of soluble saccharides by primary fermenters (e.g., Enterobacteriaceae; Degelmann et al. 2009a) yields fatty acids, alcohols, molecular hydrogen, and carbon dioxide . These compounds serve as substrates for secondary fermenters, acetogens, and methanogens (Drake et al. 2009; Drake et al. 2008; Hamberger et al. 2008; McInerney and Struchtemeyer 2008;
I NTRODUCTION 14 Cellulolytic organisms attack cellulose with their extracellular glycosidic enzyme systems, and produce cellobiose and glucose (Figure 4). Most of the hydrolysis products are directly assimilated by the cellulolytic microbe, but some products might be released into the environment (Bayer et al. 1994; Bayer et al. 2006). The removal of free cellulose-derived sugars by saccharolytic satellite species keeps these sugars at low non-inhibitory concentrations (Bayer et al. 1994; Doi 2008), and hence, prevents cellobiose- and glucose-induced inhibition of endo- and exocellulases of cellulolytic organisms (1.2.1). Products of saccharide degradation include fatty acids, alcohols, carbon dioxide, and molecular hydrogen (Figure 4). These compounds are further assimilated and dissimilated by other non-saccharolytic satellite microorganisms (e.g., aerobic heterotrophs, nitrate reducers, secondary fermenters, iron reducers, acetogens, or methanogens) to (i) carbon dioxide under oxic and (ii) carbon dioxide and methane under anoxic conditions (1.2.2) concomitantly to the reduction of alternative electron acceptors such as nitrate (NO 3- ) or ferric iron. In summary, the degradation of cellulose in a cellulosic environment is always facilitated by a complex microbial community consisting of cellulolytic and non-cellulolytic satellite organisms. 1.6. Environmental and Anthropogenic Factors Influence Bacterial Cellulose Degradation in Agricultural Soil The degradation of cellulose is an important community function in soil ecosystems, and is catalyzed by anaerobic and aerobic saccharide-degrading microorganisms (1.2.2). The bacterial community that is responsible for this process is confronted with changing environmental conditions, because agricultural soil is normally a water-unsaturated, aerated, and highly structured environment that comprises dynamic physicochemical gradients at biogeochemical interfaces (Or et al. 2007; Six et al. 2000; Totsche et al. 2010). The distribution of oxic and anoxic microzones is variable, spatially heterogeneous, and dependent on soil moisture and soil aggregate size (Dassonville et al. 2004; Or et al. 2007; Zausig et al. 1993). Oxic and anoxic compartments co-occur on a small scale in the same soil. Aerobic and anaerobic microorganisms exist in these oxic and anoxic microzones, and their metabolic activity is determined by various chemical and physical parameters (Brune et al. 2000; Or et al. 2007; Totsche et al. 2010; Yadav and Malanson 2007). 1.6.1. Changing Availability of O 2 Induces Redox Changes In water-unsaturated soils, soil moisture content is an important soil parameter that is closely related to the availability of O 2 and that can change rapidly (Denef et al. 2001; Mansfeldt 2004; Vorenhout et al. 2004). The consumption of O 2 exceeds the diffusion of O 2 at the interface of gas-filled and water-saturated pores, and promotes the formation of anoxic microzones (e.g., inside soil aggregates, Figure. 3; Brune et al. 2000; Dassonville et al. 2004; Greenwood 1961). A sudden increase of soil moisture, e.g. after a rain event, can increase the size of these anoxic microenvironments (Denef et al. 2001; Mansfeldt 2004; Vorenhout et al. 2004). O 2 is the terminal electron acceptor for aerobic respiration (Brune et al. 2000) and the depletion of O 2 facilitates short term shifts from oxidizing to reducing conditions (Denef et al. 2001; Mansfeldt 2004; Vorenhout et al. 2004). The availability of O 2 has a tremendous impact on the redox potential of soil and also on the energy metabolism/catabolism of the microorganisms (Brune et al. 2000). It is measured as a potential difference (E h in [mV]) between an inert electrode and a reference cell
I NTRODUCTION 15 (Bohn 1971; Fiedler et al. 2007). The sequential reduction of electron acceptors in aerated soil under flooded conditions correlates with the thermodynamic theory (Peters and Conrad 1996) that predicts that the reduction of electron acceptors should be sequential in the order of their standard redox potentials (Zehnder and Stumm 1988). Thus, the redox potential of soil is (i) a suitable measure for the presence of oxidizing and reducing conditions, and (ii) for prediction of redox reactions and microbial activities like denitrification, ferric iron reduction, fermentations, and methanogenesis. After the depletion of the preferred electron acceptor O 2 alternative electron acceptors, such as nitrate, manganese (Mn 2+ ), or ferric iron, are utilized by anaerobic microorganisms, and redox potentials of soil may drop to negative values as consequence of the depletion of oxidized compounds (Dassonville et al. 2004; Peters and Conrad 1996). The rate at which redox potentials drop is dependent in part on the availability of organic carbon that serves as energy and electron source for microbial metabolism, and on the concentration of alternative electron acceptors (Bohn 1971; Brune et al. 2000; Dassonville et al. 2004). The redox potential may return to more oxidizing conditions due to the increased availability of O 2 and the activation of aerobic microorganisms after a decrease of soil moisture by drought. Thus, fluctuation of the availability of O 2 (as a direct consequence of soil moisture content) is an environmental factor that impacts on the activity of microbial taxa in agricultural soils (Brune et al. 2000; Donnelly et al. 1990; Devevre and Horwath 2000; Jacobson and Jacobson 1998; Kremen et al. 2005). However, it is not known, how single members of the cellulose-degrading community respond to fluctuations in the availability of O 2 . 1.6.2. Pesticides Influence Biological Processes Microorganisms in agricultural soils are not only challenged by fluctuating availabilities of O 2 , but can also be influenced by anthropogenic disturbances (Cox et al. 1996; Johnsen et al. 2001; Katayama and Kuwatsuka 1991; Porter and Hayden 2002). The increased usage of herbicides, fungicides, and insecticides over the past 70 years has resulted in an accumulation of pesticide residues in terrestrial and aquatic environments (Akerblom 2004; Hiller et al. 2008; Thorstensen et al. 2001). Pesticides are subjected to microbial degradation in soils (Chowdhury et al. 2008; Gonzalez et al. 2006; Müller et al. 2001; Sukul and Spiteller 2001), but their degradation is usually slow compared to natural organic compounds and occurs primarily under oxic conditions (Harrison et al. 1998; Knauber et al. 2000). Pesticide degradation is generally slower under anoxic conditions (Harrison et al. 1998; Knauber et al. 2000). Some pesticides and break down products thereof are highly mobile and bioavailable in soils (Johnsen et al. 2001; Spliid et al. 2006), and thus, might impact on soil microorganisms (Chowdhury et al. 2008; Gevao et al. 2000; Katayama and Kuwatsuka 1991; Katayama et al. 1992; Wainwright 1978). The herbicide Bentazon inhibits bacterial growth in pure culture and microbial activity in soils (Cernakova et al. 1991; Galhano et al. 2009; Marsh et al. 1978). Bentazontreated soils display decreased capacities for nitrogen mineralization and dinitrogen (N 2 ) fixation (Galhano et al. 2009; Marsh et al. 1978). Similarly, the herbicide MCPA (4-Chloro-2-methylphenoxy acetic acid) reduces growth and activity of yeast, Pseudomonas putida, and Vibrio fischeri in pure culture (Ahtiainen et al. 2003; Cabral et al. 2003), but its toxicity to soil prokaryotes and associated processes is unresolved. Nonylphenol suppresses cellular respiratory activity of bacteria (Hseu 2006, Okai et al. 2000b). The fungicide Chlorothalonil reduces cellulose decomposition in aerated soils due to inhibition of cellulolytic fungi and Bacteria
I NTRODUCTION 16 (Katayama and Kuwatsuka 1991; Katayama et al. 1992; Suyama et al. 1993a; Suyama et al. 1993b). β-glucosidase (1.2.1) activity is inhibited by the fungicide Metalaxyl in soil (Monkiedje and Spiteller 2002; Sukul 2006). Thus, the application of pesticides might alter cellulose-dependent carbon flow in soil and might influence associated microbial taxa. 1.7. Hypotheses and Objectives The biological degradation of plant-derived cellulose is an important function of agricultural ecosystems. The activity of cellulose-utilizing microbes in agricultural soil is determined by several natural and anthropogenic factors, two important ones being the distribution of O 2 and the increasing usage of pesticides. Both factors might influence the cellulose-degrading community, and might alter the degradation rates of cellulose in soil. The process of cellulose degradation has been intensively studied (Bayer et al. 2006; Berghem et al. 1975; Henrissat 1994), but it is widely unresolved if the capacity of soil to degrade cellulose is also a stable trait of agricultural ecosystems when the availability of O 2 changes rapidly. Furthermore, it is unresolved how pesticides affect trophic interactions in cellulolytic communities, and thereby, influence cellulose mineralization. There is still a large unknown diversity of cellulosedegrading microorganisms, and few studies evaluated the activity and distribution of cellulolytic bacteria in agricultural soil (Bernard et al. 2007; Haichar et al. 2007). To understand carbon flow through cellulolytic soil communities, it is necessary to identify active microorganisms and to characterize their metabolic response to fluctuations of regulating environmental factors. Thus, the following hypotheses were addressed: Hypothesis 1: Different cellulolytic and saccharolytic taxa catalyze the degradation of cellulose in response on the presence or absence of O 2 . Hypothesis 2: The selective activation of functionally redundant cellulolytic and saccharolytic taxa yields a stable degradation of cellulose under fluctuating availabilities of O 2 . Hypothesis 3: Pesticides impact on the metabolic activity of cellulosedegrading microbial taxa and decrease cellulose degradation. Objectives of the doctoral thesis were to analyze the aerobic and anaerobic dissimilation and assimilation of [ 13 C]-cellulose, [ 13 C]-cellobiose, and [ 13 C]-glucose in oxic and anoxic soil microcosms of an agricultural soil. Both prokaryotic and eukaryotic taxa that incorporated [ 13 C]-labeled carbon in their nucleic acids were identified by RNA-based stable isotope probing (1.4), with the primary focus being on the prokaryotic community. Subsequently, taxon-specific 16S rRNA primers were developed to resolve the metabolic response of major cellulolytic and saccharolytic bacterial taxa of the investigated soil (i) to fluctuating availabilities of O 2 using quantitative PCR and (ii) to pesticides during the degradation of cellulose and cellulose-derived saccharides.
M ATERIAL A ND M ETHODS 17 2. M ATERIAL AND M ETHODS 2.1. Chemicals, Gases, and Media Unless otherwise stated chemicals and gases were supplied by Aldrich (Steinheim, Germany), AppliChem (Darmstadt, Germany), BioRad (Richmond, USA), Boehringer (Mannheim, Germany), Carl Roth (Karlsruhe, Germany), Eppendorf (Hamburg, Germany), Fluka (Buchs, Switzerland), Merck (Darmstadt, Germany), Sigma (Deisenhofen, Germany), and Rießner (Lichtenfels, Germany). Deionised double-distilled water (ddH 2 O) was produced with the ultrapure water purification system Seralpur Pro 90 CN (Seral Erich Alhäuser, Ransbach-Baumbach, Germany). PCR-H 2 O was prepared by filtration (∅ 0.2 µm) of ddH 2 O and autoclaving. RNAse/DNAse-free water (DEPC-H 2 O) was achieved by treating of ddH 2 O with 0.1% (v/v) diethylpyrocarbonate (DEPC). After incubation (37°C, 1 hour, slow shaking [300 rpm]) the solution was autoclaved (120°C, 1 bar, 20 minutes) for inactivation of DEPC. Different culture media (Table 2 and Table 3) were used for transformation of competent cells and cultivation of clones. (2.5.12). Composition and preparation as follows: Table 2. LB medium (Lurani-Bertani medium). Component Amount Final Concentration Tryptone 10 g 1% Yeast extract 5 g 0.5% NaCl 5 g 85 mM ddH 2 O Ad 1,000 ml - Tryptone, yeast extract, and NaCl were mixed with 980 ml ddH 2 O. The pH was adjusted to 7.0 and ddH 2 O was added to a final volume of 1,000 ml. Subsequently, the medium was autoclaved. For the use of LB plates 1.5% (w/v) agar was added to the medium before autoclaving. The medium was poured into sterile plastic petri dishes and stored at 4°C after solidification. For blue/white-screening of clones (2.5.12.3) LB-plates (LB/Ampicillin/IPTG/X- Gal-plates) were prepared that contained ampicillin, isopropyl-β-D- thiogalactopyranoside (IPTG), and 5-Bromo-4-chloro-3-indolyl-β-D-galactopyranoside (X-Gal). Autoclaved LB-medium (Table 2) was cooled down to 50°C and 1 ml ampicillin (100 mg ml -1 ), 1 ml (IPTG, 0.5 M) and 1.6 ml X-Gal (50 mg ml -1 in N,N’- dimethylformamide) were added. The medium was poured into sterile plastic petri dishes and stored at 4°C after solidification.
M ATERIAL A ND M ETHODS 18 Table 3. SOC medium (Super-Optimal-Browth medium). Component Amount Final Concentration Tryptone 2.0 g 2.0% Yeast extract 0.5 g 0.5% NaCl solution (1 M) 1.0 ml 10 mM KCl solution (1 M) 0.25 ml 2.5 mM Mg 2+ solution (2 M) a 1.0 ml 20 mM Glucose solution (2 M) 1.0 ml 20 mM ddH 2 O Ad 100 ml - a Mg 2+ solution: 24.6 g MgSO 4 , 20.3 g MgCl 2 , ad 100 ml. Tryptone, yeast extract, NaCl, and KCl were mixed with 97 ml ddH 2 O and autoclaved. After cool down Mg 2+ and glucose (both sterile filtered, ∅ 0.2 µm) were supplied, and sterile ddH 2 O was added to final volume of 100 µl. The medium was again sterile filtered and stored at -20°C. 2.2. Sampling Site and Soil Characteristics Soil samples were taken from an agricultural soil that was planted with corn and located near the research farm Klostergut Scheyern, Germany (Table 4). Soil from the upper 20 cm was sampled randomly with a spade and samples were pooled. Soil that was not used immediately was stored at 2°C. Di rectly before use soil was homogenized by sieving (mesh size: 2 mm, Retsch, Haan, Germany).
M ATERIAL A ND M ETHODS 19 Table 4. Soil characteristics a . Klostergut Scheyern, Germany Coordinates 48°30.0 ′N, 11°20.7′E Soil type b Dystric Cambisol Texture (clay : silt : sand) c 28 : 33 : 40 (± 0 : 0 : 0) ≙ clay loam (Lt 2 ) Sampling dates November 2007, April 2008/2010, June 2009 W g (%) 19.9 ± 3.6 Total C (%) c 1.2 ± 0.0 Total N (%) c 0.2 ± 0.0 C/N ratio c 6.9 ± 0.1 pH (in water) 6.5 ± 0.4 pH (in CaCl 2 ) 6.0 ± 0.1 I norganic ions d µmol g soil DW - 1 NH 4 + 2.9 ± 4.7 NO 3 - 11.1 ± 8.9 Total Fe 533.8 ± 41.0 Total Mn 17.2 ± 1.5 SO 4 2 - 0.3 ± 0.2 a Values are the means of 3 – 21 replicates with standard deviation. b According to FAO-classification (Food and Agriculture Organization of the United Nations; Fuka et al. 2008). c Department of Soil Ecology, University of Bayreuth, Germany. Lt 2 (German soil classification; Eckelmann 2005). d Institute for Central Analytics, University of Bayreuth, Germany. W g , Gravimetric water content. DW, Dry weight. 2.3. Soil Microcosms Microcosms were prepared as slurries (soil : water, 1 : 2.5) or by using field fresh soil without additional water. Values of substrate-free controls were included in any experiment to determine indigenous carbon turnover and linked processes. Substrate-free controls were used for correction of substrate-supplemented treatments to directly show substrate-related processes. Values obtained in substrate-free controls were subtracted from values in substrate-supplemented treatments at the corresponding time point. Incubations were performed at 15°C or 20°C in the dark.
M ATERIAL A ND M ETHODS 20 2.3.1. Treatments for Stable Isotope Probing (SIP) Stable isotope probing experiments were performed to resolve aerobic and anaerobic dissimilation of three model substrates, and to identify prokaryotes that assimilated substrate-derived carbon (2.5.6). Crystalline cellulose, the disaccharide cellobiose, and the monosaccharide glucose were used in either its [ 12 C]- or [ 13 C]- isotopic form (uniformly labeled; IsoLife, Wageningen, The Netherlands). Field fresh, sieved soil was mixed with 2.5 volumes of sterile oxic or anoxic double distilled water (ddH 2 O) and mixed by extensive manual shaking for 5 minutes. Slurries were placed on ice and flushed with sterile dinitrogen (100%; Rießner, Lichtenfels, Germany) or sterile air for 1 hour. Slurries were homogenized on an endover-end shaker (Reax 2, Heidolph, Schwabach, Germany) for 1.5 hours at 5°C, divided into 80 ml aliquots, and filled in butyl rubberstoppered 500 ml (anoxic) or 1 l (oxic) flasks (Müller & Krempel, Bülach, Switzerland) with dinitrogen or air as atmosphere. After the addition of substrate (2.3.1.1., 2.3.1.2) incubation was performed at 15°C on an end-over-end shaker for 84 days (cellulose) or 24 days (cellobiose, glucose). Every 7 days cellulose-supplemented treatments were flushed for 15 minutes with sterile dinitrogen or air. Cellobiose- or glucose-supplemented treatments were flushed every 2 days (oxic) or 4 (anoxic) days. Exchange of the headspace was performed to minimize ‘cross-labeling’ by either the autotrophic or heterotrophic fixation of the [ 13 C]-carbon dioxide formed during the degradation of the primary labeled substrate. 2.3.1.1. Supplementation of Cellulose 0.2 g of [ 13 C]-cellulose or [ 12 C]-cellulose was added once due to the insolubility of this polymer. The amount of supplemented cellulose corresponds to approximately 6.6 mmol of carbon based on the assumption that 1 mol cellulose (independent of degree of polymerization) is equivalent to 1 mol glucose (MW = 180 g mol -1 ). [ 13 C]- treatments were duplicated, [ 12 C]-treatments were not replicated. Endogenous carbon turnover was monitored in unsupplemented oxic and anoxic controls (2.3). 2 ml of the liquid and 0.2 ml of the gaseous phase were taken every 7 days with sterile syringes for analysis of methane, carbon dioxide and molecular hydrogen (2.4.6), soluble organic compounds (2.4.7), inorganic compounds (2.4.10), and pH measurements (2.4.4). Gases were immediately analysed after sampling, followed by the exchange of the headspace. 2 ml of the liquid phase were additionally taken for molecular analyses at day 0, 35, and 70 (2.5), and stored at -80°C until use. In collaboration with Dr. Antonis Chatzinotas (Department of Ecological Microbiology, Helmholtz Center for Environmental Research, UFZ Leipzig, Germany) cDNA (2.5.7) samples of day 0, 35 and 70 derived from oxic and anoxic [ 12 C]- and [ 13 C]-gradients were sent to UFZ Leipzig and used to determine if also eukaryotic organisms (i.e., fungi and protists) were labeled in the experiment.
M ATERIAL A ND M ETHODS 21 2.3.1.2. Supplementation of Cellobiose and Glucose The substrate, i.e., [ 12 C]- or [ 13 C]-cellobiose or [ 12 C]- or [ 13 C]-glucose, was pulsed periodically, i.e., daily in oxic treatments and every two days in anoxic treatments. In total, 0.5 ml of a substrate stock solution (40 mM) were added 12 times. The concentration of glucose and cellobiose (0.25 mM per pulse) was five times greater than maximum concentration detected in soils (Hill et al. 2008; Medeiros et al. 2006), and thus was supplied at amounts that represented a compromise between the low in situ concentrations and that needed for effective [ 13 C] incorporation in the rRNA pool. [ 13 C]-supplemented treatments were set up in triplicates, [ 12 C]-supplemented and unsupplemented control treatments (2.3) in duplicates. The total amounts of cellobiose and glucose pulsed were 0.24 mmol, which corresponds to 2.88 and 1.44 mmol of carbon, respectively. Analogous to cellulose-supplemented treatments (2.3.1.1) methane, carbon dioxide and molecular hydrogen (2.4.6), soluble organic compounds (2.4.7), inorganic compounds (2.4.10), pH measurements (2.4.4), and exchange of the headspace were performed every 2 days (oxic) or 4 days (anoxic). Additionally, 2 ml of the liquid phase were sampled for molecular analyses that were performed at day 0, 6, and 12 in oxic treatments, and at day 0, 4, 20, and 24 in anoxic treatments (2.5). 2.3.2. Treatments in Self-Constructed Incubation Chambers The effect of periodically changing O 2 availabilities on the degradation of carboxymethyl-cellulose (CMC) and cellobiose was analysed in self-constructed incubation chambers (Figure 5). Previously identified cellulolytic and saccharolytic taxa (2.5.6) were analysed for their metabolic response on the availability of O 2 . 2.3.2.1. Design of Incubation Chambers Incubation chambers (Figure 5) were constructed in close collaboration with G. Küfner, and P. Schmidt and colleagues (Wissenschaftliche Werkstätten, University of Bayreuth, Germany). The central element was a plexiglas cylinder (inner ∅: 100 mm; height: 125 mm; wall thickness: 10 mm) that was stuck to a plastic disk (150 mm x 150 mm). The cover plate (∅: 150 mm) was equipped with a closable sensor port for insertion of pH/redox sensors, two ball valves for gas in- and gas outlet, and an exchangeable septum (∅: 10 mm, thickness: 3 mm) for sampling of the liquid phase. The cover plate was fixed gastight with four thread-forming screws. The gas inlet valve was linked by a plexiglas tube with the gas distribution plate located at the bottom of the chambers. Synthetic air or dinitrogen was supplied via the gas inlet to aerate the slurries from below. Gas was released from the gas outlet.
M ATERIAL A ND M ETHODS 22 Figure 5. Incubation chamber. Sensor port was used for pH and redox potential measurements. Sterile synthetic air or dinitrogen were supplied at a gas inlet, and slurries were aerated from below by a gas distribution plate. Liquid samples were taken at the septum with a sterile syringe. Gas was released from the gas outlet. Valves of gas inlet and outlet were closed during anoxic incubation, but at the oxic phases synthetic air was permanently flushed through the soil slurry. 2.3.2.2. Incubation Conditions 50 g of field fresh, sieved soil was filled in a chamber, 250 ml sterile oxic or anoxic ddH 2 O was added, chambers were closed, and the slurries were mixed thoroughly by manual shaking. The chambers were divided into four groups: anoxic controls (permanently anoxic, no substrate supplementation), redox controls (changing O 2 status, no substrate supplementation), cellobiose-supplemented treatments (changing O 2 status), and carboxymethyl-cellulose (CMC)-supplemented treatments (changing O 2 status). Chambers were then flushed with sterile dinitrogen (100% Rießner, Germany, 3 replicates) or with synthetic air (80% dinitrogen, 20% O 2 , Rießner, Germany; 9 replicates) for 1 hour. Cellobiose and CMC were periodically supplemented (0.5 mM substrate per pulse), i.e., the substrate was supplemented after 48, 96, and 168 hours of incubation. Total amounts of cellobiose and CMC pulsed were 0.6 mmol, which corresponds to 7.2 mmol and 4.8 mmol of carbon, respectively. After 48 hours the chambers were flushed with dinitrogen for 1.5 hours to achieve anoxic conditions in treatments formerly incubated under O 2 containing atmosphere. Re-aeration with O 2 was performed after 168 hours in redox controls and substrate-supplemented treatments by constant flushing with synthetic air until the end of incubation. Duration of the anoxic incubation was used since similar periods of O 2 depletion are likely to occur in upland soils (Pett-Ridge and Firestone 2005). Treatments were set up in triplicates and chambers were incubated at 20°C. 2 Sensor port with redox electrode Gas outlet Plexiglas tube Gas distribution plate Septum Gas inlet
M ATERIAL A ND M ETHODS 23 ml of the liquid phase were taken for determination of soluble organic compounds (2.4.7). Additionally 2 ml liquid sample were taken after 0, 48, 102, 168, and 196 hours for molecular analyses (2.5), i.e., quantification of the transcript numbers of the following domain- and family-level taxa (2.5.10): Archaea Total , Bacteria Total , Micrococcaceae/Cellulomonadaceae (phylum Actinobacteria), Kineosporiaceae/ Nocardioidaceae (phylum Actinobacteria), Cluster I Clostridiaceae (phylum Firmicutes), Planctomycetaceae (phylum Planctomycetes), the new taxa ‘Cellu1-3’ (phylum Bacteroidetes), ‘Sphingo1-4’ (phylum Bacteroidetes), and ‘Deha1’ (phylum Chloroflexi). 2.3.3. Treatments to Resolve Impact of Pesticides on the Degradation of Saccharides Oxic and anoxic soil treatments were set up to analyse the impact of pesticides on the aerobic and anaerobic degradation of cellulose and cellobiose, and the involved prokaryotic soil community. Prokaryotic taxa that were previously identified to be part of the active cellulolytic and saccharolytic soil community (2.5.6) were quantified by 16S rRNA gene based quantitative PCR (qPCR; 2.5.10). Five pesticides were tested for potential inhibitory effects in short-term preincubations (2.3.3.1). The lack of an aqueous phase with defined volume in cellulosesupplemented soil treatments prohibits expression of values in mM. Therefore, in all treatments determining pesticide impact any concentration is given in µmol g soil DW-1 . 2.3.3.1. Pesticides Five xenobiotic compounds (Table 5) were tested for their potential to impact on the degradation of the model substrate cellobiose. Compounds comprise ingredients of commercially available pesticide formulations. The term ‘pesticide’ is used to denote these compounds throughout the work.
M ATERIAL A ND M ETHODS 30 n gas , amount of gas in gaseous phase (µmol); V gas , volume of gas phase (ml); c, measured gas concentration (ppm); V akt,mol , molar volume of gas under actual conditions (ml); p akt , actual air pressure (mbar); p ü , overpressure in incubation flasks (mbar). The molar gas volume at actual conditions (V atk,mol ) was determined based on the ideal gas law (Equation 4). Equation 4. Ideal gas law. akt akt molakt pT TVp V• • • = 1 11 , p 1 , standardized air pressure (mbar); V 1 , standardized molar gas volume (ml); T 1 , standardized temperature (K); T akt , actual temperature (K). The amount of gas physically and chemically dissolved in the liquid phase is based on Equation 5 and Equation 6. Bunsen solubility coefficient (α) is variable and is dependent on the gas type and the actual temperature (Blachnik 1998; Table 8). Equation 5. Amount of physically dissolved gas. akt üakt molakt flpgel p pp Vc Vn + •••= α , , V fl , volume of liquid phase (ml); Equation 6. Amount of chemically dissolved gas. pHpk pgelcgel a nn +− •= 10 ,, Table 8. Bunsen solubility coefficients of carbon dioxide, methane, and molecular hydrogen at different temperatures (Blachnik 1998). Bunsen solubility coefficient α α α α (in water) 293.15 K (20°C) 298.15 K (25°C) CO 2 0.850 0.740 CH 4 0.032 0.029 H 2 0.018 0.015
M ATERIAL A ND M ETHODS 31 2.4.7. Soluble Organic Compounds Concentration of alcohols, fatty acids, and sugars was measured by high performance liquid chromatography (HPLC; Ehrlich et al. 1981; Matthies et al. 1993) in liquid samples of microcosm experiments (2.3). Soluble organic compounds in cellulose-supplemented treatments (2.3.3.3) were determined after mixing soil with 2.5 volumes of ddH 2 O and homogenization on an end-over-end shaker for 10 minutes at 5°C. Liquid samples were centrifuged (1-15K Sartorius microcentrifuge, Sigma Laborzentrifugen, Osterode am Harz, Germany; 13,000 x g, 10 minutes, 4°C) and supernatant was microfiltrated (HPLC nylon filter, pore volume 0.2 µm, Infochroma, Zug, Switzerland) into flange bottles with aluminium caps (VWR International, Darmstadt, Germany). Samples were directly analysed after filtration in a Hewlett Packard 1090 Series II HPLC system (Palo Alto, USA) equipped with an autosampler (Table 9). Injection volume was 20 µl. External standards with known concentrations of soluble organic compounds were used for quantification. Table 9. Parameters of HPLC system for soluble organic compounds. Hewlett Packard 1090 Series II Detectable compounds Alcohols, fatty acids, sugars Precolumn Carbo-H 4 x 3 mm ID Column Aminex Ion Exclusion HPX-87H; 300 x 7.8 mm (BioRad, Richmond, USA) Detectors G1362A RID, G1314B VWD (210 nm) (Series 1200, Agilent Technologies, Böblingen, Germany) Mobile phase H 3 PO 4 (4 mM) Oven temp. [°C] 60 Flow rate [ml min - 1 ] 0.8 Software ChemStation (Agilent Technologies, Böblingen, Germany) temp., Temperature. RID, Refractive index detector. VWD, Variable wavelength detector.
M ATERIAL A ND M ETHODS 32 2.4.8. Cellulose Degradation of cellulose (2.3.3.3) was determined by measuring cellulose weight loss of buried cellulosic paper sheets (Munier-Lamy and Borde 2000). This procedure is a suitable method to measure degradation of crystalline cellulose (Hiroki and Watanabe 1996; Jacobson and Jacobson 1998; Kurka 2001; Munier-Lamy and Borde 2000). Sheets were dried at 40°C for 4 hours in a drying cabinet (Thermo Fisher Scientific, Waltham, USA) and weighed on a MC1 analytic AC 120 S balance (AS Wägetechnik, Garding, Germany) before insertion into the soil. Sheets were gently removed with forceps after incubation. Sheets were washed with sterile ddH 2 O to remove attached soil particles, dried, and weighed. The difference in cellulose weight loss was used for calculation of the potential inhibition by pesticides (2.3.3.4). 2.4.9. Pesticides Pesticides were extracted and quantified in pesticide-supplemented treatments (2.3.3.3; Liu et al. 2011). The pesticide extraction procedure was optimized and the extraction efficiencies were 102.4 ± 4.1% for Bentazon, 98.3 ± 2.8% for MCPA, and 92.1 ± 2.5% for Nonylphenol. Bentazon and Nonylphenol were extracted with 70% ethanol, and MCPA with 10 mM NaOH (in triplicates). One volume of soil was mixed with three volumes of solvent. Mixtures were incubated for 2 hours at 15°C on an end-over-end shaker and centrifuged (13,000 rpm, 10 minutes, room temperature). Samples were microfiltered and injected in a Agilent Technologies 1200 Series HPLC instrument (Agilent Technologies, Böblingen, Germany) that was equipped with a 1200 Series diode-array detector (Agilent 1200 series, Agilent Technologies, Böblingen, Germany) (Table 10). Pesticides were separated with on a MultoHigh 100 RP 18-5µ column (250 x 4 mm) with a pre-column (20 x 4 mm; CS Chromatographie, Langerwehe, Germany) at constant temperature (30°C). Peak purity was confirmed by online spectra from 220 to 340 nm. External standards with known concentrations of pesticide were used for calibration. Signals were recorded at 230 nm with the software ChemStation (Version B.04.01; Agilent Technologies, Böblingen, Germany) Table 10. Parameters of pesticide measurements by HPLC. Bentazon Nonylphenol MCPA Injection temp. [°C] 30 30 30 Detector temp. [°C] 30 30 30 Mobile phase 50:50 (v/v %) Acetonitrile : C 2 H 3 NaO 2 (20 mM) 70:30 (v/v %) Acetonitrile : ddH 2 O 50:50 (v/v %) Acetonitrile : C 2 H 3 NaO 2 (20 mM) Flow rate [ml min - 1 ] 1.0 2.5 1.0 Retention time [min] 7 - 9 11 - 12 7 - 9 Reference This work This work Liu et al. 2011 temp., Temperature.
M ATERIAL A ND M ETHODS 33 2.4.10. Inorganic Compounds 2.4.10.1. Nitrate (NO 3- ) and Total Amounts of Ammonium (NH 4+ ), Iron, Manganese, and Sulphate Compounds were analysed by standardized colorimetric methods by coworkers of the Institute of Central Analytics, University of Bayreuth. External standards with known concentrations were used for quantification. 2.4.10.2. Ferrous Iron (Fe 2+ ) Concentrations of ferrous iron were determined photometrically (Tamura et al. 1974). 20 µl sample were mixed with 980 µl HCl (0.5 N), incubated for 1 hour at room temperature, and microfiltrated to remove soil particles. 450 µl sample were mixed with 50 µl acetate buffer (200 g CH 3 COONH 4 , 250 ml CH 3 COOH, ad 500 ml ddH 2 O) and 50 µl of a phenanthroline solution (1 g 1,10-Phenanthrolinium-chloride- monohydrate ad 200 ml ddH 2 O), leading to the development of a red-coloured Fephenanthroline-complex. 160 µl of the mixture were immediately transferred to microtiter plates (Orange Scientific, Braine-l’Alleud, Belgium) and extinction was measured at 512 nm by a µQuant microplate spectrophotometer (BioTek, Bad Friedrichshall, Germany). The software Gen5 (Version 1.04.5; BioTek, Bad Friedrichshall, Germany) and external standards were used for quantification. 2.5. Molecular Methods 2.5.1. Extraction of Nucleic Acids Nucleic acids, i.e., DNA and RNA, were co-extracted according to a modified protocol of Griffiths et al. 2000. All steps were performed in certified RNase-and DNase-free tubes. Filtered tips were used to avoid contamination of nucleic acids with hydrolytic enzymes. Centrifugation was carried out on an analytical centrifuge (1-15K Sartorius microcentrifuge, Sigma, Osterode am Harz, Germany) at 13,000 x g for 5 minutes at 4°C unless otherwise stated. 2 ml soil slurry were transferred to 2 ml screw-capped tubes (VWR International, Darmstadt, Germany), centrifuged, and supernatant was discarded. 1 g of baked (12h, 200°C) zirconium beads (0.5 g ∅ 0.1 mm, 0.5 g ∅ 0.5 mm; CarlRoth, Karlsruhe, Germany), 500 µl preheated extraction buffer (60°C, 5% CTAB, 350 mM NaCl, 120 mM potassium phosphate buffer [containing KH 2 PO 4 and K 2 HPO 4 ], pH 8), and 500 µl phenol/chloroform/isoamylalcohol (25:24:1, pH 8) were added. Samples were bead beaten two times at 5.5. m/s in a Bead Beater (FastPrep FP 120, Thermo Savant, USA) for 30 seconds and centrifuged. Supernatant was transferred to sterile 1.5 ml tubes, 500 µl chloroform/isoamylalcohol (24:1) were added, samples were mixed, and centrifuged. Supernatant was transferred to sterile 2 ml tubes and 2 volumes of precipitation buffer (30% PEG6000, 0.1 M HEPES, pH 7) were added until the solution was clear. Nucleic acids were precipitated at room temperature (20 – 25°C) for 2 hours and centrifuged to pellet nucleic acids. The pellet was washed with 500 µl ice cold ethanol (70%, -20°C) to dissolve co-precipitated salts, centrifuged, and the supernatant was discarded. Nucleic acids were dried at room temperature and dissolved in 52 µl DNase/RNase-free ddH 2 O. Nucleic acid extraction was checked spectrophotometrically (2.5.3.1) and extracts were stored at -80°C.
M ATERIAL A ND M ETHODS 34 2.5.2. Separation of DNA and RNA 2.5.2.1. Solid Phase Columns Co-extracted DNA and RNA (2.5.1) were separated with the QIAGEN RNA/DNA Mini Kit (QIAGEN, Hilden, Germany) according to the manufacturer’s protocol. The procedure was started at step 3 of the protocol for ‘Isolation of total RNA and genomic DNA from Bacteria’, a protocol suitable for total RNA amounts smaller than 40 µg. After separation DNA was dissolved in 50 µl, RNA in 30 µl DNase/RNase-free ddH 2 O. DNA and RNA were quantified (2.5.4.2) and samples were stored at -80°C. 2.5.2.2. Enzymatic Separation An alternative enzyme-based method was applied to separate nucleic acids. RNA was obtained by digestion of DNA with DNase I (1U µl -1 ; Fermentas, St. Leon- Roth, Germany). 3 µl reaction buffer (10x) and 1 µl DNAse I were mixed with 30 µl co-extract and samples were incubated on a thermomix comfort (Eppendorf, Hamburg, Germany) at 37°C for 45 minutes under constant shaking (300 rpm). DNA was obtained by digestion of RNA with RNase A (10 µg µl -1 ; Fermentas, St. Leon- Roth, Germany). 1 µl RNase A were added to 20 µl co-extract and samples were incubated for 20 minutes at room temperature (20 – 25°C). Enzymatic digestions were stopped by purification of nucleic acids by isopropanol/Na-acetate precipitation (2.5.3.1). 2.5.3. Purification and Precipitation of Nucleic Acids 2.5.3.1. Isopropanol/Sodium Chloride Precipitation Nucleic acids that were subjected to DNase I or RNase A digestion (2.5.2.2) were purified by isopropanol precipitation (Sambrook and Russell 2001). 0.7 volumes of ice cold isopropanol (100%, -20°C) and 0.1 volum e of NaCl (5 mM) were added to 1 volume of nucleic acids. Samples were incubated for at least 12 hours at -20°C, and nucleic acids were precipitated by centrifugation (18,000 x g, 45 minutes, 15°C). The DNA or RNA pellet was washed with 500 µl ice cold ethanol (70%, -20°C), dried at room temperature, and dissolved in 22 µl DNAse/RNAse-free ddH 2 O. 2 µl of nucleic acid solution were used for quantification of DNA or RNA by fluorescence based methods (2.5.4.2). DNA and RNA samples were stored at -80°C. 2.5.3.2. Gel Extraction PCR products amplified for tRFLP analyses (2.5.11) and PCR products for the construction of gene libraries (2.5.12) were purified by gel extraction (Montage Gel Extraction Kit, Millipore, Bedford, USA) after the manufacturer’s protocol. Samples were loaded on a 1% agarose gel that was prepared with modified TAE buffer (40 mM Tris-Acetate, 0,1 mM Na 2 EDTA, pH 8,0; Millipore, Bedford, USA). Agarose gel electrophoreses was performed for 1 hour at 70 mV as described (2.5.5). Bands were excised under UV-light with a sterile knife and loaded on Montage DNA Gel Extraction Device.
M ATERIAL A ND M ETHODS 35 2.5.3.3. Filter Plates PCR products that were subjected to mung bean endonuclease digestion (2.5.11.1) and PCR products representing clone insert sequences or re-amplified 16S rRNA genes thereof (2.5.9.3, 2.5.10.2) were purified with Millipore PCR 96 Cleanup Plates (Millipore Cooperation, Bedford, USA) after manufacturer’s protocol. Samples were loaded onto the filters and plates were evacuated for 12 minutes with a suction pump (KNF Neuberger, Balterswil, Switzerland). PCR products were dissolved in 25 µl tris buffer (10 mM, pH 8.5) and transferred into sterile 0.5 ml tubes (LoBind tubes, Eppendorf, Hamburg, Germany) by pipetting. 2.5.4. Quality Control and Quantification of Nucleic Acids 2.5.4.1. Spectrophotometry Quality of nucleic acid extraction was determined with a Nanodrop ND-1000 spectrophotometer (PEQLAB Biotechnology GmbH, Erlangen, Germany) and supplied software. Absorption at 260 nm, 280 nm, and 230 nm was measured in 2 µl sample. One absorption unit at 260 nm (A 260 ) corresponds to a concentration of 50 µg DNA ml -1 and 33 µg RNA ml -1 , respectively (Lottspeich and Engels 2006; Sambrook and Russell 2001). The attenuation coefficient A 260 /A 280 gives information about possible contaminations of nucleic acids with phenol, proteins, or humic acids. An ideal coefficient lies between 1.6 and 2.0 (Sambrook and Russell 2001). Additionally the absorption spectrum should be low at 230 nm, because humic acids exhibit a strong absorption at this wavelength (Tsutsuki and Kuwatsuka 1979). All DNA and RNA samples of the current study showed a good quality and were not contaminated with humic acids. 2.5.4.2. Pico-/RiboGreen-Based Quantification DNA and RNA samples of nucleic acids after extraction (2.5.2), purified PCR products (2.5.3), and PCR products that were subjected to restriction digestion for tRFLP analysis (2.5.11.2) were quantified with a fluorescence-based method (Sambrook and Russell 2001). Fluorescence-based methods are insensitive to interferences caused by proteins or other contaminations, and allow exact quantification of very low amounts of nucleic acids (minimum: 25 pg DNA ml -1 [Jones et al. 1998] or 20 ng RNA ml -1 [Labarca and Paigen 1980]). Quantification of DNA and RNA was performed in microtiter plates with the Quant-iT-PicoGreen dsDNA reagent Kit and the Quant-iT-RiboGreen Quantitation Reagent Kit (both Invitrogen, Karlsruhe, Germany), respectively. A FLx800 microplate fluorometer (BioTek, Bad Friedrichshall, Germany) was used and quantification was performed according to the manufacturer’s protocol. For quantification of RNA an assay suitable for 1 – 50 ng ml -1 was used (i.e., ‘low range’). External standards with known concentrations of DNA or RNA were entrained in every measurement. Florescence signals were recorded with the software Gen5 (BioTek, Bad Friedrichshall, Germany).
M ATERIAL A ND M ETHODS 36 2.5.5. Agarose Gel Electrophoresis PCR products were qualified with horizontal agarose gel electrophoresis, a method that separates nucleic acids in an electrical field dependent on their size (Aaij and Borst 1972; Sambrook and Russell 2001). Agarose gels (1%) were prepared with 1x TAE buffer (40 mM Tris-HCl, 20 mM CH 3 COOH, 1 mM EDTA, pH 8). The buffered mixture was heated in a MWG 756 E microwave (Clatronic, Kempen, Germany) until agarose was fully dissolved. After cool down to 60°C 0.05 mg ml -1 ethidium bromide (3,8-Diamino-5-ethyl-6-phenyl-phenanthridiumbromid) were added for visualization. Ethidium bromide intercalates with DNA and shows a strong fluorescence under UV-light after binding to DNA. Liquid gels were filled in a gel rack, transferred to a migration chamber (Techne, Jahnsdorf, Germany) after solidification, and covered with running buffer (1x TAE). Samples were prepared by mixing 5 µl sample with 1 µl 6 x Blue Orange loading dye (Promega, Madison, USA) or self prepared 6 x loading dye (0.05% bromophenol blue, 0.05% xylene cyanol, 55% glycerol). Samples were filled into the gel slots. 2 µl of a DNA ladder (MWM-1, 200 – 10,000 bp, Bilatec, Viernheim, Germany) were used for determination of product length. Electrophoresis was performed at 120 V for 20 – 30 minutes. Visualization of bands by UV-light (302 nm) was carried out on a transilluminator (UVT-20M, Herolab, Wiesloch, Germany) and a photographically based documentation system (Canon PowerShot G5, Canon, Krefeld, Germany; software: Remote Capture). 2.5.6. rRNA Stable Isotope Probing (RNA SIP) rRNA-based stable isotope probing is a culture-independent method for identification of microorganisms that are responsible for a certain process (e.g., cellulose degradation; Manefield et al. 2002b). In this study, the protocol of Whiteley et al. 2007 was used in a slightly modified form. Stable isotope probing experiments were performed in oxic and anoxic treatments (2.3.1) [ 13 C]-labeled substrates are added to the environmental sample and incubated (Figure 6). Microbial assimilation of [ 13 C]-labeled substrates results in the incorporation of heavy carbon isotopes into their biomass, e.g., nucleic acids. As a consequence, the buoyant density of DNA and RNA increases (Meselson and Stahl 1958). Total nucleic acids are extracted and RNA is separated from DNA. [ 13 C]- labeled ‘heavy’ RNA is then separated from the ‘light’ [ 12 C]-RNA by isopycnic ultracentrifugation and subsequent fractionation of the gradients. RNA of every gradient fraction is precipitated and reversely transcribed into complementary DNA (cDNA). CDNA is used for identification of tRFs related to [ 13 C]-labeled organisms by molecular methods (e.g., tRFLP fingerprinting [2.5.11], cDNA based gene libraries [2.5.12]). However, it is likely that ‘heavy’ and ‘light’ RNA are not fully separated during centrifugation. This may be due to insufficient labeling of ‘heavy’ RNA or comigration of unlabeled ‘light’ RNA with high G+C contents (causing higher buoyant densities) towards the ‘heavy’ fractions (Lüders et al. 2004; Manefield et al. 2002a; Rangel-Castro et al. 2005). Combination of different molecular tools (e.g., tRFLP and cloning) should be taken into account to avoid false interpretation of results only gained by one methodological approach (Bent and Forney 2008).
M ATERIAL A ND M ETHODS 37 Figure 6. Schematic protocol of stable isotope probing (SIP). See text for details. 2.5.6.1. Density Gradient Centrifugation A gradient solution (buoyant density: 1.790 ± 0.005 g ml -1 ) was prepared that contained 4.61 ml caesium trifluoroacetate (CsTFA, buoyant density: 2.0 ± 0.05 g ml -1 , GE Healthcare, Buckinghamshire, UK), 0.175 ml formamide and a variable amount of gradient buffer (100mM KCl, 100 mM Tris, 1 mM EDTA, pH 8). 0.8 – 1.2 ml of gradient buffer were added until the gradient solution had a buoyant density of 1.790 ± 0.005 g ml -1 . Density of the gradient solution was determined at 25°C. 500 ng of RNA were mixed with 1 ml of gradient solution and transferred into a 4.9-ml-centrifugation tube (OptiSeal Polyallomer Tube, 13 x 48 mm; Beckmann, Palo Alto, USA). The tubes were filled with gradient solution, balanced until a maximal difference of 0.05 g was achieved and closed with a plug. Tubes were placed in an ultra-vertical VTi 65.2 Rotor Package supplied with aluminium spacers (Beckham Coulter, Fullerton, USA). Screws were closed with a torque wrench adapter (200 inch-pounds, Beckham Coulter, Fullerton, USA) to a bolting torque of 60 inchpounds. [ 12 C]- and [ 13 C]-RNA were separated by isopycnic centrifugation at 130,000 x g (37,800 rpm) at 20°C for 67 hours in a LE-70 ultra centrifuge (Beckman Coulter, Fullerton, USA). The rotor ran out non-braked. All gradients were set up with the same gradient solution to minimize potential differences that might otherwise occur by the use of gradient solutions with similar but non identical densities. Isopycnic gradients prepared from soil treatments (2.3.1) containing the same saccharide (including both the oxic and anoxic treatments) were centrifuged simultaneously. In total, 45 gradients (15 for cellulose-, 15 for cellobiose-, and 15 for glucosesupplemented treatments) were processed in three centrifugation runs. A tube containing gradient solution without RNA (= Blank) was entrained in every run to measure the density of the fractions of the gradients (2.5.6.3). Relative Fluorescence 88 bp 90 bp 105 bp Fragment Length 126 bp Extraction of total RNA [ 13 C]-RNA [ 12 C]-RNA Incubation with [ 13 C]-Substrate Density Gradient Centrifugation tRFLP Analyses Reverse Transcription of RNA into cDNA Identification of [ 13 C]-assimilating Microorganisms Collection of [ 13 C]- RNA and [ 12 C]-RNA by Fractionation Incorporation of [ 13 C] in Nucleic Acids of Active Microorganisms Molecular Analysis of [ 13 C]-RNA and [ 12 C]-RNA 16S rRNA cDNA Gene Libraries Relative Fluorescence 88 bp 90 bp 105 bp Fragment Length 126 bp Relative Fluorescence 88 bp 90 bp 105 bp Fragment Length 126 bp Extraction of total RNA Extraction of total RNA [ 13 C]-RNA [ 12 C]-RNA [ 13 C]-RNA [ 12 C]-RNA [ 13 C]-RNA [ 12 C]-RNA Incubation with [ 13 C]-Substrate Density Gradient Centrifugation tRFLP Analyses Reverse Transcription of RNA into cDNA Identification of [ 13 C]-assimilating Microorganisms Collection of [ 13 C]- RNA and [ 12 C]-RNA by Fractionation Incorporation of [ 13 C] in Nucleic Acids of Active Microorganisms Molecular Analysis of [ 13 C]-RNA and [ 12 C]-RNA 16S rRNA cDNA Gene Libraries
M ATERIAL A ND M ETHODS 38 2.5.6.2. Fractionation of Gradients Fractionation of the gradients was performed manually by a suction pump (Econo Pump1, BioRad, Hercules, USA) (Manefield et al. 2002a). The centrifugation tubes were vertically fixed in a rag. A silicon hose (inner diameter: 1.6 mm) that was linked to the suction pump was equipped with a sterile needle (23G x 1’’) and inserted carefully into the tube underneath the plug. A second needle was used to drill a hole into the bottom of the tubes. Out flowing gradient solution was collected after the suction pump was started (flow rate: 0.45 ml min -1 ) and sterile water (coloured with brilliant blue) was pumped onto the gradient solution (Figure 7). Eleven fractions (~400 µl each) were collected in 60-second-intervals from every gradient. The eleventh fraction was contaminated with coloured water and therefore discarded from the analyses. Figure 7. Fractionation of gradients. See text for details. 2.5.6.3. Measurement of the Density of the Fractions The buoyant density of fractions 1 – 10 was determined in the blank gradient (i.e., without RNA) of each centrifugation run (Manefield et al. 2002b). The fractions were tempered to 25°C in a water bath and extensively mixed. The weight of the tubes was measured on an analytical balance (Analytic AC 120 S, Sartorius, Garching, Germany) before and after 100 µl of the fraction were removed. The density of every fraction was determined 10 times. Two fractions that presumably represented labeled ‘heavy’ RNA (corresponding to fraction numbers three and four, buoyant density 1.813 ± 0.001 – 1.821 ± 0.005 g ml -1 ) were pooled. Furthermore, two fractions that represented unlabeled ‘light’ RNA (corresponding to fraction numbers eight and nine, buoyant density 1.767 ± 0.000 – 1.776 ± 0.000 g ml -1 ) (Manefield et al. 2002b) were pooled. Pooled fractions were used for further analyses.
M ATERIAL A ND M ETHODS 39 2.5.6.4. RNA Precipitation RNA in ‘light’ and ‘heavy’ fractions was precipitated according to Degelmann et al. 2009a. 200 µl of each fraction were mixed with 130 µl NaCl (3 M, pH 5.2), 13.6 µl glycogen (10 mg ml -1 ), and 1,000 µl ice cold ethanol (96%). Samples were incubated for at least 12 hours at -20°C. RNA was precipitated by centrifugation (13,000 x g, 20 minutes, 4°C). The supernatant was discarded, the RNA pellet washed with 500 µl ice cold ethanol (70% in RNAse-free ddH 2 O, -20°C), and dried at room temperature. The RNA was dissolved in 22 µl RNAse-free ddH 2 O and 2 µl of RNA solution were used for quantification by fluorescence based methods (2.5.4.2). RNA samples were directly used for reverse transcription of RNA into cDNA (2.5.7). 2.5.7. Reverse Transcription of RNA into cDNA Reverse transcriptase of RNA into single-stranded complementary DNA (cDNA) was performed with the SuperScript VILO cDNA Synthesis Kit or the SuperScript III First-Synthesis System (both: Invitrogen, Darmstadt, Germany). Random hexamers (50 ng µl -1 ) were used in all reactions. Reverse transcription PCR (RT-PCR) was conducted on a TGradient Thermocycler (Biometra, Göttingen, Germany). Samples were prepared according to the manufacturer’s protocols. Annealing of random hexamers to RNA was performed for 10 minutes at 25°C, followed by 90 minutes at 42 – 50°C for cDNA synthesis. Reactions were terminated by heat-inactivation of the enzyme for 5 minutes at 85°C. A negative control (RNase-free ddH 2 O) and a positive control (control RNA supplied by the manufacturer) were entrained in every reverese transcription. CDNA was stored at -80°C. 2.5.8. In Vitro Transcription of DNA into RNA Single-stranded RNA was prepared by in vitro transcription of pGEM-T vector DNA with the T7 transcription kit (Fermentas, St.Leon-Rot, Germany). Singlestranded RNA was used for correction of transcript numbers measured in qPCR assays (2.5.10). Vector DNA was amplified by M13-PCR (2.5.9.3). The PCR product was purified with Millipore PCR 96 Cleanup Plates (Millipore Cooperation, Bedford, USA) (2.5.3.3) and quantified with the Quant-iT-PicoGreen dsDNA reagent Kit (2.5.4.2). A maximum of 1 µg of DNA was used for in vitro transcription according to the manufacturer’s description. A control DNA included in the kit was entrained in the analysis. In vitro transcription was performed by 2 hours of incubation at 37°C and reaction was stopped by freezing samples at -20°C. Successful transcription was checked by agarose gel electrophoresis (2.5.5). Template DNA was removed by DNAse I digestion (2.5.2.2). Remaining RNA was quantified with the RiboGreen Quantitation Reagent Kit (2.5.4.2). 2.5.9. Polymerase Chain Reaction (PCR) Polymerase chain reaction (PCR) is a method to multiply fragments of DNA or cDNA by several orders of magnitude, i.e., it generates multiple copies of a particular gene sequence. The process of enzymatic amplification of these gene fragments can be divided into three steps: denaturation of the DNA template, annealing of genespecific primers to single stranded DNA, and elongation of the complementary strand by a thermally stable DNA-polymerase (Saiki et al. 1985; Saiki et al. 1988).
M ATERIAL A ND M ETHODS 46 Name of Primer Position a Sequence (5´-3´) b T m (°C) c Target genes of Reference Sph1R 670 – 685 CTG TCA ATT CCG CCT 42 ‘Sphingo1-4’ e this work Deh1R 580 – 600 CGA CTT GAA CGA CCG CCT 53 ‘Deha1’ e this work DNA Inhibition Correction Assay T7PromF n.a. TAA TAC GAC TAT AGG G 38 pGEM-T vector insert Degelmann et al. 2009b M13RevR n.a. CAG GAA ACA GCT ATG ACC 48 pGEM-T vector insert Degelmann et al. 2009b RNA Inhibition Correction Assay IhcF n.a. ATT GGG CCC GAC GTC 47 pGEM-T vector insert Wieczorek et al. 2011 IhcR n.a. ATT TAG GTG ACA CTA TAG AAT A 46 pGEM-T vector insert Wieczorek et al. 2011 a Numbers in the names of primers indicate the 5’-position of binding sites relative to the reference sequence of Escherichia coli [ACCX80725]. Abbreviations: F, forward primer; R, reverse primer. b Y, C or T (IUPAC, International Union of Pure and Applied Chemistry). c T m (basic melting temperature) of primers was calculated online with OligoCalc: http://www.basic.northwestern.edu/biotools/oligocalc.htm]; last visit: 15.07.2010; Kibbe 2007 d Bold and underlined bases in primer sequence of family-level taxa indicate mismatches to non-target control (3.4). e Names and affiliation of family-level taxa according to (3.2.4.2). n.a., Non-applicable. Primers used in qPCR reactions (Table 15) were purchased from Biomers GmbH (Biomers.net, Ulm, Germany). PCR reactions were prepared in 20 µl volumes with PCR chemicals of Bioline (Luckenwalde, Germany) (Table 16) and performed on an iQ5 iCycler (Bio-Rad, Hercules, USA).
M ATERIAL A ND M ETHODS 47 Table 16. Chemical composition of qPCR reactions. Reagents (Conc.) Domain-level assays Family-level assays Inhibition correction assays Volume [µl] Final conc. Volume [µl] Final conc. Volume [µl] Final conc. SensiMix SYBR & Fluorescein (2x) a 10 1 x 10 1 x 10 1 x Forward primer b, c (10 µM) 1.5 0.75 µM 1 – 1.5 0.5 – 0.75 µM 1 0.5 µM Reverse primer b, c (10 µM) 1.5 0.75 µM 1.5 – 3 0.75 – 1.5 µM 1 0.5 µM Template DNA or cDNA 5.0 - 5.0 - 5.0 - PCR-H 2 O Ad 20 - Ad 20 - Ad 20 - a SensiMix SYBR & Fluorescein (2 x; Bioline, Luckenwalde, Germany) contains reaction buffer, heat-activated DNA Polymerase, ultra-pure dNTPs, 6 mM MgCl 2 , stabilizers, Fluorescein, and SYBR® Green I. b Details about primers in Table 15. c The final concentration of each primer in family-level assays number 1, 3, 5 and 8 was 750 nM (Table 25; 3.4). The final concentration of each primer in family-level assays number 2, 4, 6 and 7 was 500 nM (forward primer) and 1500 nM (reverse primer) (Table 25; 3.4). conc., Concentration Thermal protocols are summarized in Table 17. Details about family-level assays are given in the results section (3.4). The fluorescence signal was recorded during elongation. Subsequent agarose gel electrophoresis (2.5.5) was used to check for the formation of unspecific bands. Specificity of domain- and family-level assays was analyzed by sequencing of qPCR amplicons (2.5.10.2).
M ATERIAL A ND M ETHODS 48 Table 17. Thermal protocols for amplification of 16S rRNA genes and clone insert sequences. Primers pairs a Familylevel assays b Gm5F / Eub534R Arc344F / Arch915R IhcF / IhcR T7PromF / M13Rev No. Step T (°C) / time (minutes:seconds) 1 Initial denaturation 94 / 8:00 94 / 8:00 94 / 8:00 94 / 8:00 94 / 8:00 Step 2 – 4 35 x 35 x 45 x 38 x 38 x 2 Denaturation 95 / 0:30 95 / 0:30 95 / 0:30 95 / 0:30- 95 / 0:30- 3 Annealing 56 – 67.5 b / 0:40 55.7 / 0:40 55.7 / 0:40 60 / 0:15 61.2 / 0:15 4 Elongation 72 / 0:40 72 / 0:30 72 / 0:30 72 / 0:45 72 / 0:45 5 Final Elongation 72 / 5:00 72 / 5:00 72 / 5:00 72 / 7:00 72 / 10:00 161 x 49 x 6 Melting curve 70 – 94 / 0:10 (+0.15°C / cycle) 70 – 94 / 0:10 (+0.5°C / cycle) Length of amplicon (bp) See Table 25 190 570 450 450 a Details about primers in Table 15. b Details about primer pairs, appropriate annealing temperature, and length of amplicon of family-level assay in the results section (3.4). bp, Base pairs. No., Number. 2.5.10.2. Preparation of Quantitative DNA Standards Each qPCR was calibrated with quantitative assay-specific DNA standards. These standards were prepared with family-level 16S rRNA gene sequences reamplified (primer pair: 27F/907RM, Table 11) from the corresponding clone insert sequence (2.5.9.3, 3.4). PCR products were purified with Millipore PCR 96 Cleanup Plates (Millipore Cooperation, Bedford, USA) (2.5.3.3) and quantified with the QuantiT-PicoGreen dsDNA reagent Kit (Invitrogen, Karlsruhe, Germany) (2.5.4.2). The number of target sequences c TS was calculated with Equation 7, including the concentration of purified standard DNA, the length (i.e., the number of bases) of the amplicon, the molecular weight of one base pair in double stranded DNA, and the Avogadro constant N A .
M ATERIAL A ND M ETHODS 49 Equation 7. Number of target molecules in purified standard solutions. A bpbases DNAstd TS N MWN c c• • = c TS , number of target sequences (targets µl -1 ); c DNAStd , concentration of standard DNA (ng DNA µl -1 ); N bases , length of the amplicon; MW bp , molecular weight of one base pair in doublestranded DNA (MW bp = 650 ng nmol -1 ); N A , Avogadro constant (N A = 6.23 × 10 23 molecules nmol -1 ). DNA standards were diluted with DEPC-H 2 O to 2 × 10 9 target molecules µl -1 and stored at -80°C. Directly before use, dilution series were prepared to achieve concentrations of 10 7 – 10 0 molecules per qPCR reaction by application of 5 µl of diluted standard DNA as template (2.5.10.1). DNA standards for quantification of Bacteria and Archaea were supplied by Katharina Palmer, Department of Ecological Microbiology, University of Bayreuth. 2.5.10.3. Evaluation of the Specificity of Assays in Soil Samples Several factors were taken into account to evaluate target group specificity of family-level and domain-level qPCR assays. Melt curves of PCR products of qPCR standards were compared with PCR products amplified from soil samples. Subsequently, the formation of unspecific bands was checked by agarose gel electrophoreses (2.5.5). QPCR products from each assay of every sample (i.e., all replicates, time points, and substrates) of treatments in self-constructed incubation chambers (2.3.2) were pooled and sent to LGC Genomics (Berlin, Germany) for cloning and sequencing (2.5.13). 17 – 24 sequences per assay were analysed with BLAST (Altschul et al. 1990; 2.5.14). 2.5.10.4. Correction of PCR Inhibition and Calculation of Transcript Numbers Co-extracted compounds (e.g. humic acids) or variable reverse transcription efficiencies may impact on qPCR measurements. This might lead to underestimation of transcript numbers (Degelmann et al. 2009b; Wieczorek et al. 2011). Thus, transcript numbers were corrected by measuring inhibition factors in all samples (Degelmann et al. 2009b; Wieczorek et al. 2011). A pheA (gene that encodes a phenol hydroxylase) containing pGEM-T vector insert was used as inhibition control DNA and as template for preparation of control RNA (2.5.8). Every environmental DNA and RNA sample was spiked with 2 × 10 9 inhibition control molecules µl -1 . Spiked RNA extracts were transcribed into cDNA (2.5.7), and all extracts were quantified with domain-, family-level, and DNA or RNA inhibition assays (Table 15 and Table 17). Corrected 16S rRNA gene transcript numbers (lg[SQ corrected ]) of Bacteria, Archaea and family-level taxa were calculated with Equation 8.
M ATERIAL A ND M ETHODS 50 Equation 8. Correction of transcript numbers by inhibition correction assays. ••= INHIBCORR Assay isINHIBCORR setINHIBCORR measuredcorrected E E SQ SQ SQSQ )lg( )lg( lg)lg( _ _ lg(SQ corrected ), corrected logarithmic values of the starting quantity (SQ) measured in domain- or family-level assays; SQ measured , starting quantity measured in domain- or family-level assays; lg(SQ INHIBCORR_set ), logarithmic values of the starting quantity of control DNA or RNA spiked into environmental extracts; lg(SQ INHIBCORR_is ), logarithmic values of the starting quantity of control DNA or RNA measured; E Assay , amplification efficiency (E) of domain- or family-level assay; E INHIBCORR , amplification efficiency of the Inhibition correction assays. Family-level assays (3.4) did not yield sufficient gene amplification in DNA samples (i.e., target sequences were below the detection limit of about 10 genes per reaction; C T > 30). Presented transcript numbers in this study refer exclusively to transcript numbers quantified in cDNA samples. The total amount of RNA that was extracted from experimental replicates was similar, but variable (Figure 9). Thus, corrected 16S rRNA transcript numbers were quantified per ng RNA rather than per gram dry weight of soil. These absolute numbers of transcripts of family-level taxa were divided by the corresponding number of transcripts of total Bacteria to further correct for variability in lysis efficiencies and the growth of Bacteria. For each replicate, these copy numbers of family-level taxa were normalized based on the highest copy number measured (i.e., the highest number obtained for each assay, including substrate-supplemented and unsupplemented microcosms, was set at 100%) and are given in percent. Data presented are means of experimental triplicates (including standard deviations) or duplicates (without standard deviation). Figure 9. Total RNA extracted from each sampling point of each treatment. Filled bars, CMC-supplemented treatments; Shaded bars, cellobiose-supplemented treatments; Empty bars, unsupplemented controls. Error bars indicate standard deviations (n = 3). Time [h] 0 48 102 168 192 ng RNA g DW-1 0 500 1000 1500 2000 2500
M ATERIAL A ND M ETHODS 51 2.5.11. Terminal Restriction Fragment Length Polymorphism (tRFLP) Analysis TRFLP is a PCR-based fingerprinting method that allows for the rapid characterization of microbial communities in various environments (Liu et al. 1997; Thies 2007). 16S rRNA or functional genes are amplified with fluorescence labeled primers (one or both), resulting in terminal-labeled PCR products (Figure 10). Digestion of these PCR products with a restriction endonuclease (type II) yields terminal-labeled restriction fragments (tRFs). These fragments are subsequently length separated on a polyacrylamide gel (PAGE) (Liu et al. 1997). Different microbial communities result in different tRF profiles in which every tRF is indicative for at least one community member (Marsh 2005). In silico digestion of ribosomal database sequences or clone insert sequences obtained from the same environmental sample allows the affiliation of single tRFs to microbial taxa. Figure 10. Scheme of tRFLP analysis. See text for details. In this study, 16S rRNA cDNA gene based tRFLP was performed to rapidly compare bacterial and archaeal soil communities involved in the consumption of supplemented carbon (i.e., cellulose, cellobiose, glucose; 2.5.6). TRF profiles were further used to identify sequences of [ 13 C]-labeled organisms in gene libraries. Although cDNA of ‘heavy’ fractions was used for cloning, unlabeled genotypes (i.e., 16S rRNA cDNA sequences) in ‘heavy’ fractions had high G+C contents (53 – 61%, data not shown) and may have thus been unfolded RNA molecules yielding buoyant densities similar to labeled genotypes (Lüders et al. 2004). Therefore, tRF profiles of Bacteria and Archaea were generated from cDNA samples derived from the ‘light’ and ‘heavy’ fraction of [ 12 C]- and [ 13 C]-stable isotope probing gradients (2.5.6.3, 2.5.7), and compared. This minimizes overestimation of the active part of the cDNA of ‘Heavy‘ or ‘Light‘ Fraction 16S rRNA Gene Amplification with Fluorescence Labeled Primers Digestion with Restriction Enzyme PAGE Comparison of tRF Profiles Relative Fluorescence 74 bp 90 bp Fragment Length [ 13 C] Relative Fluorescence 108 bp Fragment Length 123 bp [ 12 C] MspI, TaqI MspI, TaqI Terminal Labeled Amplicons Separation of tRFs In silico Identification of Labeled tRFs Sequencedependent tRFs cDNA of ‘Heavy‘ or ‘Light‘ Fraction 16S rRNA Gene Amplification with Fluorescence Labeled Primers Digestion with Restriction Enzyme PAGE Comparison of tRF Profiles Relative Fluorescence 74 bp 90 bp Fragment Length Relative Fluorescence 74 bp 90 bp Fragment Length [ 13 C] Relative Fluorescence 108 bp Fragment Length 123 bp Relative Fluorescence 108 bp Fragment Length 123 bp [ 12 C] MspI, TaqI MspI, TaqI Terminal Labeled Amplicons Separation of tRFs In silico Identification of Labeled tRFs Sequencedependent tRFs
M ATERIAL A ND M ETHODS 52 microbial community. TRF profiles of day 0 (corresponding to the autochthonous soil community) were generated exclusively for [ 13 C]-treatments, since the same homogenized soil samples was used in [ 12 C]-treatments. Primer pairs, PCR conditions, and thermal protocols as mentioned above (Table 11 - Table 13). 2.5.11.1. Mung Bean Endonuclease Digestion Prematured termination of the elongation step during PCR can lead to singlestranded parts of the amplicons. This, in consequence, might promote the formation of ‘pseudo-tRFs’, i.e., tRFs that result from false restriction digestion at non-terminal restriction sites (Egert and Friedrich 2003). Single-stranded extensions at the terminal ends were removed with mung bean endonuclease digestion (New England Biolabs, Ipswich, USA). 50 µl of purified PCR product were mixed with 5.5 µl reaction buffer NEB2 (10x) and 2 µl mung bean endonuclease (0.5 U µl -1 ). Incubation was performed at 30°C for 1 hour. Endonuclease digestion was stopped by purification of samples with Millipore PCR 96 Cleanup Plates (Millipore Cooperation, Bedford, USA) (2.5.3.3). 2.5.11.2. Restriction Digestion Subsequent restriction digestion was performed with MspI (5’ 3’restriction site: C^CGG; New England Biolabs, Ipswich, USA, bacterial 16S rRNA genes) or TaqI (5’ 3’restriction site: T^CGA; New England Biolabs, USA; archaeal 16S rRNA genes) (Degelmann et al. 2009b). 7 µl purified PCR product were mixed with 1 µl reaction buffer NEB2 (10x; MspI) or 1 µl NEB3 (10x, TaqI), 1 µl BSA (10x), and 1 µl restriction enzyme (2 U µl -1 ). Restriction digestion was performed at 37°C for 4 hours (MspI) or at 65°C for 3 hours ( TaqI). Reactions were stopped by heat inactivation of the enzyme for 5 minutes at 95°C (MspI) or 20 minutes at 80°C (TaqI). Remaining DNA was quantified with the Quant-iT-PicoGreen dsDNA reagent Kit (Invitrogen, Karlsruhe, Germany) (2.5.4.2) and diluted with PCR-H 2 O to a concentration of 0.5 ng µl -1 . 2.5.11.3. Denaturing Polyacrylamide Gelelectrophoresis (PAGE) Denaturing gel electrophoresis was performed on a NEN model 4300 DNA analyzer (Licor, Lincoln, USA) as described (Hamberger et al. 2008). Gel plates (Borofloat glass plates 25 mm x 25 mm x 5 mm, LICOR, Lincoln, USA) were thoroughly cleaned sequentially with ddH 2 O, ethanol (70%) and isopropanol (80%). For stabilization of the gel pockets 250 µl bind-silane solution (1:1 bind silane [plusOne; GE Healthcare, Piscataway, USA] and 10% CH 3 COOH) was applied to the plates at the area at which the comb has to be inserted. The plates were separated by spacers (0.2 mm tick), fixed with assembly rails and casting plate, and placed in the gel casting stand. The polyacrylamide gel was prepared by gently mixing 15 g urea (Roche Pharma, Reinach, Switzerland), 3.75 ml of 40% acrylamide-bis solution (37.5 : 1; 2.6% C; Bio-RAD, Hercules, USA), 6 ml 5 x TBE buffer (450 mM Tris, 450 mM H 3 BO 3 , 10 mM EDTA, pH 8 [Sambrook and Russell 2001]), and 9.25 ml ddH 2 O. The gel solution was filtered (∅ 0.2 µm) to remove excess undissolved salts. The polymerisation reaction was started by addition of 175 µl ammonium persulfate (APS, 440 mM) and 17 µl N,N,N,N-Tetramethylethylendiame (TEMED, ultra-pure; Invitrogen, Karlsruhe, Germany). The gel was poured and a rectangular tooth comb (48 lanes) was inserted. Polymerisation was carried out at 25°C for 45 minutes. The
M ATERIAL A ND M ETHODS 53 glass plates were placed into the DNA analyser, the buffer tanks were fixed, and filled with 1 x TBE buffer. The comb was removed and gel pockets were cleaned from precipitated urea by gently flushing with buffer. Before samples were loaded onto the gel a pre-run was performed at 1,200 V and 45°C for 25 minutes. Samples were prepared for electrophoresis by mixing 2 µl of sample with 2 µl Stop-Solution (Licor, Lincoln, USA), denaturation at 95°C for 3 minutes, and placing on ice. A 50 – 700 bp length standard (Size Standard IRDye 700; Licor Biosciences or microSTEP-24a (700); Microzone Ltd, Haywards Heath, UK) was processed analogous. 0.5 µl size standard or 0.8 µl sample were loaded onto the gel and electrophoresis was performed at 1,200 V and 45°C for 4 hours. 2.5.11.4. Analysis of tRFLP Profiles Scanned gel images were exported and analysed with GelQuest (Version 2.6.3.; Sequentix, Klein Raden, Germany). This program generates tRFLP profiles in which every peak represents one tRF. The peak height of a tRF is directly linked to the relative intensity of the fluorescence label. To compare relative numbers, the fluorescence value of a detected tRF was normalized against the fluorescence value of the respective tRF in the tRFLP profile with the lowest total fluorescence. TRFs with values below 5% were excluded. Means and standard deviations were calculated for t end measurements of [ 13 C]-supplemented treatments (2.3.1). The tRFLP profiles derived from ‘heavy’ RNA fractions of [ 13 C]- and [ 12 C]-treatments were compared for identifying tRFs of labeled genotypes. The tRFs were scored as labeled when they were present only in [ 13 C]-profiles or when the relative intensity in [ 13 C]-profiles was higher than in [ 12 C]-profiles at the same respective time interval. This procedure avoided an overestimation of labeled genotypes, as unlabeled genotypes may co-migrate towards the ‘heavy’ RNA fraction (Lüders et al. 2004; Manefield et al. 2002a; Rangel-Castro et al. 2005). TRFs were assigned to putative taxa by in silico analyses of clone insert sequences (2.5.12) and ribosomal databases using the software tool TRFCUT included in ARB (Ricke et al. 2005; 2.5.14.3). 2.5.12. Construction of Gene Libraries Gene libraries are constructed by insertion of foreign DNA (e.g. PCR products of certain genes) into a cloning vector (ligation) and further introduction of this vector (transformation) into competent host cells (Sambrook and Russell 2001; Yansich- Perron et al. 1985). Cultivation of host cells yields clones (i.e., colonies) of which the ones are picked that apparently carry cloning vectors with inserted DNA (blue/white screening). 16S rRNA cDNA genes of each Bacteria, Archaea, and family-level taxa amplified by qPCR (2.5.10, 2.3.2) were pooled (i.e., all replicates, time points, and substrates). Cloning of bacterial 16S rRNA genes amplified from cDNA of the ‘heavy’ fraction of [ 13 C]-gradients (2.3.1, 2.5.9.2) was performed with the pGEM-T vector system II (Promega, Madison, USA) after the manufacturer’s protocol. 2.5.12.1. Ligation The pGEM-T vector (size: ~3000 bp) is linearized with a single 3’-terminal thymidine at both ends. This prevents recircularization of the vector and improves the efficiency of ligation of the PCR product into the vector (due to a polymerasemediated adenosine-overhang at the 3’ end of the DNA strand; Mezei and Storts
M ATERIAL A ND M ETHODS 54 1994; Robles and Doers 1994). The vector contains a gene for ampicillin-resistance and a MCS (multiple cloning site) that is located in the lacZ operon. lacZ encodes for the enzyme β-galactosidase and its operon is disrupted by insertion of a foreign DNA at the MCS. Thus, lacZ cannot be expressed in host cells that took up a plasmid with DNA insert. The inability of such clones to produce β-galactosidase can be used for selection of insert positive clones (2.5.12.3). The optimal insert to vector molar ratio for the pGEM-T vector system is denoted to be 1:1, but successful ligation can also be achieved between ratios of 8:1 and 1:8. The appropriate amount of insert to include in the ligation reaction can be calculated for different rations (Equation 9). A ratio between 1:1 and 3:1 was used in this study, what corresponds to 15 – 45 ng of PCR product per reaction (Table 18). Equation 9. Insert : vector ratio. ratio molar vector : insert vector of size kb insert of size kb • • = vector insert ng ng Purified 16S rRNA cDNA gene PCR products (2.5.3.2) were inserted into the vector plasmid (pGEM-T) by a T4 DNA ligase. The ligation reactions were prepared (Table 15), mixed, placed into a water bath (25°C), and incubated overnight at 4°C. The optimal temperature for ligation was perambulated during the cool down of the water bath. Table 18. Ligation reaction mix. Reaction component (concentration) Amount (µl) Final concentration Rapid Ligation Buffer (2 x) a 5 1x pGEM-T vector (50 ng µl - 1 ) 1 5 ng µl - 1 T4 DNA Ligase (3 Weiss U µl - 1 ) 1 0.3 U µl - 1 PCR product (insert DNA) 0.5 - 2 15 – 45 ng PCR-H 2 O Ad 10 - a 2x Rapid Ligation Buffer contains 60 mM Tris-HCl (pH 7.8), 20 mM MgCl 2 , 20 mM DTT, 2 mM ATP, and 10% polyethylene glycol (MW8000, ACS Grade). 2.5.12.2. Transformation Vector plasmids were inserted into high-efficiency competent Escherichia coli JM 109 cells (> 10 8 cfu µg DNA-1 ). E. coli cells were purchased from Promega (Madison, USA) or prepared by Ralf Mertel at the Department for Ecological Microbiology (University of Bayreuth) according to the following procedure.
M ATERIAL A ND M ETHODS 55 Liquid E. coli cultures (in glycerol) were transferred onto LB-plates and incubated at 37°C for 12 hours. One colony was picked, transferred into 5 ml LB- medium (2.1) and incubated overnight. 1 ml overnight-culture was used as inoculum for 50 ml LB-medium and cultures were incubated at 37°C for about 5 hours until an optical density (OD 660 ) of 0.5 was reached. The cells were collected by centrifugation (4,000 rpm, 4°C, 5 minutes), resuspended in 15 ml s terile, ice-cold transformation buffer 1 (10 mM CaCl 2 , 50 mM MnCl 2 ⋅ 4 H 2 O, 30 mM potassium acetate [CH 3 CO 2 K], 100 mM RbCl, 15% glycerol, pH 5.8), and incubated on ice for 90 minutes. Subsequently, cells were collected by centrifugation and resuspended in 2 ml sterile, ice cold transformation buffer 2 (75 mM CaCl 2 , 10 mM MOPS, 10 mM RbCl, 15% glycerol, pH 8.0). Cells were stored in 220 µl portions at -80°C. Directly before use cells were thawed on ice. For transformation 220 µl competent cells were gently mixed with 2.5 µl ligation reaction and mixtures were incubated on ice for 30 minutes. Cells were placed in a preheated water bath (42°C) for 45-50 seconds (heat-shock) to force uptaking of the vector plasmid. The cells were immediately placed on ice for 2 minutes. 950 µl SOC- medium (2.1) were added and cells were incubated at 37°C for 90 minutes on a thermomixer (Eppendorf, Hamburg, Germany) under constant slow shaking (300 rpm). Subsequently, about 400 µl of transformed cells were spread over LB/Ampicillin/IPTG/X-Gal-plates (2.1) and incubated overnight at 37°C. The addition of ampicillin to agar plates exclusively allows growth of cells that acquired the ampicillin-resistance gene by uptake of the vector plasmid. 2.5.12.3. Blue/white Screening of Clones Blue/white screening (Sambrook and Russell 2001) was used to select for clones that do not only carry a vector plasmid (as selected by ampicillin resistance), but also a plasmid with a DNA insert at the MCS. IPTG induces the expression of the lac operon and hence the production of the enzyme β-galactosidase. This enzyme reacts with the colourless compound X-Gal (lactose analogue) and leads to the formation of the blue insoluble pigment 5,5’-dibromo-4,4’dichloro-indigo, causing a characteristic blue colour in the colonies. White colonies are sufficient for the production of β-galactosidase, indicating that the lacZ operon is disrupted by a DNA insert. White colonies were randomly picked with a sterile toothpick, transferred onto fresh LB/Ampicillin/IPTG/X-Gal-Plates (2.1), and incubated at 37°C overnight. Selection of positive clones that carry vector plasmids with putative 16S rRNA cDNA gene insert sequences was performed by amplification of the insert DNA by M13- PCR (2.5.9.3). 2.5.13. Sequencing Purified M13-PCR products of 16S rRNA cDNA gene insert sequences (2.5.9.3) were sent to Macrogen (Kumchun-ku Seoul, Korea) for sequencing. QPCR products of domain- and family-level assays (2.5.10.3) were sent to LGC Genomics (Berlin, Germany) for cloning and sequencing. M13uniF (Table 11) was used as sequencing primer.
R ESULTS 62 Figure 11. Degradation of cellulose (A), cellobiose (B), and glucose (C) in soil slurries. Values are the means of triplicates ([ 13 C]-cellobiose, [ 13 C]- glucose), duplicates ([ 13 C]-cellulose, [ 12 C]-cellobiose, [ 12 C]-glucose), or a single treatment ([ 12 C]-cellulose). Error bars indicate standard deviations. Concentrations of compounds in unsupplemented treatments were subtracted from concentrations of products in supplemented treatments; the difference is shown. Solid symbols, values from treatments supplemented with [ 13 C]-substrates; Open symbols, values from treatments supplemented with [ 12 C]-substrates. Symbols: ✖ cellobiose; glucose; ▼ acetate; butyrate; ✚ propionate; ● carbon dioxide; ■ molecular hydrogen; ▲ ferrous iron. Arrows indicate sampling for stable isotope probing analyses. 0 10 20 0 10 20 0 10 20 0 10 20 0 4 8 12 0 10 20 30 40 50 0 4 8 12 0 10 20 30 40 50 0 40 80 0 10 20 30 40 50 0 40 80 0 40 80 0.0 0.2 0.4 0.6 0.0 0.2 0.4 0.6 0 2 4 6 8 10 0 2 4 6 8 10 0 2 4 6 8 10 CO 2 , H 2 , Fe 2+ [mM] Fatty Acids [mM] PulsedSubstrate [mM] B Cellobiose C Glucose Pulsed Substrate [mM] Time [d] A Cellulose OXIC ANOXIC CO 2 , H 2 , Fe 2+ [mM] Fatty Acids [mM] CO 2 , H 2 , Fe 2+ [mM] Fatty Acids [mM] 0 10 20 0 10 20 0 10 20 0 10 20 0 4 8 12 0 10 20 30 40 50 0 4 8 12 0 10 20 30 40 50 0 40 80 0 10 20 30 40 50 0 40 80 0 40 80 0.0 0.2 0.4 0.6 0.0 0.2 0.4 0.6 0 2 4 6 8 10 0 2 4 6 8 10 0 2 4 6 8 10 CO 2 , H 2 , Fe 2+ [mM] Fatty Acids [mM] PulsedSubstrate [mM] B Cellobiose C Glucose Pulsed Substrate [mM] Time [d] A Cellulose OXIC ANOXIC CO 2 , H 2 , Fe 2+ [mM] Fatty Acids [mM] CO 2 , H 2 , Fe 2+ [mM] Fatty Acids [mM]
R ESULTS 63 3.2. Identification of Active Prokaryotes in [ 13 C]-Cellulose, -Cellobiose, and -Glucose Supplemented Treatments by RNA Stable Isotope Probing Stable isotope probing was used to identify prokaryotic soil organisms that directly incorporated [ 13 C]-carbon into their RNA by assimilating [ 13 C]-cellulose, -cellobiose, or -glucose under oxic and anoxic conditions (2.3.1, 3.1). Although the risk of cross-feeding by [ 13 C]-carbon dioxide was minimized (2.3.1) it cannot be excluded that some soil organisms incorporated [ 13 C]-carbon by assimilation of [ 13 C]-breakdown products (e.g., fatty acids) during prolonged incubation. Therefore, tRFLP analysis was not only performed at the end of incubation but also in between to detect time-dependent shifts in the active microbial community (Figure 11). 3.2.1. Distribution of RNA in Gradient Fractions and Selection of ’Light’ and ’Heavy’ Fractions for Molecular Analyses RNA extracted from [ 12 C]- and [ 13 C]-cellulose, -cellobiose, or -glucose treatments was subjected to isopycnic centrifugation in a density gradient to separate ‘heavy’ [ 13 C]-RNA of saccharide-assimilating, i.e., labeled, organisms from ‘light’ [ 12 C]-RNA of non-labeled organisms. A RNA-free gradient (blank) was used to determine the buoyant density of any fraction (Figure 12). The buoyant density of each fraction decreased linearly and gradients were reproducible and comparable between three independent centrifugation runs (Figure 12). Buoyant densities ranged from 1.84 ± 0.01 g ml -1 to 1.76 ± 0.01 g ml -1 . Since the gradients were highly reproducible (Figure 12) the distribution of RNA in gradient fractions was only determined exemplarily for one gradient (Figure 13).
R ESULTS 64 Figure 12. Buoyant density of gradient solution in fractions of blank gradients that lack RNA at 25°C. Closed symbols: centrifugation run 1; grey symbols: centrifugation run 2; open symbols: centrifugation run 3. Lines represent linear regressions (R 2 = 0.98). Error bars indicate standard deviations (n = 10). RNA was distributed over the whole gradient with highest concentrations in fractions number four and seven, indicating a partial separation of [ 13 C]- and [ 12 C]- RNA (Figure 12). Full separation of ‘heavy’ and ‘light’ RNA derived from pure cultures is possible (Lüders et al. 2004), but not likely for environmental samples. Based on literature (Lüders et al. 2004; Manefield et al. 2002a) two fractions that presumably represented labeled ‘heavy’ or unlabeled ‘light’ RNA were selected for further analyses. Fractions number three and four (buoyant densities of 1.813 ± 0.00 – 1.821 ± 0.01 g ml -1 ) were pooled and correspond to ‘heavy’ RNA. Fractions number eight and nine (buoyant densities of 1.767 ± 0.00 – 1.776 ± 0.00 g ml -1 ) were pooled for analyses of unlabeled ‘light’ RNA (Figure 13). RNA of ‘heavy’ and ‘light’ fractions of [ 12 C]- and [ 13 C]-treatments was transcribed into cDNA and used for comparative tRFLP analyses of bacterial and archaeal 16S rRNA cDNA sequences. Fraction 1 2 3 4 5 6 7 8 9 10 Buyoant Density [g ml-1] 1.74 1.76 1.78 1.80 1.82 1.84 1.86
R ESULTS 65 Figure 13. Distribution of RNA in gradient fractions of anoxic [ 13 C]-cellobiose treatments after 24 days of incubation . Red indicates fractions that were chosen for analyses of ‘heavy‘ RNA. Blue indicates fractions that were chosen for analyses of ‘light‘ RNA. Error bars indicat the standard deviation (n = 3). Buoyant density of ‘light’ fractions was 1.767 ± 0.00 – 1.776 ± 0.00 g ml -1 . Buoyant density of ‘heavy’ fractions was 1.813 ± 0.00 – 1.821 ± 0.01 g ml -1 . 3.2.2. TRFLP Profiles of Archaeal 16S rRNA cDNA Sequences Archaeal 16S rRNA cDNA genes were digested with TaqI, resulting in two major tRFs (142 bp, 154 bp) and some minor tRFs with relative fluorescence values below 5% (Figure 14). These tRFs were present in each sample of cellulose-, cellobiose-, or glucose-supplemented treatments (data not shown). Neither differences between [ 13 C]- and [ 12 C]-treatments nor between ‘light’ and ‘heavy’ fractions of [ 13 C]-treatments were detected (Figure 14). This general result suggested that mesophilic Archaea were of minor to no consequence to the primary consumption of supplemental substrates. Archaeal 16S rRNA cDNA genes were not used for cloning, but in silico analyses revealed that the tRFs might represent species of Methanocorpusculaceae, Methanococcaceae, and Methanobacteriaceae (tRF 142 bp). TRF 154 bp could not be assigned to known sequences. RNA [ng µl -1 ] 0.0 0.2 0.4 0.6 Fraction 1 2 3 4 5 6 7 8 9 10 3 4 8 9 RNA [ng µl -1 ] 0.0 0.2 0.4 0.6 Fraction 1 2 3 4 5 6 7 8 9 10 3 4 8 9 RNA [ng µl -1 ] 0.0 0.2 0.4 0.6 Fraction 1 2 3 4 5 6 7 8 9 10 RNA [ng µl -1 ] 0.0 0.2 0.4 0.6 Fraction 1 2 3 4 5 6 7 8 9 10 RNA [ng µl -1 ] 0.0 0.2 0.4 0.6 Fraction 1 2 3 4 5 6 7 8 9 10 33 44 88 99 RNA [ng µl -1 ] 0.0 0.2 0.4 0.6 Fraction 1 2 3 4 5 6 7 8 9 10 33 44 88 99
R ESULTS 66 Figure 14. TRFLP gel image of archaeal 16S rRNA cDNA genes . Restriction enzyme was TaqI. S, length standard 75 – 1,000 bp; LF, ‘light‘ fraction; SF, ‘heavy‘ fraction; Rep., Replicates; X, Lane was not analysed. 900 800 700 600 500 450 400 350 300 250 100 75 200 150 900 800 700 600 500 450 400 350 300 250 100 75 200 150 1000 13 LF 13 SF 13 LF 13 SF 12 SF 12 LF 12 SF 12 LF 202024 240204 24 3 Rep. 0 612 3 Rep. 0 204 24 3 Rep. 612 3 Rep. 0126 126 S S S Day XX 900 800 700 600 500 450 400 350 300 250 100 75 200 150 900 800 700 600 500 450 400 350 300 250 100 75 200 150 900 800 700 600 500 450 400 350 300 250 100 75 200 150 1000 900 800 700 600 500 450 400 350 300 250 100 75 200 150 1000 13 LF 13 SF 13 LF 13 SF 12 SF 12 LF 12 SF 12 LF 202024 240204 24 3 Rep. 0 612 3 Rep. 0 204 24 3 Rep. 612 3 Rep. 0126 126 S S S Day XX XX
R ESULTS 67 3.2.3. TRFLP Profiles of Bacterial 16S rRNA cDNA Sequences Primarily, tRF profiles analysed of the last time points were compared, i.e., for each treatment ‘light’ and ‘heavy’ fraction of oxic and anoxic treatments supplemented with either [ 13 C]-cellulose, [ 13 C]-cellobiose, [ 13 C]-glucose, [ 12 C]- cellulose, [ 12 C]-cellobiose, or [ 12 C]-glucose were compared (Figure 15 and Figure 16). Separation of [ 12 C]- and [ 13 C]-RNA during centrifugation should be reflected in different tRF patterns between ‘light’ and ‘heavy’ fractions. 3.2.3.1. Comparison of ‘Heavy’ and ‘Light’ Fractions in Oxic and Anoxic [ 12 C]- or [ 13 C]-Treatments Restriction digestion of bacterial 16S rRNA cDNA sequences derived from [ 13 C]- treatments resulted in different tRF pattern of the ‘light’ and ‘heavy’ fractions (Figure 15), confirming incorporation of [ 13 C]-carbon into RNA by parts of the microbial community and successful separation of [ 12 C]- and [ 13 C]-RNA in the gradient solution. Assuming that labeled taxa are presented by (i) every tRF that is exclusively present in the ‘heavy’ fraction but not in the ‘light’ fraction and (ii) every tRF that exhibits a higher relative fluorescence in the ‘heavy’ than in the ‘light’ fraction, 31 tRFs (in total) were putatively labeled in [ 13 C]-treatments at the end of incubation.
R ESULTS 68 Figure 15. TRFLP pattern of bacterial 16S rRNA cDNA sequences in ‘light‘ (fraction 8+9: 1.767-1.776 g ml -1 ) and ‘heavy‘ fractions (fraction 3+4: 1.181-1.183 g ml -1 ) of [ 13 C]-supplemented treatments at the latest analysed time point. (A) Cellulose treatment after 70 days. (B) Cellobiose treatment after 12 days (oxic) and 24 days (anoxic). (C) Glucose treatment after 12 days (oxic) and 24 days (anoxic). Values are the mean of triplicates ([ 13 C]-cellobiose, [ 13 C]-glucose) or duplicates ([ 13 C]-cellulose). Error bars indicate standard deviations. Bars that represent putatively labeled tRFs (i.e., tRFs exclusively occured in ’heavy‘ fractions or are higher in relative fluorescence than the corresponding tRFs in ‘light‘ fractions) are marked with an asterisk. B A ** * * ** * * * **** * * * * * ** T-RF Length [bp] C ******* * * * * 60 30 0 Oxic ‘light‘ Oxic ‘heavy‘ Anoxic ‘heavy‘ Anoxic ‘light‘ 60 30 60 30 60 30 100 50 0 Oxic ‘light‘ Oxic ‘heavy‘ Anoxic ‘heavy‘ Anoxic ‘light‘ Relative Fluorescence [%] 100 50 100 50 100 50 80 40 0 Oxic ‘light‘ Oxic ‘heavy‘ Anoxic ‘heavy‘ Anoxic ‘light‘ 80 40 80 40 80 40
R ESULTS 69 Comparisons of ‘light’ and ‘heavy’ fractions may lead to an overestimation of the active part of the microbial community, since tRFs in the ‘heavy’ fraction do not necessarily result from labeled RNA, but can result from co-migrated unlabeled RNA (Lüders et al. 2004). The presence of unlabeled tRFs in ‘heavy’ fractions becomes apparent by comparison of ‘heavy’ and ‘light’ fractions of [ 12 C]-treatments (Figure 16) that never received [ 13 C]-carbon. In total, eight tRFs were exclusively present in ‘heavy’ fractions of [ 12 C]-treatments, although a [ 13 C]-labeling was impossible. These unlabeled genotypes in ‘heavy’ fractions could be related to sequences that had high G+C contents (53 – 61%, 3.2.4.2.2). They may have thus been unfolded RNA molecules yielding buoyant densities similar to labeled genotypes (Lüders et al. 2004). To avoid this bias identification of labeled tRFs was based on the comparison of ‘heavy’ fractions of [ 12 C]-treatments with ‘heavy’ fractions of [ 13 C]-treatments (3.2.3.2).
R ESULTS 70 Figure 16. TRFLP pattern of bacterial 16S rRNA cDNA sequences in ‘light‘ (fraction 8+9: 1.767-1.776 g ml -1 ) and ‘heavy‘ fractions (fraction 3+4: 1.181-1.183 g ml -1 ) of [ 12 C]-supplemented treatments at the latest analysed time point. (A) Cellulose treatment after 70 days. (B) Cellobiose treatment after 12 days (oxic) and 24 days (anoxic). (C) Glucose treatment after 12 days (oxic) and 24 days (anoxic). Fluorescence values were normalized against the lowest single fluorescence; values below 5% were excluded from the analysis. Bars that represent tRFs that exclusively occur in ’heavy‘ fractions are marked with an asterisk. ** * * *** * 60 30 0 60 30 60 30 60 30 80 40 0 Relative Fluorescence [%] 80 40 80 40 80 40 80 540 0 80 40 80 40 80 40 T-RF Length [bp] 69 91 143 153 163 172 205 279 288 295 325 433 438 447 456 469 474 480 486 490 494 497 500 507 511 520 523 526 540 547 552 603 772 891 903 920 80 134 149 156 164 201 208 283 292 298 327 436 443 452 466 473 476 483 488 492 496 498 503 508 514 522 525 534 542 549 559 690 870 899 912 B A C Oxic ‘light‘ Oxic ‘heavy‘ Anoxic ‘heavy‘ Anoxic ‘light‘ Oxic ‘light‘ Oxic ‘heavy‘ Anoxic ‘heavy‘ Anoxic ‘light‘ Oxic ‘light‘ Oxic ‘heavy‘ Anoxic ‘heavy‘ Anoxic ‘light‘ ** * * *** * 60 30 0 60 30 60 30 60 30 80 40 0 Relative Fluorescence [%] 80 40 80 40 80 40 80 540 0 80 40 80 40 80 40 T-RF Length [bp] 69 91 143 153 163 172 205 279 288 295 325 433 438 447 456 469 474 480 486 490 494 497 500 507 511 520 523 526 540 547 552 603 772 891 903 920 80 134 149 156 164 201 208 283 292 298 327 436 443 452 466 473 476 483 488 492 496 498 503 508 514 522 525 534 542 549 559 690 870 899 912 80 134 149 156 164 201 208 283 292 298 327 436 443 452 466 473 476 483 488 492 496 498 503 508 514 522 525 534 542 549 559 690 870 899 912 B A C Oxic ‘light‘ Oxic ‘heavy‘ Anoxic ‘heavy‘ Anoxic ‘light‘ Oxic ‘light‘ Oxic ‘heavy‘ Anoxic ‘heavy‘ Anoxic ‘light‘ Oxic ‘light‘ Oxic ‘heavy‘ Anoxic ‘heavy‘ Anoxic ‘light‘
R ESULTS 71 3.2.3.2. Identification of Labeled tRFs by Comparison of ‘Heavy’ Fractions between [ 12 C]- and [ 13 C]-Treatments and Phylogenetic Affiliation A conservative approach was performed to identify labeled tRFs (3.2.3.1). TRFs were scored as labeled when they were only present in [ 13 C]-profiles (e.g., tRF 507 bp, Figure 17) or when the relative intensity in [ 13 C]-profiles was significantly higher than in [ 12 C]-profiles at the same respective time interval (e.g., tRF 490 bp, Figure 17). Labeled tRFs were assigned to family-level taxa based on gene library analyses (Table 19 – Table 22). 3.2.3.2.1. Bacteria that Incorporated [ 13 C]-Carbon in Oxic Treatments [ 13 C]-cellulose supplementation resulted in twelve labeled tRFs of which tRFs 438, 466, 490, 507, 525, and 540 bp were observed at days 35 and 70, whereas tRFs 80, 443, 456, 473, 497, and 503 bp were only detected at day 70 (Figure 17). No tRF was dominant or contributed to more than 30% to the total fluorescence what correlated with the frequency of corresponding labeled genotypes obtained from gene libraries from oxic cellulose-supplemented treatments (Table 22, 3.2.4.2.1). Ten tRFs were identified as known and new family-level taxa within the phyla Actinobacteria, Alphaproteobacteria, Bacteroidetes, Betaproteobacteria, Chloroflexi, Deltaproteobacteria, Firmicutes, Gammaproteobacteria, Planctomycetes, and two deep-branching groups within Bacteria (‘Bac2’) and Proteobacteria (‘Prot1’) (Table 19). Two tRFs (503 bp, 540 bp) could not be identified.
R ESULTS 78 [ 13 C]-cellobiose supplementation resulted in three labeled tRFs of which tRF 523 bp was present at day 4, 20, and 24, and exhibited the highest relative fluorescence (Figure 21). TRF 494 bp was only detected at day 4, tRF 903 at day 4 and 20, but disappeared at day 24 (Figure 21). The most dominant tRF 523 bp affiliated with the phylum Firmicutes (Table 20 and Table 22) what corresponds to the prevalence of this phylum in gene libraries (Table 22). TRF 494 bp represented members of the phyla Gammaproteobacteria and Deltaproteobacteria, taxa of minor importance in both relative fluorescence and frequencies in gene libraries (Table 20 and Table 22). TRF 903 bp could not be identified (Table 20). Figure 21. TRFLP pattern of bacterial 16S rRNA cDNA sequences in ‘heavy’ fractions (fraction 3+4: 1.181-1.183 g ml -1 ) of cellobiose-supplemented anoxic treatments . (A) [ 13 C]-cellobiose treatment after 0, 4, 10, and 24 days. (B) [ 12 C]-cellobiose treatment after 20 and 24 days. Time points were chosen concerning changes in processes during substrate utilization (Figure 11). The [ 12 C]-gradient for day 0 and 4 was not generated. Fluorescence values were normalized against the lowest single fluorescence; values below 5% were excluded from the analysis. Error bars indicate standard deviations. [ 13 C]-labeled tRFs are indicated by their length and marked by an arrow (494 bp; 523 bp; 903 bp). A [ 13 C]-Cellobiose (Anoxic) Relative Fluorescence [%] 0 days 20 days 24 days B [ 12 C]-Cellobiose (Anoxic) 4 days 20 days Relative Fluorescence [%] tRF Length [bp] 70 160 480 920 24 days tRF Length [bp] 494bp ➤ 70 160 480 920 523bp ➤ 903bp ➤ 75 50 25 523bp ➤ 903bp ➤ 523bp ➤ 523bp ➤ 523bp ➤ 903bp ➤ 75 50 25 75 50 25 75 50 25 75 50 25 75 50 25 A [ 13 C]-Cellobiose (Anoxic) Relative Fluorescence [%] 0 days 20 days 24 days B [ 12 C]-Cellobiose (Anoxic) 4 days 20 days Relative Fluorescence [%] tRF Length [bp] 70 160 480 920 24 days tRF Length [bp] 494bp ➤ 70 160 480 920 523bp ➤ 903bp ➤ 75 50 25 75 50 25 523bp ➤ 903bp ➤ 523bp ➤ 523bp ➤ 523bp ➤ 903bp ➤ 75 50 25 75 50 25 75 50 25 75 50 25 75 50 25 75 50 25 75 50 25 75 50 25 75 50 25 75 50 25
R ESULTS 79 [ 13 C]-glucose supplementation resulted in three labeled tRFs (Figure 22) of which only tRF 520 bp was also present in [ 12 C]-treatments. TRF 508 and 520 bp were only labeled at day 10 and 24, whereas tRF 496 was labeled at all time points (Figure 22). TRF 520 bp was the dominant labeled tRF and affiliated with the phylum Firmicutes which also represented the most frequent taxon in the corresponding gene library (Table 20 and Table 22). TRF 496 bp was assigned to the phylum Gammaproteobacteria (Table 20) which was a minor taxon in gene libraries (Table 22). TRF 508 bp could not be identified, although it exhibited a relative fluorescence of about 10%. Figure 22. TRFLP pattern of bacterial 16S rRNA cDNA sequences in ‘heavy’ fractions (fraction 3+4: 1.181-1.183 g ml -1 ) of glucose-supplemented anoxic treatments . (A) [ 13 C]-glucose treatment after 0, 4, 10 and 24 days. (B) [ 12 C]-glucose treatment after 0, 4, and 24 days. Time points were chosen concerning changes in processes during substrate utilization (Figure 11). The [ 12 C]-gradient for day 10 was not generated. Fluorescence values were normalized against the lowest single fluorescence; values below 5 % were excluded from the analysis. Error bars indicate standard deviations. [ 13 C]-Labeled tRFs are indicated by their length and marked by an arrow (496 bp; 508 bp; 520 bp). A [ 13 C]-Glucose (anoxic) Relative Fluorescence [%] 0 days 0 days 4 days 24 days B [ 12 C]-Glucose (anoxic) 4 days 10 days 24 days Relative Fluorescence [%] tRF Length [bp] 70 290 480 540 70 290 480 540 tRF Length [bp] 496bp 508bp 520bp ➤ ➤ ➤ 520bp ➤ 520bp ➤ 520bp ➤ 520bp ➤ 496bp ➤ 496bp ➤ 508bp ➤ 60 40 20 60 40 20 60 40 20 60 40 20 60 40 20 60 40 20 60 40 20 A [ 13 C]-Glucose (anoxic) Relative Fluorescence [%] 0 days 0 days 4 days 24 days B [ 12 C]-Glucose (anoxic) 4 days 10 days 24 days Relative Fluorescence [%] tRF Length [bp] 70 290 480 540 70 290 480 540 tRF Length [bp] 496bp 508bp 520bp ➤ ➤ ➤ 520bp ➤ 520bp ➤ 520bp ➤ 520bp ➤ 496bp ➤ 496bp ➤ 508bp ➤ 60 40 20 60 40 20 60 40 20 60 40 20 60 40 20 60 40 20 60 40 20 60 40 20 60 40 20 60 40 20 60 40 20 60 40 20 60 40 20 60 40 20
R ESULTS 80 TRFLP analyses and 16S rRNA cDNA gene libraries of anoxic [ 13 C]-treatments showed that the phylum Firmicutes was prevalent in all treatments, i.e., cellulose-, cellobiose-, and glucose-supplemented treatments (Table 22). Almost all labeled genotypes of the phylum Firmicutes were related to the family Clostridiaceae. Different sub-groups within this family were selectively activated in regard to the type of substrate. Sequences of Cluster III Clostridiaceae were closely related (> 97% sequence similarity) to the strict anaerobic cellulolytic species Clostridium cellulolyticum and C. thermocellum (Magnusson et al. 2009; Petitdemange et al. 1984; Zhang and Lynd 2005), whereas sequences derived from [ 13 C]-cellobiose and –glucose treatments were closely related (> 98% similarity) to the saccharolytic species C. butyricum and C. vincentii (Figure A2; Mountfort et al. 1997; Skerman et al. 1980). An unclassified cluster (‘Cellu1’) within the phylum Bacteroidetes and the family Kineosporiaceae (phylum Actinobacteria) were major active taxa in cellulosesupplemented treatments (Table 22, Figure A1 – Figure A3), but not in cellobiose- or glucose-supplemented treatments. The lack of these taxa in cellobiose and glucosesupplemented treatments (Table 22) indicates that labeling by cross-feeding on cellulose-derived cellodextrins was unlikely. Table 20. Identification and occurrence of labeled tRFs in anoxic [ 13 C]- treatments. Treatment tRF [bp] Identity [phylum: family] a Label b [day] Cellulose-supplemented Anoxic 143 Actinobacteria: Intrasporangiaceae, Kineosporiaceae, Streptomycetaceae Deltaproteobacteria: Geobacteraceae 70 149 Actinobacteria: Cellulomonadaceae Alphaproteobacteriaceae: Hyphomicrobiaceae 70 205 Bacteroidetes: ‘Cellu1’ c Bacteria: ‘Bac1’ c , ‘Bac3’ c Firmicutes: Clostridiaceae, Paenibacillaceae, ‘Clos1’ c 35, 70 490 Betaproteobacteria: Rhodocyclaceae Deltaproteobacteria: Pelobacteraceae 35 522 Firmicutes: Clostridiaceae, ‘Clos2’ c 35, 70 540 Bacteroidetes: ‘Cellu2’ c , ‘Cellu3’ c 35 899 n.i. 35
R ESULTS 81 Treatment tRF [bp] Identity [phylum: family] a Label b [day] Cellobiose - supplemented Anoxic 494 Gammaproteobacteria: Aeromonadaceae, ‘Gam1’ c Deltaproteobacteria: Pelobacteraceae 4 523 Firmicutes: Clostridiaceae, ‘Clos4’ c 4, 20, 24 903 n.i. 4, 20 Glucose - supplemented Anoxic 496 Gammaproteobacteria: Aeromonadaceae, Enterobacteriaceae 4, 10, 24 508 n.i. 10, 24 520 Firmicutes: Clostridiaceae 10, 24 n.i., Not identified. a Based on experimental gene libraries. b First labeling at noted day of incubations in ‘heavy’ RNA fractions. c New family-level taxon based on 16S rRNA gene similarities < 87% to next cultivated species. 3.2.4. Bacterial 16S rRNA cDNA Gene Libraries of ‘Heavy’ Fractions To identify tRFs related to labeled taxa, six 16S rRNA cDNA gene libraries were constructed, i.e., 16S rRNA cDNA genes derived from ‘heavy’ fractions of oxic and anoxic [ 13 C]-cellulose, -cellobiose, or -glucose treatments were used for cloning. 3.2.4.1. Rarefaction and Coverage In total, 828 sequences (> 350 bp length) were analysed. The number of clones per library ranged between 104 and 177 (Table 21). The coverage was above 90% in five of six libraries, indicating that the major diversity on family-level (> 87% sequence similarity) was covered by the analyses (Table 21). Some libraries exhibited a high number of sequences that could not be assigned to labeled tRFs. These sequences were defined to represent unlabeled bacterial taxa (Table 23), partly resulting in a low proportion of labeled clones in ‘heavy’ fractions (Table 21).
R ESULTS 82 Table 21. Characteristics of 16S rRNA cDNA gene libraries of ’heavy’ fractions of oxic and anoxic [ 13 C]-cellulose, [ 13 C]-cellobiose, and [ 13 C]-glucose treatments. Number of clones per library Number of family-level OTUs Labeled clones [%] Unlabeled clones [%] Coverage [%] Cellulose-supplemented Oxic 163 93 23.3 76.7 64.4 Anoxic 177 27 89.8 10.2 93.8 Cellobiose-supplemented Oxic 122 19 45.2 54.8 93.4 Anoxic 130 17 64.6 35.4 94.7 Glucose-supplemented Oxic 104 15 63.6 36.4 96.2 Anoxic 132 28 23.3 76.7 90.9 Rarefaction analyses supported the above mentioned findings (i.e., detection of the major diversity), as indicated by the out-plateauing rarefaction curves (Figure 23). The rarefaction curve of oxic cellulose-supplemented treatments was steep (Figure 23 A) what corresponded to the low coverage of 64% of this gene library (Table 21). Gene libraries of oxic and anoxic cellobiose- and glucose-supplemented treatments showed both a good coverage (Table 21), and rarefaction curves nearly outplateaued (Figure 23 B and C). In general, differences in the diversity within all treatments were observed. More family-level OTUs were detected in oxic cellulose- and oxic cellobiose-supplemented treatments than in the corresponding anoxic treatments, whereas oxic glucose-supplemented treatments showed less OTUs than the corresponding anoxic treatments (Table 21). The total number of detected OTUs was highest in cellulose-supplemented treatments.
R ESULTS 83 Figure 23. Rarefaction curves of 16S rRNA cDNA gene sequences of ‘heavy‘ fractions of oxic and anoxic [ 13 C]-treatments . (A) Cellulose-supplemented; (B) Cellobiose-supplemented; (C) Glucose-supplemented. , oxic treatments; , anoxic treatments. Solid lines indicate the 95% confidence intervals. 0 25 50 75 100 A Number of OTUs 0 5 10 15 20 B Number of Clones 0 25 50 75 100 125 150 175 0 10 20 30 C 0 25 50 75 100 A Number of OTUs 0 5 10 15 20 B Number of Clones 0 25 50 75 100 125 150 175 0 10 20 30 C
R ESULTS 84 3.2.4.2. Grouping of Sequences into Family-level OTUs and their Phylogenetic Affiliation 16S rRNA cDNA gene sequences were grouped into family-level OTUs if sequence similarity was higher than 87 % (Yarza et al. 2008). TRFLP analysis was used to differentiate between genotypes of labeled and non-labeled taxa detected in 16S rRNA cDNA gene libraries. 3.2.4.2.1. Labeled Taxa Fourty-eight family-level taxa within ten bacterial phyla were identified as organisms that assimilated carbon derived from [ 13 C]-cellulose, -cellobiose, or -glucose. Twenty-eight of these taxa did not closely affiliate with known families (Table 22, Figure A1 - Figure A3). Active taxa differed in regard to the substrate type and the availability of O 2 , indicating that (i) cellulose supplementation selected for other taxa than cellobiose and glucose supplementation did, and (ii) oxic conditions selected for other taxa than anoxic conditions did (Table 22). Taxa that showed high relative abundances in gene libraries of oxic glucose- and cellobiose-supplemented treatments were the families Intrasporangiaceae (5 – 17%) and Micrococcaceae (30%) of the phylum Actinobacteria (Table 22, Figure A1). In contrast, major active taxa in anoxic glucose- and cellobiose-supplemented treatments were closely related to saccharolytic Clostridiaceae (Cluster I Clostridiaceae) of the phylum Firmicutes (19 – 53%) (Table 22, Figure A2). Planctomycetaceae (Planctomycetes), a deepbranching new family-level taxon (‘Deha1’) of the phylum Chloroflexi, and four new taxa (‘Sphingo1-4’) of Sphingobacteriales (Bacteroidetes) were the most abundant groups in gene libraries of oxic cellulose-supplemented treatments and exhibited relative abundances of almost 4%, 5%, and 3%, respectively (Table 22, Figure A3). Kineosporiaceae (Actinobacteria), cellulolytic Clostridiaceae (Cluster III Clostridiaceae, phylum Firmicutes), and three new deep-branching family-level taxa (‘Cellu1-3’) of the phylum Bacteroidetes were dominant in gene libraries of anoxic cellulose-supplemented treatments and had relative abundances of about 16%, 28%, and 22%, respectively (Table 22, Figure A1 – Figure A3). These groups were analysed for their response to changing availabilities of O 2 and pesticides in further experiments (3.6; 3.5).
R ESULTS 85 Table 22. Relative abundances of labeled OTUs obtained from 16S rRNA cDNA gene libraries from ‘heavy’ fractions of [ 13 C]-treatments and their phylogenetic affiliation. Relative Abundances [%] Phyla and families OTUs a Cellulose Cellobiose Glucose oxic anoxic oxic anoxic oxic anoxic Actinobacteria Cellulomonadaceae 22, 23, 78 1.23 0.57 0.81 - - - Intrasporangiaceae 5, 60, 79 0.61 1.14 4.84 - 16.8 - Kineosporiaceae 2, 46 - 21.7 4.03 - 5.61 - Micrococcaceae 3, 18, 42, 54, 57, 58, 59, 61, 80 - - 29.8 - 29.9 - Mycobacteriaceae 56, 62 - - 2.24 - - - Nakamurellaceae 55 - - - - 0.93 - Nocardiaceae 44, 65 - - 1.61 - - - Nocardioidaceae 13, 47, 63 0.61 - - - 4.67 - Streptomycetaceae 11 - 2.29 - - - - ‘Micro1’ (Micrococcineae) b 34 0.61 - - - - - ‘Micro2’ (Micrococcineae) b 41 - - 0.81 - - - ‘Micro3’ (Micrococcineae) b 43 - - - - 0.93 - ‘Micro4’ (Micrococcineae) b 48 0.61 - - - - - ‘Actino1’ (Actinomycetales) b 40 0.61 - - - - - Alphaproteobacteria Hyphomicrobiaceae 72 - 0.57 - - - - ‘Rhizo1’ (Rhizobiales) b 75 0.61 - - - - - ‘Rhizo2’ (Rhizobiales) b 76 0.61 - - - - -
R ESULTS 86 Relative Abundances [%] Phyla and families OTUs a Cellulose Cellobiose Glucose oxic anoxic oxic anoxic oxic anoxic Bacteroidetes ‘Sphingo1’ (Sphingobacteriales) b 28 1.23 - - - - - ‘Sphingo2’ (Sphingobacteriales) b 36 0.61 - - - - - ‘Sphingo3’ (Sphingobacteriales) b 38 0.61 - - - - - ‘Sphingo4’ (Sphingobacteriales) b 50 0.61 - - - - - ‘Cellu1’ c 6 - 13.7 - - - - ‘Cellu2’ c 14 - 1.71 - - - - ‘Cellu3’ c 74 - 0.57 - - - - Betaproteobacteria Nitrosomonadaceae 67 - - - - 0.93 - Oxalobacteraceae 66 0.61 - - - - - Rhodocyclaceae 7 - 4.00 - - - 0.75 ‘Burk1’ (Burkholderiales) b 24 - - - - 0.93 - Chloroflexi ‘Deha1’ (Dehalococcoidetes) b 8 4.91 - - - - - Deltaproteobacteria Geobacteraceae 27, 35 - 0.57 - - 0.93 - Pelobacteraceae 12 - 1.14 - 0.77 - - ‘Desu1’ (Desulfuromonadales) b 19 1.23 - - - - - ‘Myxo1’ (Myxococcales) b 31 0.61 - - - - -
R ESULTS 87 Relative Abundances [%] Phyla and families OTUs a Cellulose Cellobiose Glucose oxic anoxic oxic anoxic oxic anoxic Firmicutes Clostridiaceae 1, 4, 16, 20, 21, 25, 51, 52, 53, 68 - 28.0 - 53.1 - 18.8 Paenibacillaceae 15 - 0.57 - - - - ‘Clos1’ (Clostridiales) b 29 - 1.14 - - - - ‘Clos2’ (Clostridiales) b 37 - 0.57 - - - - ‘Clos3’ (Clostridiales) b 49 0.61 - - - - - ‘Clos4’ (Clostridiales) b 82 - - - 0.77 - - Gammaproteobacteria Aeromonadaceae 70, 73 - - - 0.77 - 2.26 Enterobacteriaceae 9 - - - - - 1.50 ‘Chrom1’ (Chromatiales) b 26 0.61 - - - - - ‘Gam1’ e 81 - - - 0.77 - - Planctomycetes Planctomycetaceae 10, 32, 33 3.68 - - - - - Proteobacteria ‘Prot1’ f 77 0.61 - - - - - Bacteria ‘Bac1’ g 17 - 1.71 - - - - ‘Bac2’ h 30 0.61 - - - - - ‘Bac3’ i 64 - 1.14 - - - - -, Not detected. a Identification of OTUs by BLASTn (http://blast.ncbi.nlm.nih.gov/). b New family-level taxon based on 16S rRNA gene similarities < 87% to next cultivated species (Yarza et al. 2008). c Next cultivated species: Prolixibacter bellaniivorans (AY918928; 86% 16S rRNA gene similarity). d Sequences derived from cellulose-supplemented treatments clustered to Cluster III Clostridiaceae, sequences from cellobiose- and glucose-supplemented treatments clustered to Cluster I Clostridiaceae (according to Collins et al. 1994).
R ESULTS 94 rRNA cDNA gene libraries of ‘heavy’ fraction but not identified as labeled taxon (3.2.4.2.2). Cloning of qPCR products amplified by primers specific for cellulolytic Clostridiaceae (Cluster III Clostridiaceae) revealed that none of the analysed sequences was related to sequences of this taxon (Table 26). In contrast to pure target DNA and non-target DNA for which the assay was specifically working, it was not possible to amplify any target sequence in environmental samples. Assays designed for quantification of the new family-level taxa ‘Cellu1-3’, ‘Sphingo1-4’ and ‘Deha1’ yielded 96%, 61%, and 80% (Table 26) specificity, respectively. Non-target sequences that were amplified by these assays were related to various organisms in the phyla Cyanobacteria, Deltaproteobacteria, Firmicutes, Gemmatimonadetes, and Nitrospirae. Although published protocols were used for the quantification of total Bacteria and Archaea, the assays were checked for their specificity. The Bacteria-targeting primers were highly specific and did not yield any non-target taxon (i.e., sequences were not non-bacterial) (Table 26). In contrast, approximately 44% of the sequences that were amplified with the Archaea-targeting primers were non-archaeal (i.e., sequences were bacterial).
R ESULTS 95 Table 26. Specificity of qPCR assays. Target sequences [%] a Non-target sequences [%] a Affiliation of nontargets (phylum, number of sequences) Family - level a ssays 1 Micrococcaceae & Cellulomonadaceae 100.0 (19) 0.0 (0) --- 2 Kineosporiaceae & Nocardioidaceae 78.9 (15) 21.1 (4) Acidobacteria, 1 Cyanobacteria, 2 Firmicutes, 1 3 Cluster I Clostridiaceae 84.2 (16) 15.8 (3) Gammaproteobacteria, 3 4 Cluster III Clostridiaceae 0.0 (0) 100.0 (20) Acidobacteria, 6 Actinobacteria, 1 Alphaproteobacteria, 6 Deltaproteobacteria, 5 Firmicutes, 2 5 Planctomycetaceae 100.0 (20) 0.0 (0) --- 6 ‘Cellu1-3’ b (Bacteroidetes) 95.5 (21) 4.5 (1) Cyanobacteria, 1 7 ‘Sphingo1-4’ b (Bacteroidetes) 61.1 (11) 38.9 (7) Deltaproteobacteria, 3 Gemmatimonadetes, 4 8 ‘Deha1’ b (Chloroflexi) 80.0 (16) 20.0 (4) Nitrospirae, 2 Firmicutes, 2 Domain - level a ssays 9 Bacteria 100.0 (20) 0.0 (0) --- 10 Archaea 55.6 (10) 44.4 (8) Actinobacteria, 1 Alphaproteobacteria, 1 Bacteroidetes, 1 Betaproteobacteria, 2 Firmicutes, 1 Planctomycetes, 2 a Parenthetical values are the number of sequences. b Name and affiliation as defined elsewhere (3.2.4.2.1).
R ESULTS 96 3.5. Effect of Pesticides on the Degradation of Cellulose and Cellobiose The pesticides Bentazon, Chlorothalonil, MCPA, Metalaxyl, and Nonylphenol were pre-screened for their potential to impact on the aerobic and anaerobic degradation of cellobiose (Figure 24). Figure 24. Degradation of supplemental cellobiose in soil treatments in the presence of pesticides . (A), (C): Pesticide without apparent effect (Chlorothalonil, Metalaxyl). (B), (D): Pesticide with apparent effect (Bentazon, MCPA, Nonylphenol). Symbols: Bentazon-supplemented treatments; Chlorothalonil-supplemented microcosm; MCPA-supplemented treatments; Metalaxyl-supplemented treatments; Nonylphenol-supplemented microcosms; Control treatments without pesticides (red). Closed symbols, cellobiose. Open symbols, glucose. Error bars indicate standard deviations (n =3). Cellobiose was hydrolysed without apparent delay in oxic and anoxic treatments, resulting in an accumulation of glucose. Parallel to the hydrolysis of cellobiose that lasted 20 hours in pesticide-free control treatments, released glucose was consumed. Concentrations of glucose decreased in pesticide-free controls within 10 hours (oxic) and 48 hours (anoxic) to values under or near the detection limit (~0.5 µmol g soil DW-1 ). Chlorothalonil and Metalaxyl did not apparently affect the aerobic and anaerobic degradation of cellobiose and the subsequent consumption of 0 6 12 18 24 0.0 0.1 0.2 0.3 0.4 0.0 0.1 0.2 0.3 0.4 A B Oxic 0 12 24 36 48 C D Anoxic Time [h] Cellobiose, Glucose [µmol g DW-1 ] 1.2 0.9 0.6 0.3 0.0 1.2 0.9 0.6 0.3 0.0 0 6 12 18 24 0.0 0.1 0.2 0.3 0.4 0.0 0.1 0.2 0.3 0.4 A B Oxic 0 12 24 36 48 C D Anoxic Time [h] Cellobiose, Glucose [µmol g DW-1 ] 1.2 0.9 0.6 0.3 0.0 1.21.2 0.90.9 0.60.6 0.30.3 0.00.0 1.2 0.9 0.6 0.3 0.0 1.21.2 0.90.9 0.60.6 0.30.3 0.00.0
R ESULTS 97 glucose (Figure 24A and C). In contrast, the presence of Bentazon, MCPA, and Nonylphenol resulted in a slower hydrolysis of cellobiose and a strong inhibition of the subsequent consumption of glucose (Figure 24B and D). Complete cellobiose hydrolysis was delayed for 4 hours compared to pesticide-free controls. Furthermore, the concentration of free glucose doubled in Bentazon-, MCPA-, and Nonylphenolsupplemented treatments under oxic and anoxic conditions resulting in remaining glucose at the end of incubation (Figure 24B and D). Based on these results, Bentazon, MCPA, and Nonylphenol were chosen as model pesticides to analyse the impact on the degradation of cellulose and cellobiose under oxic and anoxic conditions (Figure 25 and Figure 26; Table 27; Table A2, 2.3.3.3; 2.3.3.2). Different concentrations were applied to simulate in situ relevant (‘low’, 3.5.1) and elevated (‘high’, 3.5.2) concentrations of pesticides. Low amounts of pesticides were only used in soil slurries supplemented with cellobiose. 3.5.1. In Situ-relevant (‘Low’) Concentrations of Pesticides Addition of cellobiose stimulated production of carbon dioxide by approximately 60% and small amounts of molecular hydrogen accumulated under anoxic conditions (Figure 25). Methane and ferrous iron were not detected in any of the treatments. pH was stable at 6.0 (data not shown). Supplemental cellobiose was hydrolysed without apparent delay in both oxic and anoxic treatments, leading to a transient accumulation of glucose (Figure 25). Both cellobiose and glucose were totally consumed within 10 hours under oxic conditions, and within 42 hours under anoxic conditions in control treatments lacking pesticides (Figure 25). Carbon recovery was 47% and 55% in oxic and anoxic control treatments, respectively. No apparent effect on the hydrolysis of cellobiose and the production of carbon dioxide was evident under ‘low’ concentration of pesticides (Figure 25; Table 27; Table A2), whereas a slight inhibition of the consumption of glucose occurred. Glucose was not totally consumed after 10 hours in oxic treatments supplemented with Bentazon or Nonylphenol (Figure 25) what refers to an inhibition of 12% and 25% by Bentazon and Nonylphenol, respectively (Table 27). Addition of pesticides to anoxic treatments affected consumption of glucose more pronounced than addition of pesticides to oxic treatments. Likewise, the total consumption of glucose in anoxic treatments was slower in the presence of pesticides than in pesticide-free control treatments what refers to an inhibition of 24 – 37% (Table 27). Production of molecular hydrogen under anoxic conditions was less in the presence of MCPA and negligible in the presence of Bentazon or Nonylphenol at 42 hours (Table A2).
R ESULTS 98 Figure 25. Degradation of supplemental cellobiose in treatments under ‘low‘concentration of pesticides. Pesticide concentrations: Bentazon: 0.1 µmol g soil DW-1 ; Nonylphenol: 0.4 µmol g soil DW-1 ; MCPA, 0.01 µmol g soil DW-1 . Values are the means of duplicates. Error bars indicate standard deviations. Symbols: treatments supplemented with Bentazon (black), with Nonylphenol (white), with MCPA (grey), without pesticide supplementation (green, dashed line), or without substrate and addition of pesticides (red, dotted line). carbon dioxide; molecular hydrogen; ✖ cellobiose; glucose. Arrows, time point for calculations of inhibitory effect of supplemented pesticides as presented in Table 27 and Table A2. 3.5.2. Elevated (‘High’) Concentrations of Pesticides High concentrations of pesticides simulated the accumulation of pesticides that can occur in soil by heterogenous distribution (Marsh et al. 1978), especially in anoxic zones where pesticide degradation is slower compared to oxic zones. In treatments with higher concentration of pesticides cellobiose was not completely hydrolysed after 10 hours in oxic treatments (Figure 26), i.e., a reduction of the consumption of cellobiose of 38%, 38%, and 34% in the presence of Bentazon, MCPA, and Nonylphenol, respectively, was evident (Table 27). Likewise, the consumption of cellobiose-derived glucose and the production of carbon dioxide were lower in oxic treatments with high concentration of pesticides compared to treatments with low concentration of pesticides and controls lacking pesticides (Table 27). Between 0.3 ± 0.0 and 0.6 ± 0.0 µmol glucose g soil DW-1 were consumed after 10 days in oxic pesticide-supplemented treatments whereas the total amount of 1.4 ± 0.1 µmol glucose g soil DW-1 was consumed in the same period in CO 2 , H 2 [µmol g DW-1 ] 0 2 4 6 8 Time [h] 0 5 10 15 20 25 Sugars [µmol g DW -1 ] 0.0 0.4 0.8 1.2 Time [h] 0 10 20 30 40 50 Oxic Anoxic CO 2 , H 2 [µmol g DW-1 ] 0 2 4 6 8 Time [h] 0 5 10 15 20 25 Sugars [µmol g DW -1 ] 0.0 0.4 0.8 1.2 Time [h] 0 10 20 30 40 50 Oxic Anoxic
R ESULTS 99 unsupplemented controls (Table A3). Production of carbon dioxide at high concentration of pesticides was almost as low as the soil indigenous production in substrate-free control treatments. In the presence of Bentazon and MCPA (Figure 26, Table A3) a reduction of 68% and 56% was measured under oxic conditions (Table 27). Nonylphenol showed only a slight effect on the production of carbon dioxide under oxic conditions (Table 27). Supplemental cellobiose hydrolysis was not affected in anoxic treatments supplemented with high concentration of pesticides (Figure 26), but a reduced consumption of glucose was observed in the presence of Bentazon and MCPA (Figure 26). At high concentration of pesticides, consumption of glucose was delayed for more than 12 hours, resulting in the detection of residual glucose at the end of incubation (Figure 26). The amount of utilized glucose was 0.2 ± 0.1 µmol g soil DW-1 , and 0.3 ± 0.0 µmol g soil DW-1 in the Bentazon and MCPA- supplemented treatments, respectively (Table A3), what refers to a reduction of the consumption of glucose by more than 70% under high concentration of pesticides (Table 27). Interestingly, the initial anaerobic consumption of glucose was also slower under high Nonylphenol concentrations, but accelerated during incubation. This resulted in similar amounts of glucose consumed in both Nonylphenolsupplemented and unsupplemented control treatments (Figure 26, Table A3). Production of carbon dioxide at high concentration of pesticides was reduced by about 51 – 65% by MCPA and Bentazon, respectively (Table 27). No inhibition of the production of carbon dioxide or molecular hydrogen by Nonylphenol was obvious under anoxic conditions in which production of both gaseous compounds was stimulated, and concentrations exceeded control treatments at 42 hours (Figure 26). The anaerobic production of molecular hydrogen was negligible in the presence of Bentazon and MCPA (Table A3). Furthermore, the lack of fermentation products in these short incubations (Figure 24 – Figure 26; 3.5) demonstrates that strict anaerobes or facultative aerobes need time to form detectable amounts of molecular hydrogen, fatty acids, or alcohols after the depletion of O 2 . At high concentration of pesticides the pH immediately decreased to 5.4 ± 0.6 after the addition of pesticides, but stabilized at these value over the incubation period (data not shown).
R ESULTS 100 CO 2 , H 2 [µmol g DW-1 ] 0 2 4 6 8 10 Time [h] 0 5 10 15 20 25 Sugars [µmol g DW-1 ] 0.0 0.5 1.0 1.5 Time [h] 0 10 20 30 40 50 Oxic Anoxic CO 2 , H 2 [µmol g DW-1 ] 0 2 4 6 8 10 Time [h] 0 5 10 15 20 25 Sugars [µmol g DW-1 ] 0.0 0.5 1.0 1.5 Time [h] 0 10 20 30 40 50 Oxic Anoxic Figure 26. Degradation of supplemental cellobiose in treatments under ‘high‘concentration of pesticides . Pesticide concentration: Bentazon: 8.5 µmol g soil DW-1 ; MCPA, 2.4 µmol g soil DW-1 ; Nonylphenol: 3.5 µmol g soil DW-1 . Values are the means of duplicates. Error bars indicate standard deviations. Symbols: treatments supplemented with Bentazon (black), with Nonylphenol (white), with MCPA (grey), without pesticide supplementation (green, dashed line), or without substrate and addition of pesticides (red, dotted line). carbon dioxide; molecular hydrogen; ✖ cellobiose; glucose. Arrows, time point for calculations of inhibitory effect of supplemented pesticides as presented in Table 27 and Table A3. The impact of pesticides on the degradation cellulose was investigated by insertion of cellulose paper sheets into sieved fresh soil and destructive sampling, i.e., removing of one replicate from the analysis, every 7 days. Approximately 50% of the supplemented cellulose was degraded during the 70-day incubation under both oxic and anoxic conditions in pesticide-free control treatments, i.e., about 12 µmol cellulose g soil DW-1 was consumed (Figure 27, Table A3). Hydrolytic products of cellulose (cellobiose or glucose) were not detected in either oxic or anoxic treatments (detection limits approximated 0.5 µmol g soilDW-1 ). Organic products were not detected during the first four weeks, but small amounts of acetate, propionate, and butyrate accumulated in anoxic cellulose-supplemented treatments (< 5 µmol g soil DW-1 ) lacking pesticides, but not in oxic treatments. Production of carbon dioxide was stimulated up to 50% by the addition of cellulose under both oxic and anoxic conditions without pesticides (Figure 27, Table A3). Carbon recovery was 106% and 81% in oxic and anoxic control treatments, respectively. Addition of cellulose also stimulated ferric iron reduction under anoxic conditions. Ferrous iron reached concentrations of up to about 190 µmol g soil DW-1 . In treatments without substrate, ferrous iron was not
R ESULTS 101 detected (Figure 27, Table A3). Molecular hydrogen and methane were not detected in any of the cellulose-supplemented treatments. pH was stable at 6.2 (data not shown). Bentazon, MCPA, and Nonylphenol at 2.4 µmol g soil DW-1 impaired the degradation of cellulose under oxic conditions by 41%, 62%, and 47%, respectively. Bentazon and MCPA reduced production of carbon dioxide by 20 – 32% (Table 27). Nonylphenol had no apparent inhibitory effect (Figure 27). In contrast, Nonylphenol stimulated production of carbon dioxide under oxic conditions (Table A3) what can be explained by the degradation of this pesticide. After 28 weeks 2.4 µmol Nonylphenol g soil DW-1 were totally consumed in oxic treatments making a second application of Nonylphenol to remaining replicates nessecary (Figure 27). Concentrations of Bentazon and MCPA decreased slowly in oxic cellulosesupplemented treatments, resulting in about 50% residual pesticide at the end of incubation (Figure 27). Pesticide-impaired degradation of supplemental cellulose was more evident under anoxic than under oxic conditions, i.e. less than 1 µmol cellulose g soil DW-1 was consumed (Figure 27, Table A3). This corresponds to an inhibition of more than 90% by all pesticides (Table 27). Accumulation of organic products (i.e., acetate, propionate, and butyrate) did not exceed 1.5 µmol g soil DW-1 (data not shown) and production of carbon dioxide was impaired by 85%, 85%, and 44% for Bentazon, MCPA, and Nonylphenol, respectively (Table 27). Reduction of ferric iron was also reduced, i.e., 60 – 96% less ferrous iron accumulated in the presence of pesticides (Table A3). Concentration of pesticides was stable in anoxic cellulose-supplemented treatments (Figure 27). Figure 27. Degradation of supplemental cellulose in soil treatments in the presence of pesticides at 2.4 µmol g soil DW-1 . T end values are the means of triplicates. Error bars indicate standard deviations. Symbols: treatments supplemented with Bentazon CO 2 [µmol g DW-1 ] 0 50 100 150 Pesticide [µmol g DW-1 ] 0 1 2 3 Time [d] 0 14 28 42 56 70 Cellulose [µmol g DW-1 ] 0 10 20 30 Time [d] 0 14 28 42 56 70 Fe 2+ [µmol g DW-1 ] 0 50 100 150 Oxic Anoxic CO 2 [µmol g DW-1 ] 0 50 100 150 Pesticide [µmol g DW-1 ] 0 1 2 3 Time [d] 0 14 28 42 56 70 Cellulose [µmol g DW-1 ] 0 10 20 30 Time [d] 0 14 28 42 56 70 Fe 2+ [µmol g DW-1 ] 0 50 100 150 Oxic Anoxic
R ESULTS 102 (black), with MCPA (grey), with Nonylphenol (white), and without pesticide supplementation (green, dashed line), or without substrate and addition of pesticides (red, dotted line); carbon dioxide; ▼ pesticide (dashed line); cellulose; ▲ ferrous iron. Arrows, time point for calculations of inhibitory effect of supplemented pesticides as presented in Table 27 and Table A3. Nonylphenol was applied twice (day 0, day 28) in oxic treatments. Table 27. Inhibition of saccharide degradation, carbon dioxide and hydrogen emission, and production of ferrous iron in the presence of pesticides in soil treatments compared to control treatments. Inhibition [%] Oxic Anoxic Bentazon MCPA Nonylphenol Bentazon MCPA Nonylphenol ‘ Low’ concentration of pesticides – Cellobiose - supplemented a Cellobiose consumption d 11.1 ± 9.5 n.i. 15.4 ± 6.3 n.i. n.i. n.i. Glucose consumption d 12.2 ± 6.3 n.i. 25.1 ± 8.4 26.6 ± 13.5 23.5 ± 2.5 37.3 ± 2.3 CO 2 production d n.i. n.i. n.i. 13.6 ± 11.3 13.4 ± 1.4 n.i. H 2 production d - - - 100.0 ± 0.0 89.2 ± 1.5 100.0 ± 0.0 ‘ High’ concentration of pestici des – Cellobiose - supplemented b Cellobiose consumption d 38.3 ± 7.0 37.8 ± 6.0 34.4 ± 0.3 n.i. n.i. n.i. Glucose consumption d 81.0 ± 0.1 59.1 ± 4.3 61.1 ± 7.5 81.8 ± 6.0 73.4 ± 0.9 n.i. CO 2 production d 68.2 ± 3.3 55.6 ± 5.3 32.6 ± 2.4 64.7 ± 5.1 51.0 ± 5.9 n.i. H 2 production d - - - 100.0 ± 0.0 100.0 ± 0.0 n.i. ‘High’ concentration of pesticides – Cellulose - supplemented c Cellulose degradation e 40.8 ± 12.0 61.6 ± 2.6 47.4 ± 11.0 94.8 ± 5.0 93.8 ± 3.6 97.0 ± 3.4 CO 2 production e 20.3 ± 6.5 31.5 ± 4.5 n.i. 84.6 ± 0.8 85.6 ± 2.1 44.2 ± 12.8 Fe 2+ production e - - - 95.7 ± 0.5 95.4 ± 0.3 59.9 ± 1.0
R ESULTS 103 a 0.01 to 0.4 µmol pesticide g soil DW-1 . b 2.4 to 8.5 µmol pesticide g soil DW-1 . c 2.4 µmol pesticide g soil DW-1 . d Inhibition of saccharide degradation, carbon dioxide and hydrogen emission was calculated based on the differences in the concentration of products at 10 hours (oxic) and 42 hours (anoxic) referred to the control treatment (Table A2). e Inhibition of cellulose degradation, carbon dioxide and hydrogen emission, and ferric iron reduction was calculated based on the differences in the concentration of products at 70 days of incubation referred to the control treatment (Table A3). .i., No inhibition, i.e., amount of product was higher in pesticide-supplemented treatments than in control treatments or standard deviations where higher than the mean value. -, Not determined. 3.5.3. 16S rRNA Content of Bacterial Taxa 16S rRNA transcript numbers of total soil Bacteria and five saccharide-utilizing family-level taxa (Micrococcaceae/Cellulomonadaceae [Actinobacteria], Cluster I Clostridiaceae [Firmicutes], Planctomycetaceae [Planctomycetes], and the new taxa ‘Cellu1-3’ and ‘Sphingo1-4’ [Bacteroidetes]), were determined in soil samples of cellulose-supplemented treatments using reverse transcriptase quantitative PCR (RTqPCR). Taxa were quantified in Bentazon-supplemented, MCPA-supplemented, and pesticide-free control treatments, but not in Nonylphenol-supplemented treatments. RNA content of the total bacterial soil community decreased in all treatments from 6.4 × 10 10 transcripts ng RNA-1 to significantly lower values (Figure 28) independent of the presence or absence of pesticides. This consistent decrease might be explained by various factors, e.g., the death of non-saccharide-utilizing bacteria due to substrate limitation or grazing of Bacteria by eukaryotes feeding on bacteria (e.g., protozoa, ciliates). Bentazon and MCPA application lowered the number of transcript of total soil Bacteria in both oxic and anoxic treatments at the end of the experiment compared to pesticide-free treatments (Figure 28). RNA content decreased from 3.8 × 10 10 transcripts ng RNA-1 in pesticide-free controls to 1.5 × 10 10 transcripts ng RNA-1 in Bentazon-supplemented, and 1.2 × 10 10 transcripts ng RNA-1 in MCPA-supplemented treatments (Fig. 31) under oxic conditions. In anoxic treatments Bentazon and MCPA application lowered the number of 16S rRNA genes from 3.0 × 10 10 transcripts ng RNA-1 to 1.3 × 10 10 transcripts ng RNA-1 and 0.6 × 10 10 transcripts ng RNA-1 , respectively (Figure 28).
R ESULTS 110 Transcripts ng RNA-1 048 102 168 192 Time [h] 10 10 10 8 10 6 10 4 * ** * * * * Figure 31. Numbers of 16S rRNA gene transcripts of Bacteria and Archaea in treatments in incubation chambers . The filled bars represent values obtained from CMC-supplemented treatments. The shaded bars represent values obtained from the cellobiose-supplemented treatments. The empty bars represent values obtained from unsupplemented control treatments. Error bars indicate standard deviations (n = 3). Bar code: non-hatched, Bacteria; hatched, Archaea. An asterisk indicates that a transcript number is statistically different from the transcript number of the previous time point (t-Test). The synthesis of 16S rRNA of family-level taxa was differentially affected by addition of CMC and cellobiose (Figure 32 and Table A1). A selective activation and/or repression of functionally redundant taxa was observed (Figure 32). No family-level taxon was stimulated without addition of substrate, i.e., the amount of 16S rRNA genes did not change significantly in unsupplemented controls (Table A1). The relative amount of 16S rRNA of Micrococcaceae/Cellulomonadaceae increased significantly from 9.8 × 10 -5 to 5.6 × 10 -5 transcripts transcripts Bacteria-1 in the first oxic period of cellobiose-supplemented treatments, but not in CMC-supplemented treatments (Figure 32A, Table A1). Anoxic incubation (48 – 168 hours) resulted in a significant decrease of transcript numbers of Micrococcaceae/Cellulomonadaceae in cellobiose-supplemented treatments (Figure 32A). 16S rRNA gene transcript numbers of Planctomycetaceae behaved inversely (Figure 32D). Addition of cellobiose did apparently not affect RNA content of Planctomycetaceae (Figure 32D, Table 24), but addition of CMC resulted in a significant increase from 1.4 × 10 -6 to 1.8 × 10 -5 transcripts transcripts Bacteria-1 during re-aeration (Table A1). Members of Kineosporiaceae/Nocardioidaceae were not stimulated by either CMC or cellobiose (Figure 32B). In contrast, relative transcript numbers of this taxon decreased during the oxic and the first half of the anoxic period (0 – 168 hours) from 6.3 × 10 -7 to 2.1 × 10 -7 transcripts transcripts Bacteria-1 and from 9.3 × 10 -7 to 1.3 × 10 -7 transcripts transcripts Bacteria-1 in CMC- and cellobiose-supplemented treatments, respectively (Table A1). Similar results were obtained for the uncultured taxon ‘Deha1’ (phylum Chloroflexi), for which stimulation by addition of substrates was not apparent (Figure 32F, Table A1). Transcript numbers of Cluster I Clostridiaceae and the new familylevel taxon ‘Cellu1-3’ (phylum Bacteroidetes) increased under anoxic conditions
R ESULTS 111 (48 - 68 h) in cellobiose-supplemented treatments but not under oxic conditions (Figure 32C and E). Transcript numbers of Cluster I Clostridiaceae and ‘Cellu1-3’ increased to 1.9 × 10 -7 and 1.0 × 10 -7 transcripts transcripts Bacteria-1 after 168 hours of incubation in the presence of cellobiose, respectively (Table A1). CMC did not appear to stimulate these taxa, neither in the oxic period nor in the anoxic period (Figure 32C and E, Table A1). The uncultured taxon ‘Sphingo1-4’ (phylum Bacteroidetes) was the only taxon that was slightly stimulated by both substrates (Figure 32G, Table A1). In the second half of the anoxic period 16S rRNA genes increased from 8.4 × 10 -9 to 2.2 × 10 -8 transcripts transcripts Bacteria-1 and from 2.1 × 10 -6 to 4.5 × 10 -6 transcripts transcripts Bacteria-1 in CMC- and cellobiosesupplemented treatments, respectively (Table A1).
R ESULTS 112 Figure 32. Effects of CMC and cellobiose on taxon-specific 16S rRNA gene transcripts. The shaded areas correspond to the anoxic periods. Symbols: filled, values obtained from CMC-supplemented treatments; grey, values obtained from cellobiosesupplemented treatments; empty, values obtained from unsupplemented control treatments. Values are the means of two replicates (without standard deviation) or three replicates (with standard deviation). The highest number of taxon-specific transcripts per assay (i.e., in CMC, cellobiose, and unsupplemented control treatments) was set at 100% (absolute and reference values are marked in Table A1). An asterisk indicates that a transcript number is statistically different from the transcript number of the previous time point (t-Test). 0 50 100 150 0 50 100 150 0 50 100 150 200 0 50 100 150 A Micrococca- /Cellulomonadaceae B Kineosporia- /Nocardioidaceae C Cluster I Clostridiaceae D Planctomycetaceae E ‘Cellu1-3’ F ‘Deha1’ 0 50 100 150 200 0 1 50 100 150 G ‘Sphingo1-4’ * ** * * * * * * * * * ** Change in Transcripts [%] ** * Time [h] Time [h]
R ESULTS 113 A shift in the ratio of family-level taxa was also observed. Micrococcaceae/Cellulomonadaceae became the most abundant of all detected groups in cellobiose-supplemented treatments after the initial oxic phase. Such an enrichment of rRNA of these Actinobacteria was not observed in CMC-supplemented treatments (Figure 33). In contrast, 16S rRNA cDNA genes of the new family-level taxon ‘Deha1’ (Chloroflexi) comprised a high proportion of the autochthonous soil microbial community (Figure 33), but neither this taxon nor other detected taxa were remarkably activated in CMC- or cellobiose-supplemented treatments (Figure 32). Figure 33. Relative abundances of family-level taxa in (A) CMC- supplemented treatments, (B) cellobiose-supplemented treatments, and (C) unsupplemented control treatments. 100% corresponds to the total number of transcripts of all taxa (numbers above bars in [× 10 -3 transcripts ng RNA-1 ]). Time [h] 0 48 102 168 192 [%] 0 20 40 60 80 100 0.9 0.40.5 0.5 0.7 A Time [h] 0 48 102 168 192 1.4 6.8 6.2 3.9 4.4 B Time [h] 0 48 102 168 192 [%] 0 20 40 60 80 100 1.2 0.9 0.8 0.9 0.7 C Clostridia III 'Cellu' ( ‘Cellu1-3‘ (Bacteroidetes) Cluster I Clostridiaceae ‘Deha1‘ (Chloroflexi) Kineosporiaceae & Nocardioidaceae Kineosporia-/Nocardioidaceae Micrococcaceae & Cellulomonadaceae Clostridia Micrococca-/Cellulomonadaceae Planctomycetaceae ‘Sphingo1-4‘ (Bacteroidetes) % %
D ISCUSSION 114 4. D ISCUSSION Cellulose is the most abundant biopolymer on earth and its biological degradation in soil is a major process in the global carbon cycle (Bayer et al. 2006; Falkowski et al. 2000; Lal 2008; Lynd et al. 2002). Agricultural soils account for about 40% of the world’s terrestrial surface area and are hot spots for the turnover of organic carbon (Körber et al. 2009). Agricultural soil is a highly structured environment in which oxic and anoxic microzones co-occur on a small scale (Figure 3), and in which the distribution and dimension of such zones can change rapidly (Dassonville et al. 2004; Or et al. 2007; Six et al. 2000; Totsche et al. 2010; Zausig et al. 1993). This compartmentalization facilitates the coexistence and simultaneous activity of aerobic and anaerobic soil microorganisms that catalyse the degradation of plant-derived cellulose under contrasting availabilities of O 2 . However, less is known about the phylogenetic identities of cellulose-degrading microbial communities in situ and their metabolic response to rapid changes in O 2 . Another crucial factor is the increasing usage of pesticides in agricultural ecosystems. Application of herbicides, fungicides, and insecticides has resulted in an accumulation of pesticide residues in terrestrial and aquatic habitats (Akerblom 2004; Hiller et al. 2008; Thorstensen et al. 2001). Pesticides can impact on soil communities (Cox et al. 1996; Johnsen et al. 2001; Katayama and Kuwatsuka 1991; Porter and Hayden 2002), and might alter the cellulose-dependent carbon flow and the activity of involved microbial taxa in agricultural ecosystems. Trophic interactions between cellulose-degrading microorganisms are complex and many uncultured species are likely important to the degradation of plant-derived carbon (Haichar et al. 2007; Li et al. 2009). To understand the process of cellulose degradation in such a dynamic environment, it is essential (i) to identify involved cellulose-degrading soil communities and (ii) to investigate the impact of O 2 and pesticides on active microbial taxa. 4.1. Aerobic and Anaerobic Degradation of Saccharides in An Aerated Agricultural Soil A community function, e.g., the utilization of glucose or the turnover of nitrogen, may remain stable under changing O 2 states (Picek et al. 2000; Santruckova et al. 2004) although the composition of the microbial community shifts (Pett-Ridge and Firestone 2005; Pett-Ridge et al. 2006). In this regard, the capacity of the saccharidedegrading community to readily consume supplemental substrates was not affected by the availability of O 2 or changing redox potentials (Figure 11, Figure 27, Figure 30). Supplemented carboxymethyl-cellulose (CMC) was consumed without apparent delay in oxic and anoxic treatments (Figure 30A), indicating that the saccharidedegrading community of the soil was poised to both aerobically and anaerobically consume this substrate. In contrast to CMC-supplemented treatments, a short lag phase was observed in treatments that received cellulose in the insoluble form, i.e., cellulosic paper sheets (Figure 27). The lag phase before measurable degradation of cellulose paper suggests an initial colonization of the substrate by cellulolytic microorganisms prior to its biological degradation (Munier-Lamy and Borde 2000). A lower abundance of accessible regions for enzymatic hydrolysis in crystalline cellulose compared to soluble cellulose-derivates might also account for the slow degradation of cellulose paper. Endo- and exoglucanases excreted by cellulolytic microorganisms attack crystalline cellulose at internal amorphous regions, and increase the amounts of reducing and non-reducing chain ends (Bayer et al. 1998a; Bayer et al. 1998b; Bisaria and Ghose 1981; Lynd et al. 2002; Teeri 1997; Figure 2).
D ISCUSSION 115 Thus, prolonged incubation promotes growth of cellulolytic microorganisms on cellulose sheets by an increase of substrate accessibility. The lack of detectable cellodextrins, cellobiose, and glucose during cellulose degradation (Figure 11A, Figure 27, Figure 30A) suggests a very efficient assimilation of released hydrolysis products by cellulolytic microorganisms and saccharolytic satellite organisms (Bayer et al. 1994; Bayer et al. 2004; Stanier 1942; Xie et al. 2007; Figure 4, 1.5), especially under oxic conditions. Under anoxic conditions, anaerobic cellulolytic microorganisms attack cellulosic substrates mainly by an extracellular enzyme complex (i.e., cellulosome; 1.2.1.2) that is covalently linked with the cell envelope and attaches tightly to the substrate by carbohydrate binding domains (Bayer et al. 1998b; Beguin and Aubert 1994; Desvaux 2005b; Lynd et al. 2002). Thus, the release of soluble hydrolysis products is minimal (Bayer et al. 1994; Beguin and Aubert 1994). Cellobiose was hydrolyzed, resulting in a transient accumulation of glucose (Figure 11B, Figure 24 – Figure 26, Figure 30B). This phenomenon may have been caused by extracellular β-glucosidases that hydrolyzed cellobiose faster than glucose was consumed by microorganisms (Hong et al. 1981; Lynd et al. 2002). Cellobiose also accumulated in oxic stable isotope probing treatments with prolonged incubation (Figure 11B) what suggests a subsequent product-mediated inhibition of cellobiose hydrolysis/uptake by increasing concentrations of glucose (Corazza et al. 2005; Gong et al. 1977; Kajikawa and Masaki 1999). However, it remains unclear, why the hydrolysis of cellobiose was inhibited in oxic treatments of stable isotope probing experiment, but not in other experiments. Nevertheless, supplemented cellobiose and glucose were aerobically and anaerobically consumed. Carbon recovery was below 60% in oxic treatments supplemented with cellobiose or glucose (3.1, 3.5) which indicates that a substantial amount of carbon was assimilated during aerobic conversion. This trend is consistent with studies that assessed the incorporation of carbon from [ 13 C]- and [ 14 C]-labeled cellulose, cellobiose, or glucose into microbial biomass of forest, savannah, and loamy soil (DeForest et al. 2004; Fontaine et al. 2004; Schneckenberger et al. 2008). According to literature, carbon dioxide was the main end product of the degradation of supplemented saccharides under oxic conditions (Bayer et al. 2006; Beguin 1990, Beguin and Aubert 1994; Schmidt and Ruschmeyer 1958). These results demonstrate that processes that are linked to the aerobic degradation of saccharides are similar in various types of soil and might be catalyzed by microbial communities with similar ecosystem functions. Carbon recovery was above 100% in anoxic treatments of stable isotope probing experiments (3.1) suggesting that supplemental substrates augmented the concomitant utilization of soil indigenous carbon and enhanced turnover of microbial biomass (Blagodatskaya and Kuzyakov 2008; de Nobili et al. 2001). This so-called ‘Priming Effect’ results in increased production of carbon dioxide due to stimulation of degradation of already available carbon compounds by soil microorganisms after addition of [ 13 C]-cellulose or other organic carbon compounds (Blagodatskaya et al. 2007; Blagodatskaya and Kuzyakov 2008; Fontaine et al. 2004; Zyakun and Dilly 2005). Products that are indicative of fermentative metabolisms accumulated under anoxic conditions (Figure 11, Figure 30), a result consistent with previous laboratory studies that have investigated the effects of O 2 limitation on aerated soils (Degelmann et al. 2009a; Küsel and Drake 1995). In flooded or wetland soils, cellulose, and its initial main breakdown products cellobiose and glucose are decomposed by anaerobes into fatty acids, alcohols, molecular hydrogen, and carbon dioxide, physiological events that can be coupled to acetogenesis and methanogenesis (Drake et al. 2009; Westermann 1996). However, hydrogenotrophic
D ISCUSSION 116 acetogenesis (i.e., the hydrogen-dependent reduction of carbon dioxide to acetate by acetogens [Drake et al. 2006; Drake et al. 2008]) was likely of minor importance since the headspace was regularly flushed with dinitrogen (2.3.1). The production of acetate, butyrate, propionate, and molecular hydrogen (Figure 11, Figure 30, 3.1, 3.6.1) suggests that a combination of fermentation types was active under anoxic conditions, including mixed acid fermentation and Clostridia-typical propionate or butyrate fermentation (Buckel 2005; Gottschalk 1986; White 2007). These fermentative metabolisms are likely sources of molecular hydrogen (Buckel 2005; White 2007). In contrast to forest soils that can form huge amounts of succinate and ethanol via Enterobacteriaceae-facilitated mixed acid fermentation under anoxic conditions (Degelmann et al. 2009a), the agricultural soil investigated in the present study did not yield these products in significant amounts, suggesting that they were not main products of sugar fermentation under experimental conditions or subjected to consumption. This corresponds to the low relative abundance of active Enterobacteriaceae in anoxic treatments (Table 22). In contrast, addition of organic carbon to anoxic soil (i.e., rice field or fen soil) results in an immediate accumulation of molecular hydrogen, acetate, and propionate (Hamberger et al. 2008; Penning and Conrad 2007; Wüst et al. 2009). Interestingly, supplementation of cellobiose stimulated the production of lactate under anoxic conditions in treatments that were subjected to fluctuating availabilities of O 2 (2.3.2), and lactate was not rapidly consumed after re-aeration (Figure 30B). In contrast, lactate was not a detectable compound in stable isotope probing experiments (Figure 11; 3.1). Instead, propionate accumulated (Figure 11). Propionate is classically produced by strict anaerobic bacteria of the Propionibacteriaceae (Actinobacteria), Veillonellaceae, Clostridiaceae, and ‘Streptopeptococcaeae’ (Firmicutes; Graham et al. 2011). Propionate-producing bacteria can ferment glucose, alcohols, and some amino acids to propionate (Piveteau 1999), but their preferred substrate is lactate (Graham et al. 2011). Lactate can either be reduced via the acrylate pathway to acetate and propionate, or converted stepwise via the succinate-propionate pathway to succinate and propionate (Graham et al. 2011). Succinate can be an intermediate or an end product of Propionibacteria-related fermentation (Piveteau 1999). Neither succinate nor Propionibacteria-related sequences were detected in stable isotope probing experiments (Figure 11, Table 19, Table 20, Table 23) what suggests that the production of propionate was not catalyzed by Propionibacteriaceae. The lack of lactate and succinate during long anoxic incubation (up to 84 days in stable isotope probing experiments; 3.1) and their detection during shorter anoxic incubation (< 10 days; 3.6.1) indicates that these compounds are intermediates of the degradation of saccharides, are rapidly turned over, and likely function as precursors of propionate- and acetate-formation by the microbial soil community. Formate that was produced in traces during anoxic periods was also not a stable intermediate (data not shown). A similar route of carbon turnover is also supposed to be relevant in fen soil or in the gut of earthworms (Hamberger et al. 2008; Wüst et al. 2011). Parallel to the degradation of supplemented saccharides nitrate was consumed in most oxic and anoxic incubations (Figure 29, Figure 30; 3.1, 3.6.1). This indicates that nitrate was subjected to assimilation and dissimilation. Aerobic and anaerobic assimilation of nitrate via nitrite (NO 2- ) and ammonium (NH 4+ ) into biomass (Fuchs 2007; Rudolf and Kroneck 2005; Tiedje 1988), anaerobic dissimilation of nitrate to nitrite (DNR; Stolz and Basu 2002) or dinitrogen (denitrification; Payne 1981), and anaerobic dissimilation of nitrate to ammonium (DNRA; Fuchs 2007; Simon 2002) may account for the consumption of nitrate. The capability to assimilate nitrate is common for most bacteria, fungi, algae, and higher plants, and necessary to
D ISCUSSION 117 incorporate nitrogen into cellular components (Rudolf and Kroneck 2005). In contrast, dissimilation of nitrate is catalyzed almost exclusively by Prokaryotes and is performed to gain energy under anoxic conditions (Tiedje 1988). Concentrations of nitrate increased in CMC-supplemented and unsupplemented treatments during the oxic period (3.6.1), and decreased under anoxic conditions (Figure 29, Figure 30). This suggests that the production of nitrate by nitrification exceeded the consumption of nitrate (Sprent et al. 1987; Yu et al. 2007), and that nitrate was anaerobically assimilated or utilized as electron sink by the soil community via dissimilatory pathways (Conrad 1996, Dassonville et al. 2004; Peters and Conrad 1996; Yu et al. 2007). Available nitrate might also account for the apparent lack of ferrous iron and fermentation product formation (Dassonville et al. 2004; Picek et al. 2000). Thermodynamically high redox potentials do not favor most fermentations and the reduction of nitrate or ferric iron (Bohn 1971). The reduction of nitrate and ferric iron in water-unsaturated bulk soil rather occurs at 230 mV and 150 mV, respectively (Mansfeldt 2004). Values of 300 mV (nitrate reduction) and 50 mV (ferric iron reduction) are measured in water-saturated rice field soil, respectively (Yu et al. 2007). Most fermentations (except lactate fermentation by aerotolerant lactic acid bacteria; Brioukhanov and Netrusov 2007; Chen et al. 2005; Matthies et al. 2004) occur optimally at redox potentials below -100 mV (Dassonville et al. 2004; Picek et al. 2000). Hence, the presence of nitrate in CMC-supplemented and unsupplemented treatments might have been responsible for the stabilization of redox potentials above 320 mV (Figure 29, Figure 30A). In contrast to CMC- and unsupplemented treatments, a rapid decrease of the redox potential in cellobiose-supplemented treatments occurred when O 2 was removed. The redox potential dropped more than 700 mV within 10 h (Figure 30B). It is likely that this change in redox potential was due to the metabolic activity of saccharide-utilizing microbes that coupled the oxidation of saccharides to the reduction of high redox potential electron acceptors such as oxygen, nitrate, and ferric iron (Figure 11, Figure 30B; Dassonville et al. 2004; Küsel et al. 2002; Lovley et al. 2004; Peters and Conrad 1996). This conclusion is consistent with the labeling of strict anaerobic and facultative aerobic taxa capable of the reduction of nitrate or ferric iron. For example, species of Cellulomonadaceae, Rhodocyclaceae, Clostridiaceae, Geobacteraceae, Pelobacteraceae and Aeromonadaceae perform denitrification, reduce nitrate to nitrite or ammonium, or dissimilate ferric iron under anoxic conditions (Colwell et al. 1986; Kang et al. 2007; Lee 2006a; Lovley et al. 2004; Shida et al. 1997; Tiedje 1988; van Keulen et al. 2007; Wiegel et al. 2006). It is generally supposed that microorganisms that reduce ferric iron do not directly utilize complex organic matter (Lovley 2006; Lovley et al. 2004), but preferribly utilize fermentation products, such as fatty acids and alcohols, as electron donors (Lovley et al. 1993; Lovley et al. 1995). However, some iron reducers can also utilize glucose (e.g., Aeromonadaceae, Clostridiaceae; Dobbin et al. 1999; Pham et al. 2003). Hence, it can be assumed that species of Aeromonadaceae and Clostridiaceae catalyzed the saccharolytic reduction of ferric iron in the experiments. Concomitant to the reduction of alternative electron acceptors with positive redox potentials and the high amounts of readily accessible organic carbon, the decreasing redox potential favored fermentation processes in cellobiosesupplemented treatments during the anoxic period (Figure 30B). Re-aeration of cellobiose-supplemented treatments resulted in a strong increase of the redox potential and fast consumption of fatty acids and ferric iron (Figure 30), the latter processes likely being mediated by heterotrophic aerobes.
D ISCUSSION 118 Methane was not produced in any experiment under anoxic conditions. This result is in contrast to wetland soils that yield methane from saccharides via ‘intermediary ecosystem metabolism’ (i.e., by the trophically linked processes that precede methanogenesis [Drake et al. 2009]; 1.2.2). The production of methane in acidic fen soil can be stimulated by methanogenic precursors like formate or carbon dioxide, and non-methanogenic precursors like xylose or glucose (Hamberger et al. 2008; Hunger et al. 2011). This is coincident with the occurrence of active methanogens and typical fermenting species in fen soil, suggesting that fermentation processes and methanogensis are trophically linked (Hamberger et al. 2008; Hunger et al. 2011). Furthermore, methanogens can be cultivated from aerated soils (Peters and Conrad 1995), and some species tolerate oxygen (Krätzer et al. 2011; Liu et al. 2008; Tholen et al. 2007). Methane production can occur in aerated soil samples if they are incubated for long periods (> 100 days) under anoxic conditions (e.g., Küsel and Drake 1996; Sexstone and Mains 1990; Vor et al. 2003). However, methanogens cannot utilize alternative electron acceptors. Thus, the primarily oxic nature of the investigated soil, the short incubation time, and the presence of alternative electron acceptors do not favor methanogenesis (Küsel and Drake 1994; Küsel et al. 1999). Taking all this findings into account, it can be suggested that (i) the degradation of cellulose and cellulose-derived carbon in agricultural soil is independent of the availability of O 2 and (ii) that the activity of saccharide-utilizing soil microorgansims alters the soil redox potential. 4.2. Active Saccharide-Degrading Prokaryotic Communities Different aerobic and anaerobic bacterial communities assimilated [ 13 C]- cellulose and [ 13 C]-cellulose-derived saccharides which resulted in the detection of 48 labeled family-level taxa (Table 22). The majority of these families had low relative abundances (< 3%) and were likely of minor importance for the degradation of cellulose (4.2.3). In contrast, some family-level taxa comprised higher relative abundances (> 3%) what suggests that these groups are cellulolytic and saccharolytic key taxa in the cellulose-linked carbon flow (4.2.1, 4.2.2). Archaea were not labeled in the current study indicating that the experimental conditions did not favour the activity of archaeal community members, and that supplemented [ 13 C]-saccharides were not assimilated by the archaeal soil community. This is in contrast to studies on fen soil and anoxic municipal waste landfill cover soil (Hamberger et al. 2008; Li et al. 2009; Wüst et al. 2009). Crenarchaeota are common in the environment and can dominate the ammoniumoxidizing community in agricultural soil (Aravalli et al. 1998; Gubry-Rangin et al. 2010; Leinigner et al. 2006). Most abundant Archaea in aerated soils are nonmethanogenic mesophiles of Thaumarchaea (Pester et al. 2011). These mesophilic organisms are generally assumed to be nitrifiers. Nonetheless, the largest proportion of the diversity of soil Archaea is physiologically uncharacterized (Bates et al. 2011; Pester et al. 2011). In some archaeal genomes cellulase genes were detected (Ando et al. 2002; Bauer et al. 1999), but known cellulolytic Archaea are extreme thermophiles (Graham et al. 2011; Perevalova et al. 2005; Robb et al. 2011). Thus, it is unlikely that cellulolytic Archaea play an important role in cellulose degradation in the investigated temperate soil community.
D ISCUSSION 119 4.2.1. Key Taxa that Degraded Saccharides under Oxic Conditions Although fungal growth on cellulose sheets was observed (data not shown) bacteria contributed to the degradation of cellulose sheets under oxic conditions (3.2.3.2.1). Cellulolytic and saccharolytic aerobic taxa detected in this study have been previously identified in and isolated from soil by the use of crystalline cellulose (Haichar et al. 2007; Lo et al. 2009; Patel and Vaughn 1973; Ulrich et al. 2008). In contrast, prokaryotic communities that degrade plant organic matter and leaves are dissimilar to active taxa of the current study (Bernard et al. 2007; Lee et al. 2011; Tanahasi et al. 2005). These differences might be based on the composition of plant litter. Plant cell walls consist of a lignocellulose complex, in which cellulose is embedded in lignin, hemicellulose and other polymers such as starch and proteins (Chang 2007; Kumar et al. 2008a; Malherbe and Cloete 2002). These polymers as well as other carbonaceous plant-derivates (e.g., photosynthetic products) are subjected to microbial mineralization (Boddy et al. 2007; Nultsch 2000; Six et al. 2004) and yield a broader diversity of soil organisms than pure cellulose. Bacterial communities responsible for the degradation of [ 13 C]-labeled rice callus are dominated by saccharolytic members of Gammaproteobacteria, Flavobacteria, Sphingobacteria or Acidobacteria (Lee et al. 2011). Due to the low cellulose content of rice callus, active bacteria incorporate [ 13 C]-carbon derived from water soluble organic compounds, but less from cellulose (Lee et al. 2011). Likewise, rice straw is mainly degraded by Alpha-, Beta-, Gamma-, and Deltaproteobacteria, Bacteroidetes, Flavobacteria, Spirochaeta, and Verrucomicrobia (Bastian et al. 2009; Bernard et al. 2007; Tanahasi et al. 2005). Actinobacteria were the most labeled Bacteria in the stable isotope probing experiment and had highest abundances compared to other quantified saccharidedegrading taxa under oxic conditions. Actinobacteria comprise the majority of known aerobic cellulolytic isolates (de Boer et al. 2005; Lynd et al. 2002) and are one of the best studied phyla in the domain Bacteria (Stackebrandt and Schuhmann 2006). Actinobacteria include strict and facultative aerobic species that often build mycelium-like structures, i.e., hyphae (Eriskon 1949; Goodfellow and Williams 1983; Stackebrandt and Schuhmann 2006; Ventura et al. 2007). Hyphae-forming cellulolytic members may have an ecological advantage. By penetrating cellulosic material they achieve a better access to attackable regions and bring their cellulases in closer proximity to the substrate (Abdulla and El-Shatoury 2007; Chater et al. 2010; Eriksson et al. 1990; Lynd et al. 2002). Cellulolytic and saccharolytic Actinobacteria are often identified as parts of the bacterial community that is involved in the aerobic degradation of plant-derived carbon (e.g., Dumova and Kruglov 2009; Hofsten et al. 1971; Lee et al. 2011; Lo et al. 2009; Patel and Vaughn 1973; Tanahasi et al. 2005; Ulrich et al. 2008). Interestingly, cellulolytic Actinobacteria were not frequently labeled in oxic cellulose-supplemented treatments (Table 22, Figure A1). Only some sequences affiliated with Cellulomonas composti, a facultative cellulolytic bacterium of the family Cellulomonadaceae (Kang et al. 2007). This family consists of primarily aerobic species that are well known for their ability to degrade crystalline cellulose (An et al. 2005; Bagnara et al. 1985; Kang et al. 2007; Lo et al. 2009; Patel and Vaughn 1973; Ulrich et al. 2008). They are numerically important in soil and involved in the degradation of rice straw or the degradation of other plant organic matter (Lo et al. 2009; Stackebrandt and Keddie 1986; Stackebrandt et al. 2006; Tanahasi et al. 2005). The low coverage of the corresponding gene library (Table 21; 64%) might account for the low relative abundance of cellulolytic Actinobacteria that degraded crystalline cellulose under oxic conditions. With the
D ISCUSSION 126 cellulolytic activity is linked to species of Cluster IV and XIV (Uz and Ogram 2006). Anaerobic cellulolytic Clostridia can be limited to a range of carbohydrates (Li et al. 2009; Lynd et al. 2002), growing best on cellulose and cellodextrins, but showing reduced growth on cellobiose and glucose (Ng and Zeikus 1982; Zhang and Lynd 2005). For example, Clostridium thermocellum has bioenergetic benefits by growth on cellulose-derived cellodextrines which is a result of the lower energy input when cellodextrines are taken up, and are then intracellularly cleaved as compared to up take of cellobiose and glucose (Zhang and Lynd 2005). Growth of the anaerobe Bacteroides cellulosolvens is only supported by cellulose and cellobiose, but not by glucose or other carbon sources (Murray et al. 1984). Aerobic cellulolytic species can also have different preferences towards cellulose and soluble sugars. Molar growth yields of Actinotalea fermentans (Yi et al. 2007) are higher on cellulose compared to glucose (Bagnara et al. 1987). Another point that might account for the observed differences is the tight attachment of the cellulosome of cellulolytic anaerobes to cellulose (Bayer et al. 1994a; Bayer et al. 2006). As a consequence, the release of hydrolysis products, e.g., cellobiose or glucose, is minimal (Bayer et al 1994a; Bayer et al. 2006) what was presumably growth-limiting for saccharolytic satellite organisms in anoxic cellulose-supplemented treatments. Supplementation of soluble saccharides might have further favored saccharolytic organisms rather than cellulolytic ones. Saccharolytic bacteria are well adapted to rapidly utilize cellobiose or glucose (Lynd et al. 2002), but certain cellulolytic species, e.g., Sporocytophaga myxococcoides, need time for adaptation to soluble sugars (Sjipstein and Farhaeus 1949). Hence, an out-competition of cellulolytic taxa by saccharolytic taxa in cellobiose- and glucose-supplemented treatments is likely. These conclusive results suggest that (i) a large uncultured diversity of soil Bacteria was involved in the degradation of cellulose and products of its hydrolysis, and (ii) the active saccharolytic community differed phylogenetically from the active cellulolytic community under and between oxic and anoxic conditions (Hypothesis 1, 1.7). 4.2.5. Degradation of Cellulose is a Stable Community Function in the Investigated Soil The distribution of oxic and anoxic microzones in aerated soils is highly dynamic and may change rapidly (1.6.1). It is conceivable that both the degradation of cellulose and the active saccharide-degrading community is impacted by fluctuations of the availability of O 2 , e.g., after precipitation events. However, there was no apparent delay in the consumption of supplemented saccharides during oxic and anoxic periods (Figure 30). Other functions of the soil microbial community are also not dependent on the availability of O 2 and the redox potential. For example, although glucose consumption rates in aerated soil can vary in response to O 2 , the capacity of the microbial community to consume glucose is independent of redox potential (Picek et al. 2000; Santruckova et al. 2004). Also, the availability of O 2 does not appreciably affect nitrogen turnover rates in tropical forest soil, even though the presence or absence of O 2 engages different taxa of the microbial community that are involved in nitrogen turnover (Pett-Ridge and Firestone 2005; Pett-Ridge et al. 2006). Likewise, selective activation of aerobic and anaerobic saccharide-degraders under contrasting availabilities of O 2 allowed for a continous degradation of cellulose in this study (Figure 32). Micrococcaceae and Cellulomonadaceae were the taxa that responded most pronounced to the supplementation of saccharides under oxic conditions (Figure 32A). A slight stimulation was also evident for Planctomycetaceae and the
D ISCUSSION 127 new family ‘Sphingo1-4’ (Bacteroidetes) during re-aeration (Figure 32D and G). Cluster I Clostridiaceae and the new Bacteroidetes-related taxa ‘Sphingo1-4’ and ‘Cellu1-3’ responded positively under anoxic conditions (Figure 32C, E and G). Thus, it can be concluded that the rapid metabolic response of functionally redundant taxa, i.e., taxa that have the capacity to utilize the same substrate under contrasting environmental conditions (e.g., oxic and anoxic), contributes to the stability of a particular function of the microbial community despite differing availabilities of O 2 (Hypothesis 2, 1.7). 4.3. Diversity of Metabolic Active but not [ 13 C]-Labeled Bacteria Thirty-five of 83 detected family level-taxa were not labeled in stable isotope probing experiments (Table 23; 3.2.3.2). RNA stable isotope probing was applied, because it is more advantageous than DNA or PLFA (phospholipid fatty acid) stable isotope probing (Boschker et al. 1998; Manefield et al. 2002a; Radajewski et al. 2000). RNA is synthesised at higher rates than DNA, labeling of RNA occurs without cellular replication, and RNA provides sequence information that allows a higher phylogenetic resolution than PLFA analysis (Manefield et al. 2002a; Manefield et al. 2002b; Radajewski et al. 2003). Furthermore, DNA persists in the environment after cell death and cannot be easily used to define the metabolic state (i.e., active or inactive) of the organism it belongs to (Coolen and Overmann 1998). In contrast, RNA is labile and degrades rapidly after cellular inactivation or cell death (Kramer and Singelton 1993; Lee and Kemp 1994). Hence, the RNA-based detection of nonlabeled families indicates that these taxa were metabolically active (Mengoni et al. 2005; Urich et al. 2008). The majority of non-labeled taxa belonged to the phyla Actinobacteria (21.1%), followed by little abundant taxa of Deltaproteobacteria (6.0%), Betaproteobacteria (5.2%) and Alphaproteobacteria (3.9%), Gammaproteobacteria (2.0%), Firmicutes (1.9%), Bacteroidetes (1.4%), Acidobacteria (1.1%), Chloroflexi (0.5%), Verrucomicrobia (0.5%), and Fusobacteria (0.1%, Table 23). Detected non-labeled actinobacterial families can aerobically and anaerobically utilize a wide range of carbon compounds present in the organic fraction of soil (e.g., sugars, starch, pectin) or are able to grow autotrophically (Chen et al. 2004; Evtushenko and Takeuchi 2006; Jackson et al. 1995; Kim et al. 2007; Kim et al. 2008; Lynd et al. 2002; Lu et al. 2010; Stackebrandt and Schumann 2000; Tamura et al. 1999; Wade et al. 1999). Cellulolytic activity within non-labeled Actinobacteria is only reported for Micromonosporaceae (Gallagher et al. 1996; Lynd et al. 2002; Wilson 1992). Families of the phylum Bacteroidetes and of the phylum Deltaproteobacteria include members of the so-called ‘gliding bacteria’, i.e., non-flagellated bacteria that are highly motile due to surface-associated movement (Nett and König 2007). Some gliding bacteria are important degraders of cellulose in soil (Nett and König 2007; Reichenbach 2006) or have a predatory lifestyle (Kersters et al. 2006). Acidobacteria, Firmicutes, and Gammaproteobacteria represent also phyla with cellulolytic members (Humphry et al. 2003; Kim 1995; Kumar et al. 2008a) that can be parts of the rumen ecosystem in which they primarily ferment amino and fatty acids (Cook et al. 1994; van Gylswyk 1995; Janssen and O'Farrell 1999). Cellulolytic Acidobacteria are slowgrowing constituents of the plant-decomposing community in Sphagnum peat, but were likely out-competed by more efficient utilizers of plant-derived carbon, e.g., Cytophagceae species (Eichhorst et al. 2011; Pankratov et al. 2008; Pankratov et al. 2011).
D ISCUSSION 128 The remaining family-level taxa lack described cellulolytic species (Table 23), but some process cellulase genes (Reinhold-Hurek et al. 1993; Robledo et al. 2008), or are parts of the cellulose-degrading community in soil and cellulose-fed enrichments (e.g., Anaerolineaceae, Alcaligenaceae, and ‘Xiphinematobacteriaceae’; Dumova and Kruglov 2009; Haichar et al. 2007; Ishii et al. 2008; Patel and Vaughn 1973). It is supposed that most of these families act as satellite organisms and secondary utilizers of cellulose-derived degradation products (Dumova and Kruglov 2009; Hofsten et al. 1971; Ishii et al. 2008). Only 30% of the ribosomal RNA of an organism has to be labeled until it can be detected by RNA stable isotope probing (Jehmlich et al. 2008). The large number of non-labeled families might indicate that these Bacteria incorporated amounts of [ 13 C]-carbon of less than 30% into their RNA, assimilated soil indigenous carbon, or fed on other bacteria. The latter two assumptions are supported by the enhanced turnover of soil indigenous carbon as reflected in imbalanced carbon recovers under anoxic conditions (4.1) and the physiological capabilities of these family-level taxa, which include predation and saccharide utilization. 4.4. Diversity of Active Eukaryotes The capablitiy to degrade cellulose is also a wide spread trait of microbial Eukarya (1.3.1) and different eukaryotic families were labeled in oxic [ 13 C]-cellulose treatments (Table 24). Although soil fungi are major degraders of cellulosic biomass under oxic conditions (de Boer et al. 2005; Lynd et al. 2002) no fungal species was labeled by [ 13 C]-cellulose, neither under oxic nor under anoxic conditions. The lack of fungi under anoxic conditions is consistent with the previous findings that anaerobic cellulolytic fungi are well-known from the intestine of ruminant and non-ruminant herbivores, but not from soil (Gordon and Phillips 1998; Lynd et al. 2002; Teunissen and DenCamp 1993), and further indicates that fungi played no role under anoxic conditions. Under oxic conditions, a niche differentiation of cellulolytic fungi and cellulolytic bacteria, which is governed by a set of variable factors, e.g., accessibility of cellulose fibres, excretion of antifungal compounds by bacteria, or pH, is likely (Bastian et al. 2009; de Boer et al. 2005; Lynd et al. 2002). Plant material with low lignin content may be more readily accessed by cellulolytic bacteria, and aerobic cellulolytic bacteria may be more competitive at neutral pH and high moisture contents (de Boer et al. 2005; Hiroki and Watanabe 1996). However, the growth of fungi was observed in microcosms with wet soil and cellulosic paper sheets (2.3.3.3) which indicates that fungi were active if the experimental conditions are suitable. Well-known fungi that contribute to the degradation of cellulosic material in aerated soil are related to the phyla Ascomycota (e.g., Aspergillus, Fusarium, Trichoderma), Basidiomycota (e.g., Agaricus, Cryptococcus, Phanerochaete), and Mucoromycotina (e.g., Mucor, Rhizopus) (Bastias et al. 2009; Eriksson 1978; Hammel 1997; Herculano et al. 2011; Lynd et al. 2002; Mandels and Reese 1999; Mahmood et al. 2006). It is likely that soil fungi were participants in the consumption of cellulose in this study, but the experimental conditions in stable isotope probing experiments (i.e., the near neutral pH, the purity of the cellulose, high moisture content, and constant shaking of slurries) have favored cellulolytic bacteria both under oxic and anoxic. Further research is needed to resolve the role of fungi for the degradation of plantderived saccharides in the investigated soil.
D ISCUSSION 129 In contrast to fungal species, protists were labeled in oxic cellulose treatments (i.e., Eustigamataceae, Bodonidae, Chrysophyceae, Opistonectidae, Mallomonadaceae, Stramenopiles; Table 24), but none of the detected taxa has been described as being cellulolytic (Berner and Berner 1993; Hall 1953; Laybourn- Parry and Parry 2000). Cellulolytic protists are frequently detected in gut ecosystems of ruminants and wood-feeding termites in which they catalyze the breakdown of cellulosic plant material (Bohatier et al. 1990; Breznak and Brune 1994; Coleman 1984; Hidayat et al. 1993; Wheeler et al. 2007). For example, Ophryoscolecidaerelated ciliates are important anaerobic degrader of cellulose in the rumen of sheep (Coleman 1978) and important constituents of the cellulolytic eukaryotic community in cattle (Kudo et al. 1990). Taxa detected in the current study (Table 24) did not affiliate with species of cellulolytic protists known from ruminal ecosystems what indicates that the primary oxic nature of the soil did not favor anaerobic protists. Opisthonecta-related species are aerobic heterotrophic free-living ciliates that occupy ephemeral water bodies or moisture films around soil particles (Bochdansky and Huang 2010; Laybourn-Parry and Parry 2000; Williams and Clamp 2007). Members of this genus are bacteriovorus and often found in nutrient rich wastewaters (Ettl 2001; Foissner 1978). Bodonidae are to the most frequently detected heterotrophic flagellates in several ecosystems (Patterson and Lee 2000; Scheckenbach et al. 2006). Members of these genera are frequently associated with aggregates of organic matter, e.g., in activated sludge, on which they graze very effectively on attached bacteria (Laybourn-Parry and Parry 2000; von der Heyden and Cavalier- Smith 2005). Some species, e.g., members of the genus Bodo, are able to survive also at low concentrations of prey and can therefore outcompete other bacteriovorus protists that need higher bacterial densities for multiplication (Laybourn-Parry and Parry 2000; Sleigh 2000; Zubkov 1995). Similar, Chrysophyceae-related flagellates prefer bacteria as energy and carbon source (Caron et al. 1986; Yubuki et al. 2008). Chrysophyceae are important constituents of the marine plankton (Throm 1997), but can also be found in soil samples of different geographical regions (Boenigk et al. 2005). They display an aerobic heterotrophic as well as an autotrophic lifestyle or are mixotrophs, i.e., they combine photoautotrophy with heterotrophy (Sleigh 2000). Members of Chrysophyceae and other protists can be involved in the composting of rice straw and orchard grass, but their role in the degradation process seems to be of minor importance (Cahyani et al. 2004; Lee et al. 2011). The taxa Eustigmataceae (algal stramenopiles) and Leukarachnion (amoeboid stramenopiles) occur in marine and freshwater habitats as well as in soils (Andersen 2004; Grant et al. 2009; Hoek et al. 1995), and include photoautotrophs, and heterotrophs that are osmotrophic (Berner and Berner 1993; Grant et al. 2009). Some stramenopiles are phagotrophic bacterivours, i.e., they feed on living and dead bacteria (Laybourn-Parry and Parry 2000; Sleigh 2000). Members of the detected eukaryotic families require moist zones in soil, but can also form cysts during periods of dryness or nutrient limitation (Hall 1953; Findenig et al. 2010; Laybourn-Parry and Parry 2000). In water-saturated habitats phototrophic microeukaryotes are important primary producers of biomass and drive the carbon flow through marine and freshwater ecosystems (Hall 1953; Throm 1997). Most soil flagellates feed on bacteria or assimilate organic compounds (Ekelund and Rönn 1994; Murase et al. 2006; Rönn et al. 2002), but their role for the degradation of cellulosic material is largely unresolved. It is likely that heterotrophic protists incorporated [ 13 C]-carbon in the experiments during the ingestion of [ 13 C]- labeled saccharide-degrading bacteria.
D ISCUSSION 130 4.5. Pesticides have Minor Effects on the Metabolism of Cellulose-Degrading Communities Soil microorganisms in agricultural soils may not only be grazed by bacteriovorus protists, but also impaired by toxicity of pesticides, such as herbicides, fungicides, insecticides or break down products thereof (Chowdhury et al. 2008; Gevao et al. 2000; Katayama and Kuwatsuka 1991; Katayama et al. 1992; Wainwright 1978). The extensive application of pesticides might alter cellulosedependent carbon flow through the microbial community. Thus, saccharolytic and cellulolytic taxa that were identified as key organisms of the degradation of cellulose and its breakdown products (2.3.3.3) were analysed for their metabolic response on presence and absence of pesticides (Table 28). The fungicides Metalaxyl and Chlorothalonil did apparently not affect the hydrolysis of cellobiose and subsequent consumption of glucose at concentrations 10 – 33fold above the concentrations recommended for agriculture use (Figure 24; 2.3.3.1). Although Metalaxyl may inhibit β-glucosidase activity in soil (Monkiedje and Spiteller 2002; Sukul 2006), and Chlorothalonil reduces cellulose degradation in aerated soils (Katayama and Kuwatsuka 1991; Katayama et al. 1992; Suyama et al. 1993b; Suyama et al. 1993a), the saccharolytic community of the investigated soil is likely insensitive to both fungicides or negative effects are not reflected in changed processes (Figure 24). Metalaxyl and Chlorothalonil can be decomposed by microorganisms (Fogg et al. 2003; Monkiedje et al. 2003), but it is unlikely that they were degraded during the short incubation time (2 – 4 days) of the experiments of this work (2.3.3.1). Half-life times of Metalaxal and Chlorothalonil are 9 to 66 days (Table 5). Hence, the fungicides likely can interact with the microbes in soil slurries. Several reasons that account for the lack of inhibition at such ‘high’ concentrations of pesticides are conceivable: (i) Pesticide-sensitive saccharide-degrading fungi were not favoured under the experimental conditions and bacteria were pesticideinsensitive. (ii) Sorption of Metalaxyl and Chlorothalonil to soil particles lowered their biotoxicity due to reduced interactions with soil microorganisms (Fogg et al. 2003; Katayama et al. 1991; Sigler et al. 2003). (iii) Parts of the saccharide-degrading community were selectively inhibited by the pesticides, but other functionally redundant and pesticide-insensitive taxa mimicked this effect. However, it remains unclear if and how both fungicides impacted the saccharide-degrading community, but the effects of Metalaxyl and Chlorothalonil on the degradation of saccharides seem to be minimal. In contrast, the pesticides Bentazon, MCPA and Nonylphenol reduced the degradation of cellulose and cellulose-derived sugars (Figure 25 – Figure 27, Table 28; 3.5). Aerobic and anaerobic hydrolysis of cellobiose was apparently not reduced by low concentrations of Bentazon, MCPA and Nonylphenol (Figure 25, Table 27), suggesting that these pesticides have minimal effects on the activity of βglucosidases or endoglucanases (Figure 2) at in situ-relevant concentrations. MCPA that is applied at recommended dose does not affect either carbon dioxide production, or O 2 uptake and N-mineralization in a cropland soil, and aerobic cellulose degradation is only little reduced even when MCPA is directly spread on cellulose sheets (Grossbard 1971; Schröder 1979). Although Nonylphenol can also inhibit cell growth of Bacteria in pure culture and reduces the nitrifying-activity in activated sludge (Dokianakis et al. 2006; Okai et al. 2000a), it has no impact on cellulolytic members of the bacterial soil community, e.g., Cytophaga-related species (Chang et al. 2007).
D ISCUSSION 131 In accordance with previous studies (Cernakova et al. 1991; Hseu 2006; Marsh et al. 1978; Piutti et al. 2002), the toxic effect of Bentazon, MCPA, and Nonylphenol on the degradation of cellulose-derived carbon, and linked processes was dosedependent and only evident at concentrations far above values that are typical in pesticide-treated cropland soils (Figure 26, Table 27). The degradation of crystalline cellulose was impaired both under oxic and anoxic conditions, i.e. Bentazon, MCPA, and Nonylphenol reduced the amount of degraded cellulose by about 40 – 60% under oxic conditions and by more than 90% under anoxic conditions (Figure 27, Table 27), indicating that the anaerobic cellulolytic community was highly sensitive to the toxic effects of the pesticides. A similar degree of inhibition has been reported for Oxyfluorfen, Picloram, Metolachlor, and Alachlor under oxic conditions (Kucharski and Wyszkowska 2008; Sahid and Yap 1994; Sahid and Ramil 1995). Application of these pesticides may decrease cellulase activity and cell numbers of cellulolytic bacteria (Sahid and Yap 1994; Sahid and Ramil 1995). Sulfosulfuron inhibits the growth of cellulolytic bacteria and other functional groups even at concentrations that are typical of field use (Kucharski and Wyszkowska 2008). Similar to cellulosesupplemented treatments (Figure 27) anaerobic processes, especially the consumption of glucose, were strongly inhibited by all pesticides at high concentrations (Figure 26, Table 27). This is in congruence with studies that showed that Bentazon can lower the total microbial biomass after application to soil (Piutti et al. 2002), and can reduce nitrogen mineralization, dinitrogen fixation, soil respiration, and the cultivability of cellulolytic soil bacteria, celluloytic fungi and Actinomycetes (Ahtiainen et al. 2003; Marsh et al. 1978, Cernakova et al. 1991). A generally increased sensitivity of anaerobes is likely, since some fungicides may reduce cellulose decomposition and may impair cellulolytic Clostridia as it has been shown for a flooded soil (Katayama and Kuwatsuka 1991). This is consistent with the lower production of carbon dioxide and molecular hydrogen in anaerobic pesticide-supplemented treatments, and the reduction of ferric iron to levels similar to the control without pesticide (Figure 27, Table 27). It is conceivable that Bentazon, MCPA, and Nonlyphenol impacted on membrane- and cytosol-associated proteins, and thus did not only specifically inhibit the activity of cellulolytic and saccharolytic taxa, but also the activity of anaerobes (e.g., iron reducers) that may consume fermentation products of anaerobic cellulose degradation, i.e., organic acids and alcohols. This conclusion correlates with reduced transcript numbers of total soil Bacteria and saccharide-degrading taxa at the end of the experiment (Figure 28, Table 28). It can be speculated that impairment of anaerobic processes in wellaerated soils is of minor importance for the degradation of cellulose since this process is mainly catalyzed by aerobic organisms in oxic zones. In permanently anoxic soil, e.g., flooded rice fields, pesticides may have stronger impacts on the active microbial community and may significantly inhibit the cellulose-linked carbon flow. Development of primers that are specific for Cluster III Clostridiaceae failed (Table 26, 3.4), but Cluster I Clostridiaceae were inhibited by pesticides. A major proportion of known anaerobic cellulolytic bacteria that have been isolated belong to Cluster III Clostridiaceae. Thus, it cannot be excluded that members of Cluster III Clostridiaceae were also impaired and contributed to the effect of reduced degradation of crystalline cellulose. However, the saccharolytic taxon Micrococcaceae and the cellulolytic taxon Cellulomonadaceae were apparently not influenced by the application of pesticides since their RNA content was not lowered at the end of the incubation (Table 28). Actinobacteria are well known for their
D ISCUSSION 132 insensitivity to xenobiotics and their potential to degrade a wide range of pesticides, e.g., benzonitrile herbicides and para-nitrophenol (Chikere et al. 2009; Nielsen et al. 2011; Vesela et al. 2010), and thus, may have contributed to the degradation of cellulose despite the presence of pesticides. The capability of some Bacteria to degrade xenobiotics (e.g., Ekelund et al. 1993; Hseu 2006; Porter and Hayden 2002) is further reflected in the rapid decrease of Nonylphenol under oxic conditions (Figure 27). Although no degradation products of Nonylphenol were detected, the production of carbon dioxide was less reduced in Nonylphenol-supplemented treatments than in the other pesticide-supplemented treatments (Figure 27, Table 27). This suggests that carbon dioxide that results from the degradation of Nonylphenol masked the impact of Nonylphenol on carbon dioxide production that results from degradation of cellulose. Interestingly, growth of fungi on cellulose sheets was observed (data not shown) that suggests that cellulolytic fungi were insensitive to the pesticide toxicity, as it was shown in previous studies (Girlanda et al. 2009; Koroleva et al. 2002; Pampulha et al. 2007). However, the impact of pesticides on fungi in the current study remains speculative. Exposure of the soil organisms to high concentrations of pesticides is possible although when pesticides are applied at recommended rate. Heterogeneous distribution of pesticides in crop fields may lead to local maxima of pesticide concentration in soil that exceed reported mean values (Marsh et al. 1978). Plants can accumulate pesticides in their tissue (Bokern and Harms 1997; Porter and Hayden 2002). Likewise, cellulolytic microorganisms that use these pesticide-treated plants as carbon and energy source can be exposed to higher concentrations as it was originally applied. Nevertheless, current and previous data suggest (i) that the anaerobic saccharide-degrading community is more sensitive to the toxicity of pesticides than the aerobic one and (ii) that the impact of Bentazon, MCPA and Nonylphenol on the saccharide-degrading community is minimal at in situ–relevant concentrations (Hypothesis 3, 1.7). 4.6. Conclusions and Model The results of experiments that were conducted within the doctoral project illustrate the complexity of the cellulose-linked food web. Prokaryotic and eukaryotic taxa that are actively involved in the turnover of carbon in soil slurries of an aerated agricultural soil were identified. Cellulose and cellulose-derived saccharides are utilized by a highly diverse aerobic and anaerobic microbial community, of which a large proportion was represented by new cellulolytic and saccharolytic species (Hypothesis 1, 1.7; Figure 34). Fluctuations of the availability of O 2 did not impact on the capability of soil microbial community to consume supplemented carbon compounds (Hypothesis 2, 1.7; Figure 34). Agricultural soils are often treated with pesticides, which might have impacts on the aerobic and anaerobic degradation of plant-derived saccharides and the involved microbial community (Hypothesis 3, 1.7; Figure 34). The results of the presented experiments do not support this hypothesis. Tested pesticides did not impact on the degradation of saccharides at in situ-relevant concentrations. It is possible that single saccharide-degrading taxa were impacted by pesticides, particular anaerobes, but this effect apparently plays no role for the capacity of the soil microbial community to degrade the supplemented saccharides.
D ISCUSSION 133 Figure 34. Conceptual model of processes related to the degradation of cellulose, major cellulolytic and saccharolytic bacterial taxa, and the impact of pesticides in the presence and absence of O 2 in an aerated agricultural soil under experimental conditions. The gray shaded boxes in the lower portion of the model illustrate different events that occur during the conversion of saccharides to products. The relative amounts of degradation products are indicated by different font sizes. Proposed names of bacterial taxa in quotation marks represent new family-level taxa without cultured representatives (Table 22). Bacterial taxa in brackets have not been previously shown to be cellulolytic, but were labeled in cellulose-supplemented treatments. The tapered box for redox potential indicates that the redox potential became more negative in the anoxic period. The tapered box for high concentrations of pesticides indicates that high concentrations only inhibited the utilization of saccharides under anoxic conditions. Low [Pesticide]: No Effect + O 2 CO 2 Cellulose Cellobiose Glucose Micrococcaceae Intrasporangiaceae Micrococcaceae ‘Sphingo‘ (Bacteroidetes) [‘Deha‘ (Chloroflexi)] [Planctomycetaceae] Consumption of Cellulose, Cellobiose, and Glucose Fermentations and Reduction of NO 3- and Fe 3+ Aerobic Respiration CO 2 , H 2 , Acetate, Propionate CO 2 , H 2 ,Acetate, Propionate, Butyrate, Lactate, Isobutyrate, Succinate ‘Sphingo‘ (Bacteroidetes) ‘Cellu‘ (Bacteroidetes) Clostridiaceae I ‘Sphingo‘ (Bacteroidetes) [Kineosporiaceae] ‘Cellu‘ (Bacteroidetes) Clostridiaceae III Clostridiaceae I ‘High‘ Hydrolysis High [Pesticide] (-) Redox Potential (+) Hydrolysis – O 2
D ISCUSSION 134 Agricultural soils are usually well-aerated soils. Nonetheless, oxic and anoxic microzones co-occur (1.6) in which aerobic and anaerobic microbial activities take place (Figure 3). The collective results of aerobic and anaerobic degradation processes show that cellulose, cellobiose, and glucose are mineralized to carbon dioxide by aerobic respiration under oxic conditions and that the cellulose, cellobiose, and glucose are converted to fermentation products, concomitant with the apparent reduction of nitrate and ferric iron under anoxic conditions (Figure 11, Figure 25 – Figure 27, Figure 30). Acetate is a dominant fermentation product during anaerobic degradation, followed by propionate, butyrate and traces of lactate, isobutyrate and succinate. The detected product spectra correlate well with the identification of different aerobic and anaerobic celluloytic and saccharolytic bacterial taxa and their known metabolic capabilities (Table 22; 4.2, 4.2.2). [ 13 C]-cellulose is mainly degraded by Cluster III Clostridiaceae, and new family-level taxa within Bacteroidetes under anoxic conditions, whereas new family-level taxa of the phyla Bacteroidetes and Chloroflexi are key degraders under oxic conditions. Active sub-communities in [ 13 C]- cellobiose and [ 13 C]-glucose treatments differ from cellulose degrading communities, and are dominated by Cluster I Clostridiaceae under anoxic conditions, whereas Intrasporangiaceae and Micrococcaceae are prevalent in oxic treatments. In addition, the new family-level taxa ‘Sphingo1-4’ and ‘Cellu1-3’ were stimulated by the addition of cellobiose during the anoxic period under fluctuating availabilities of O 2 (Figure 32E and G). However, the role of other active taxa, i.e., Kineosporiaceae, Planctomycetaceae, and the new family-level taxon ‘Deha1’, was not fully resolved. These taxa likely participate in the degradation of plant-derived saccharides in the investigated soil, but it is still unclear at which level they are involved in the celluloselinked food web. In summary, Cluster III Clostridiaceae and Bacteroidetes-related taxa represented the major cellulolytic constituents of the microbial community in slurries of the investigated agricultural soil, whereas Cluster I Clostridiaceae, and the actinobacterial families Micrococcaceae and Intrasporangiaceae were saccharolytic satellite organisms that utilized cellulose-derived hydrolysis products. The rapid response of these microbial taxa altered the redox potential during the degradation of saccharides (Figure 30), but the rate at which plant-derived saccharides are degraded was largely independent of the availability of O 2 . Selective activation of functionally redundant taxa that coexist in oxic and anoxic microzones of aerated agricultural soil likely makes the degradation of cellulose a stable function even when environmental factors change rapidly. Pesticides did apparently not impair the activity of the cellulose degrading-community at in situ-relevant concentrations of pesticides. The anaerobic bacterial community seemed sensitive to pesticides, but significant effects are only observed at very high concentrations that are unlikely to occur in agricultural soil. 4.7. Limitation of the Applied Methods RNA stable isotope probing was used as a method to select for bacterial taxa that actively consumed [ 13 C]-cellulose and cellulose-derived hydrolysis products, i.e., cellobiose and glucose, in experiments performed with soil from an agricultural ecosystem. Although these saccharides were applied at concentrations greater than maximum concentrations detected in soils (Hill et al. 2008; Medeiros et al. 2006), they represent a compromise between low in situ concentrations and concentrations that are needed for detectable incorporation of [ 13 C] in the rRNA pool. The choice of
D ISCUSSION 135 gradient fractions that are representative for labeled and unlabeled RNA was based on the distribution of RNA in the gradient (Figure 13). The ‘heavy’ fractions are enriched for [ 13 C]-RNA, but it is likely that they contain unlabeled RNA (Manefield et al. 2002a; Manefield et al. 2002b; Radajewski et al. 2003). To ensure that detected taxa of these ‘heavy’ fractions were isotopically labeled, tRFLP analyses and the construction of gene libraries were additionally applied. It cannot be excluded that the study missed single cellulolytic and saccharolytic soil taxa due to the constrictive type of data analysis. Some tRFs could not be identified and some active taxa might have also been present in non-analysed fractions. Nevertheless, the identification of 48 active family-level taxa and the dominance of well-characterized cellulolytic and saccharolytic microorganisms (e.g., Cluster I and III Clostridiaceae or Micrococcaceae) in tRFLP profiles and gene libraries (Table 22; 3.2.3.2.1, 3.2.3.2.2) indicates that the majority of the cellulose-degrading community was covered by the applied way of data analysis. It was possible to monitor community dynamics during the degradation of saccharides (Table 19, Table 20). Although labeling of some taxa at a later time point due to assimilation of [ 13 C]-carbon dioxide cannot be excluded, cross-feeding should have been minimal as the gaseous products were periodically removed and growth of photoautotrophic organisms was prevented by dark incubation. Taxa that were labeled at later time points in cellulose treatments might have needed longer for the competitive utilization of cellulose or might have utilized soluble hydrolysis products from cellulose, e.g., sugars or fatty acids (Bayer et al. 2006). The degradation of [ 13 C]-enriched wheat straw, rice callus, or cellulose by agricultural soil communities also reveals [ 13 C]-labeling of microbial taxa after prolonged incubation of which is not known if they represent cellulolytic taxa (Bernard et al. 2007; Lee et al. 2011; Li et al. 2009). Stable isotope probing can shed light on microbial food webs in various habitats (e.g. Hamberger et al. 2008; Hunger et al. 2011; Murell and Radajewski 2000; Wüst et al. 2009), but cross-feeding complicates interpretation of labeling patterns, especially for newly discovered taxa with unknown metabolic potentials. Therefore, the labeling of known and new cellulolytic and saccharolytic taxa was qualified by taxon-specific qPCR approaches in the current study (3.4). The degradation of CMC does not necessarily require endoglucanases what makes CMC a substrate that can also be used by saccharolytic organisms (Bisaria and Ghose 1981). For example, Pantoea species lack endoglucanse-encoding genes and utilizes CMC, but not crystalline cellulose (Adams et al. 2011). However, the degree of substitution (DS), i.e., the number of hydroxyl groups that are esterified with carboxymethyl groups per glucose unit, of the used CMC was high (i.e., 0.65 - 0.9). The higher the DS of CMC is, the more difficult the substrate is to degrade and the more similar it is to crystalline cellulose (Reese et al. 1950). Therefore, the used form of CMC was a suitable analog for crystalline cellulose. Taking all these considerations into account it can be concluded that the identified cellulolytic and saccharolytic taxa are likely key microorganism for the degradation of cellulose and cellulose-derived carbon in the investigated agricultural soil.