Microbiologyofphaseseparatedreactorsystems
forbiomethanation
athightemperatures(5575°C)
VorgelegtvonDipl.Biol.AntjeRademacher
VonderFakultätIIIProzesswissenschaften
derTechnischenUniversitätBerlin
zurErlangungdesakademischenGrades
DoktorderNaturwissenschaften
Dr.rer.nat.
genehmigteDissertation
Promotionsausschuss:
Vorsitzender: Prof.Dr.‐Ing.Sven‐UweGeißen
Berichter: Prof.Dr.UlrichSzewzyk
Berichterin: Prof.Dr.ElisabethGrohmann
Berichter: Dr.MichaelKlocke
TagderwissenschaftlichenAussprache:13.06.2013
Berlin2013
D83
II
III
Die vorliegende Arbeit wurde am Leibniz-Institut für Agrartechnik Potsdam-Bornim e.V.
(Abteilung Bioverfahrenstechnik) unter der Anleitung von Herrn Dr. M. Klocke in der
Zeit von 2009 bis 2012 erstellt.
Gefördert wurde diese Arbeit vom Bundesministerium für Bildung und Forschung
(BMBF) durch den Projektträger Jülich (Förderkennzeichen 03SF0349C).
IV
DANKSAGUNG
Im Folgenden möchte ich mich herzlich bei den Menschen bedanken, die mich
während meines Promotionsvorhabens begleitet und unterstützt haben.
Allen voran möchte ich mich bei Herrn Dr. Michael Klocke für seine wissenschaftliche
Betreuung und Unterstützung bedanken. Er war mir stets Ansprechpartner und
bereicherte meine Arbeit mit Ideen, Anregungen und konstruktiver Kritik.
Dank gilt auch Herrn Prof. Dr. Ulrich Szewzyk für die Begleitung meines
Promotionsvorhabens, die Anregungen zu meiner Arbeit sowie für deren
Begutachtung.
Ebenso gilt mein Dank Frau Prof. Dr. Elisabeth Grohmann für ihre Anregungen und die
konstruktive Kritik zu meiner Arbeit sowie für die Übernahme des Gutachtens.
Mein herzlicher Dank gilt Dipl.-Ing. Mandy Schönberg und Dipl.-Ing. Carsten Joost für
die gute Zusammenarbeit im BMBF-Projekt, für den Aufbau und Betrieb der
Versuchsanlage und die Bereitstellung von Versuchsdaten für die weitere Auswertung.
Des Weiteren möchte ich den Mitarbeiterinnen und Mitarbeitern der AG Analytik des
ATBs für die Analyse der Reaktorproben danken, sowie Kerstin Mundt für ihre
kompetente Unterstützung im Labor. Zusätzlich möchte ich den vielen weiteren
Kolleginnen und Kollegen des ATBs, insbesondere der Abteilung 1, für die angenehme
Arbeitsatmosphäre danken. An dieser Stelle seien Angelika H., Antje F., Beate K.,
Edith N., Ingo Ba., Ingo Be., Johanna K., Kay B., Kathrin H., Katrin G., Lena H.,
Matthias B., Sarah H., Susanne K. und Susanne T. namentlich genannt. Vielen Dank
für eure Unterstützung, die fruchtvollen Gespräche und die netten Stunden.
Mein weiterer Dank gilt Herrn Dr. Andreas Ulrich für seine Hilfe und Unterstützung bei
der Einführung der TRFLP-Methodik am ATB.
Nicht zuletzt gilt mein besonderer Dank meinen Eltern sowie meinem Freund für die
andauernde Unterstützung, Geduld und den notwendigen Rückhalt während der
letzten Jahre.
V
EIDESSTATTLICHEERKLÄRUNG
Ich erkläre an Eides Statt, dass die vorliegende Dissertation in allen Teilen von mir
selbständig angefertigt wurde und die benutzten Hilfsmittel vollständig angegeben
worden sind. Veröffentlichungen von irgendwelchen Teilen der vorliegenden
Dissertation sind von mir, wie umseitig dargelegt, vorgenommen worden.
Weiter erkläre ich, dass ich nicht schon anderweitig einmal die Promotionsabsicht
angemeldet oder ein Promotionseröffnungsverfahren beantragt habe.
__________________________ ___________________________
Ort, Datum Unterschrift
VI
TeiledieserArbeitwurdenbereitswiefolgtveröffentlicht:
Rademacher A, Zakrzewski M, Schlüter A, Schönberg M, Szczepanowski R,
Goesmann A, Pühler A & Klocke M (2012) Characterization of microbial biofilms in a
thermophilic biogas system by high-throughput metagenome sequencing. FEMS
Microbiology Ecology 79, 785-799, DOI: 10.1111/j.1574-6941.2011.01265.x
Rademacher A, Nolte C, Schönberg M & Klocke M (2012) Temperature increases
from 55 to 75 °C in a two-phase biogas reactor result in fundamental alterations within
the bacterial and archaeal community structure. Applied Microbiology and
Biotechnology 96, 565-576, DOI: 10.1007/s00253-012-4348-x
Eikmeyer F, Rademacher A, Hanreich A, Hennig M, Jaenicke S, Maus I, Wibberg D,
Zakrzewski M, Pühler A, Klocke M & Schlüter A (2013) Detailed analysis of
metagenome datasets obtained from biogas-producing microbial communities residing
in biogas reactors does not indicate the presence of putative pathogenic
microorganisms. Biotechnology for Biofuels 6, 49, DOI:10.1186/1754-6834-6-49
Rademacher A, Hanreich A, Bergmann I & Klocke M (2012) Black-Box-Biogasreaktor -
Mikrobielle Gemeinschaften zur Biogaserzeugung. Biospektrum 18, 727-729,
DOI:10.1007/s12268-012-0253-1
Hanreich A, Rademacher A & Klocke M (2012) Mikrobielles Leben im Biogasreaktor –
Einblicke in einen komplexen und dynamischen Mikrokosmos. In Biogas Potenziale -
Erkennen, Erforschen, Erwirtschaften, p. 114-123. Potsdam, Leibniz-Institut für
Agrartechnik Potsdam-Bornim e.V.
BMBF-Schlussbericht (2013) Bioprozesstechnik der Hydrolyse: Verfahrenstechnische
Optimierung des Bioleaching-Verfahrens unter thermophilen bis hyperthermophilen
Bedingungen und Analyse der hydrolytischen Mikroflora (Förderkennzeichen
03SF0349C). eds. Schönberg M, Rademacher A, Klocke M & Linke B, Hannover,
TIB/UB.
VII
ZUSAMMENFASSUNG
In den letzten Jahren nahm die Zahl an landwirtschaftlichen Biogasanlagen in
Deutschland stetig zu. Im Jahr 2011 wurden 7.215 Biogasanlagen mit einer installierten
elektrischen Leistung von 2.904 MW betrieben. Thermophile Temperaturen sowie eine
räumliche Trennung der Prozessphasen Hydrolyse und Acidogenese von der
Methanogenese können gegenüber den häufig verwendeten, mesophilen, einphasigen
Rührkesselreaktoren den Biogasprozess verbessern und stabilisieren. Für eine weitere
Optimierung solcher Anlagen ist ein detailliertes Wissen über die bakterielle und
archaeelle Gemeinschaft, die am Abbau der Biomasse und der nachfolgenden
Methanproduktion beteiligt sind, unabdingbar.
In dieser Studie wurden deshalb drei identische Leach-bed Biogasreaktorsysteme mit
räumlich getrennten Prozessphasen untersucht, die jeweils mit Roggen-
Ganzpflanzensilage und Stroh mit einer Verweilzeit von 21 Tagen betrieben wurden.
Jedes Reaktorsystem bestand aus einem Leach-bed Reaktor (LBR), dessen
Temperatur schrittweise von 55 auf 75 °C erhöht wurde, einem
Prozessflüssigkeitsspeicher und einem nachgeschalteten Anaerobfilter (AF), der
konstant während des gesamten Versuchs bei 55 °C betrieben wurde. Verschiedene
kultivierungsunabhängige Methoden wurden genutzt um die mikrobielle Gemeinschaft
im Reaktorsystem zu charakterisieren und zu quantifizieren sowie deren Dynamik zu
verfolgen. Neben der Analyse von Genbibliotheken, TRFLP Fingerprintanalysen,
Metagenomanalysen und qPCR Analysen wurden die Proben auch mittels Fluoreszenz
in situ Hybridisierung sowie DAPI- und Propidiumiodid-Färbung untersucht.
Die Ergebnisse der Studie zeigten, dass die beabsichtigte Phasentrennung zwischen
der Hydrolyse und Acidogenese im LBR und der Methanogenese im AF durch die
Etablierung von spezialisierten Gemeinschaften zumindest teilweise erreicht wurde. Im
Vergleich zum AF wurden im LBR verstärkt hydrolytische Bakterien und somit ein
erhöhtes genetisches Potential zum Abbau von Kohlenhydraten festgestellt. Während
einer 21-tägigen Fermentation bei konstanter Temperatur zeigte diese bakterielle
Gemeinschaft dynamische Veränderungen, die mit Konzentrationsänderungen der
Fermentationsprodukte und somit dem Grad des Biomasseabaus einhergingen.
Auch bei einer schrittweisen Erhöhung der Temperatur im LBR kam es zu
Veränderungen in der bakteriellen Gemeinschaft. Bis zu einer Reaktortemperatur von
VIII
65 °C dominierten Vertreter der Clostridia. Ab 70 °C kam es zu einer verstärkten
Ansiedlung von Vertretern der Klassen Bacteroidia, Thermotogae und Bacilli. Diese
scheinen somit vor allem bei hyperthermophilen Temperaturen wichtig für den
anaeroben Abbau von Biomasse zu sein. Entsprechende Veränderungen zeigten sich
auch bei der Analyse des genetischen Potentials für den anaeroben Abbau von
Biomasse. Ab einer Reaktortemperatur von 70 °C veränderte sich das genetische
Potential bestimmte polysaccharidabbauende Glycosidhydrolasen zu exprimieren.
Parallel dazu reduzierte sich die Abbaurate der eingebrachten Biomasse deutlich.
Mittels einer Anreicherung mit Kompost konnten leichte Verbesserungen in der
Reaktorleistung bei 70 °C erzielt werden, die jedoch nicht von Dauer waren. Um eine
langanhaltende positive Veränderung erzielen zu können, scheint somit eine
kontinuierliche Anreicherung notwendig zu sein.
Die methanogene Gemeinschaft im AF zeigte nur leichte Variationen in ihrer
Zusammensetzung während der Temperaturerhöhung im LBR. Sowohl Vertreter der
Methanobacteriales als auch der Methanosarcinales wurden identifiziert. Es ist somit
wahrscheinlich, dass verschiedene methanogene Stoffwechselwege, wie die
acetoklastische und hydrogenotrophe Methanogenese, im AF parallel ablaufen.
Des Weiteren wurde das Reaktorsystem hinsichtlich seines potentiellen Risikos,
Wachstum und Verbreitung von Pathogenen zu gewährleisten, untersucht. Die hierfür
untersuchten Metagenome zeigten jedoch, dass das Risiko für die Verbreitung von
potentiellen Pathogenen durch die Ausbringung von Gärresten auf landwirtschaftlich
genutzte Ackerflächen als sehr gering angesehen werden kann.
Die Studie gibt Aufschluss über Struktur und Dynamik der bakteriellen und archaeellen
Gemeinschaft sowie über das genetische Potential zum anaeroben Abbau von
pflanzlichen Polysacchariden bei hohen Temperaturen in zweiphasigen Leach-bed
Biogassystemen. Insgesamt wurde die beste Reaktorleistung bei Temperaturen von
55 bis 60 °C festgestellt, ein Temperaturbereich, der sich vorteilhaft auf die mikrobielle
Gemeinschaft im Reaktorsystem auswirkt. Des Weiteren zeigen die Ergebnisse
potentielle prozessrelevante Bakterien sowie Glycosidhydrolasen auf, die als
Biomarker für eine Überwachung von weiteren thermophilen Biogassystemen genutzt
werden können. Die Ergebnisse dieser Studie dienen somit als Grundlage für eine
Prozessüberwachung und eine zukünftige Optimierung von thermophilen
Biogasprozessen mit Prozessphasentrennung.
IX
ABSTRACT
In recent years, the number of agricultural biogas plants, as a means of generation of
renewable energy, has risen constantly in Germany. In 2011, 7,215 agricultural biogas
plants were operated with a total installed electric capacity of 2,904 MW.
Thermophilic temperatures and a spatial separation of the process phases hydrolysis
and acetogenesis from methanogenesis are known strategies for improving and
stabilizing biogas production. A deep understanding of the underlying bacterial and
archaeal community involved in the breakdown of plant-derived biomass and the
subsequent production of methane in phase-separated, thermophilic systems is of
major importance for further improvement.
Pursuant to this aim, phase-separated leach-bed biogas systems, which were supplied
with rye silage and straw lasting for 21 days, were analyzed. Each system consisted of
a leach-bed reactor (LBR), whose temperature was increased stepwise from 55 to
75 °C, a leachate storage reactor and a downstream anaerobic filter reactor (AF),
whose temperature remained at 55 °C throughout the experiment. Various culture-
independent methods were used for the characterization, quantification and monitoring
of the microbial community within these biogas systems. The culture-independent
methods were based on the genetic information of cells, applying gene library
construction, TRFLP fingerprinting, metagenomic and qPCR analyses and on the
microscopical quantification of intact, but not cultivated cells.
The results indicated that the intended spatial separation of process phases, such as
hydrolysis and acidogenesis, performed by different hydrolytic and fermentative
bacteria, from the methanogenesis phase, performed by methanogenic archaea, was
achieved at least in part within these biogas systems. For instance, hydrolytic bacteria
and hence the genetic potential to degrade carbohydrates were strongly increased in
the LBR in comparison to the AF.
In addition, changes in the bacterial community were detected in the LBR during the
21 days of fermentation as a consequence of changes in the VFA concentration due to
the anaerobic digestion process.
Further, the stepwise increase in the fermentation temperature of the LBR also led to
alterations within the bacterial community. Above 65 °C, the community changed from
X
being Clostridia-dominated toward being dominated by members of the Bacteroidia,
Clostridia, Thermotogae and Bacilli. These groups seemed to be important for the
anaerobic degradation of plant-derived biomass at hyperthermophilic temperatures. In
addition to these results, the genetic potential for the expression of glycoside
hydrolases, enzymes catalyzing the hydrolysis of glycosidic linkages of carbohydrates,
was also affected by the temperature and changed strongly at an LBR temperature of
70°C.
Simultaneously with the changes in the bacterial community at 70 °C, the reactor
performance also decreased strongly. A bioaugmentation with compost at this
temperature led to slight improvements in the reactor performance, which did not
persist at 75 °C. This indicated that a permanent positive effect of bioaugmentation can
possibly only be realized by a continuous application.
The methanogenic community in the AF showed slight alterations during temperature
increase in the LBR, which was most likely affected by the changes in the intermediate
production of the bacterial community. The monitored archaeal community was mainly
composed of members of the Methanobacteriales and Methanosarcinales. This
indicated that different pathways for methane production, such as the acetoclastic and
hydrogenotrophic pathway, occurred simultaneously in the AF.
Furthermore, samples of both the LBR and the AF were analyzed to assess the
potential risk for the growth of pathogens in such biogas systems. It was shown that the
operation of phase-separated thermophilic biogas systems presents only a very low
risk for an unintended proliferation of putative pathogens and hence for a potential
infection of humans, animals or plants through the application of digestate on farmland.
The present study revealed the composition and dynamics of the microbial community
and their genetic potential for carbohydrate degradation in two-phase leach-bed biogas
systems at thermophilic to hyperthermophilic temperatures. Temperatures of 55 to
60 °C in the LBR had a positive effect on the microbial community responsible for the
production of biogas, leading to the best reactor performance. Furthermore, the results
indicated potentially process-relevant bacteria and glycoside hydrolases, which may
serve as target for the monitoring of thermophilic biogas reactors in future. Hence, the
results gained in this study provide a promising basis for the monitoring and the
prospective improvement of thermophilic biogas systems with phase-separation.
XI
TABLEOFCONTENTS
1 INTRODUCTION ..................................................................................................... 1
1.1 Renewable energy sources ............................................................................... 1
1.2 Agricultural biogas production ......................................................................... 3
1.3 The biogas production process ........................................................................ 5
1.3.1 Anaerobic digestion of plant-derived biomass .................................................. 5
1.3.2 Production of precursors to methanogenesis ................................................... 8
1.3.3 The production of methane ............................................................................... 9
1.4 Purported proliferation of pathogens in biogas systems ............................ 11
1.5 Characterization and monitoring of the microbial community .................... 13
1.5.1 Analysis of bacterial and archaeal rrs gene sequences ................................. 13
1.5.2 Terminal restriction fragment length polymorphism ........................................ 15
1.5.3 Analysis of microbial metagenomes by high-throughput DNA sequencing .... 18
1.5.4 Quantification of bacterial rrs genes by quantitative PCR .............................. 21
1.5.5 Microscopical analyses ................................................................................... 22
1.6 The Aims of this study ..................................................................................... 23
2 MATERIAL AND METHODS ................................................................................. 25
2.1 The two-phase leach-bed biogas systems .................................................... 25
2.1.1 Reactor setup and operation .......................................................................... 25
2.1.2 Analysis of process parameters ...................................................................... 27
2.1.3 Reactor sampling ............................................................................................ 28
2.2 Extraction and quantification of genomic DNA ............................................. 29
2.2.1 DNA extraction protocols ................................................................................ 29
2.2.2 DNA purity and quantification ......................................................................... 32
2.3 Construction of rrs gene libraries .................................................................. 33
2.4 Terminal restriction fragment length polymorphism analysis ..................... 35
2.5 Metagenomic analysis applying 454-pyrosequencing technology ............. 39
2.6 Quantitative real-time PCR .............................................................................. 40
2.7 Microscopical analyses ................................................................................... 43
XII
3 RESULTS ............................................................................................................. 47
3.1 Operation of the two-phase leach-bed biogas systems .............................. 47
3.2 Analysis of bacterial and archaeal rrs gene sequences .............................. 52
3.2.1 Bacterial community composition in the LBR ................................................. 52
3.2.2 Archaeal community composition in the AF ................................................... 53
3.2.3 Statistical analyses of the rrs gene libraries ................................................... 54
3.3 Characterization and monitoring of the microbial community by TRFLP
analyses ........................................................................................................... 56
3.3.1 Establishment and optimization of TRFLP assays for bacterial and archaeal
community analyses ...................................................................................... 56
3.3.2 Identification of terminal restriction fragments ............................................... 61
3.3.3 Similarity of the bacterial community in the three biogas systems ................. 64
3.3.4 Impact of the substrate-attached bacterial community on the bacterial
community in the biogas reactor .................................................................... 65
3.3.5 Changes in the bacterial community in the LBR during the fermentation of one
load of substrate ............................................................................................ 66
3.3.6 Impact of temperature increase on the bacterial community in the LBR ....... 69
3.3.7 Impact of temperature increase on the bacterial community in the AF .......... 73
3.3.8 Changes in the archaeal community in the AF during the fermentation of one
load of substrate ............................................................................................ 74
3.3.9 Impact of temperature increase on the archaeal community in the AF .......... 75
3.3.10 Impact of temperature increase on the archaeal community in the LBR ....... 76
3.4 454-pyrosequencing of microbial metagenomes ......................................... 77
3.4.1 Phylogenetic assignment of metagenomic sequences to Bacteria ................ 79
3.4.2 Genetic potential for the degradation of plant-derived biomass ..................... 84
3.4.3 Phylogenetic assignment of metagenomic sequences to selected pathogens ..
....................................................................................................................... 89
3.4.4 Phylogenetic assignment of metagenomic sequences to Archaea ................ 92
3.5 Quantification of bacterial rrs genes ............................................................. 94
3.5.1 Specificity of the primer set 304 ..................................................................... 95
3.5.2 Quantification of Bacteria and a putatively process-relevant bacterium ........ 97
3.6 Microscopical analyses of the microbial biogas community ...................... 99
XIII
3.6.1 Fluorescence in situ hybridization and total cell count analyses .................... 99
3.6.2 Propidium iodide analysis ............................................................................. 101
4 DISCUSSION ....................................................................................................... 102
4.1 Methodical aspects ........................................................................................ 102
4.1.1 DNA extraction .............................................................................................. 102
4.1.2 TRFLP analysis ............................................................................................ 103
4.1.3 Metagenomic and bioinformatic analysis ...................................................... 106
4.1.4 Quantification of microorganisms ................................................................. 107
4.2 Operation of two-phase leach-bed biogas systems ................................... 109
4.2.1 Performance of the biogas systems during temperature increase ............... 109
4.2.2 Phase-separation of the two-phase biogas system ...................................... 110
4.3 Bacterial community in the two-phase biogas system ............................... 111
4.3.1 Dynamic changes in the bacterial community in the LBR during the 21-day
fermentation .................................................................................................. 112
4.3.2 Composition and dynamics of the bacterial community in the LBR during
temperature increase .................................................................................... 113
4.4 Genetic potential for anaerobic digestion of plant-derived biomass ........ 116
4.5 Putative pathogens in the two-phase biogas system ................................. 119
4.6 Archaeal community in the two-phase biogas system ............................... 121
4.6.1 Composition of the archaeal community in the LBR during temperature
increase ........................................................................................................ 121
4.6.2 Composition of the archaeal community in the AF ....................................... 122
4.7 Syntrophic interactions of VFA-oxidizing bacteria and methanogens ..... 124
4.8 Determination of cell densities of Bacteria and Archaea ........................... 125
5 OUTLOOK ........................................................................................................... 129
6 REFERENCES .................................................................................................... 131
7 APPENDIX........................................................................................................... 149
XIV
LISTOFFIGURES
Figure 1.1 Proportion of renewable energy based on the total energy consumption in
Germany 2011 ........................................................................................... 2
Figure 1.2 Scheme of the anaerobic digestion process .............................................. 9
Figure 1.3 Short workflow of the sample preparation for the TRFLP analysis .......... 16
Figure 1.4 Short workflow of the 454-pyrosequencing analysis ............................... 20
Figure 2.1Scheme of the thermophilic two-phase leach-bed biogas reactor system ..
................................................................................................................ 26
Figure 2.2 Experimental runs of the three two-phase biogas systems and sampling of
the first biogas system ............................................................................. 29
Figure 3.1 Concentration of total VFA (A), acetic acid (B), n-butyric acid (C),
propionic acid (D), valeric acid (E) and ethanol and propanol as sum (F)
shown for the first leach-bed reactor system during the 21-day
fermentation process at all specified temperature regimes ..................... 50
Figure 3.2 Results of the bacterial rrs gene sequence analyses of the LBR at 55 and
75 °C ....................................................................................................... 52
Figure 3.3 Results of the archaeal rrs gene sequence analyses of the packing’s
biofilm derived from the AF at an LBR temperature of 55 °C .................. 54
Figure 3.4 Bacterial TRFLP profiles obtained by four different DNA extraction
protocols .................................................................................................. 57
Figure 3.5 TRFLP profiles obtained by the use of different restriction enzymes and
their combinations ................................................................................... 59
Figure 3.6 Triplicate TRFLP samples indicated with standard deviation .................. 60
Figure 3.7 Comparison of TRF profiles based on TRF area and height ................... 61
Figure 3.8 Impact of the substrate-attached bacterial community on the bacterial
community in the biogas system ............................................................. 65
Figure 3.9 Bacterial community dynamics and VFA concentration during the 21-day
fermentation of one load of substrate at LBR temperatures of 55 °C (A),
60 °C (B), 65 °C (C), 70 °C (D) and 75 °C (E) ......................................... 68
Figure 3.10 Impact of temperature increases in the LBR (55 - 75 °C) on the bacterial
community composition in the LBR ......................................................... 70
Figure 3.11 TRFLP profiles (A) and NMDS plot of TRFLP profiles (B) obtained from
the three biogas reactors at LBR temperatures of 65 and 70 °C ............. 71
XV
Figure 3.12 Direct response of the bacterial community in the LBR to the temperature
increase of 70 °C ...................................................................................... 72
Figure 3.13 Bacterial community attached to the AF packings during temperature
increases in the LBR from 55 to 75 °C ..................................................... 73
Figure 3.14 Changes in the archaeal community composition during the experimental
runs .......................................................................................................... 75
Figure 3.15 Phylogenetic assignment of metagenomic sequences to Bacteria using
the RDP classifier (grey) and CARMA (black) showing the prevalent
bacterial classes ....................................................................................... 80
Figure 3.16 Phylogenetic assignment of metagenomic sequences to Bacteria using
CARMA (A) and the RDP classifier (B) showing the prevalent bacterial
genera ...................................................................................................... 81
Figure 3.17 Phylogenetic assignment of EGTs encoding carbohydrate degrading
enzymes as revealed by CARMA ............................................................ 83
Figure 3.18 Differences in GH family abundances between the metagenomic samples
P 55, D 65 and D 70 and the digestate sample derived from the LBR at
55 °C (D 55) ............................................................................................. 87
Figure 3.19 Phylogenetic assignment of pEGTs to selected plant (1) and human or
animal pathogens (2) as revealed by CARMA ......................................... 89
Figure 3.20 Phylogenetic assignment of metagenomic sequences to Archaea using
CARMA (black) and the RDP classifier (grey) showing the prevalent
archaeal orders ........................................................................................ 92
Figure 3.21 Phylogenetic assignment of metagenomic sequences to Archaea using
CARMA (A) and the RDP classifier (B) showing the prevalent archaeal
genera ...................................................................................................... 93
Figure 3.22Phylogenetic assignment of EGTs encoding methanogenic enzymes as
revealed by CARMA ................................................................................. 94
Figure 3.23 Quantification of Bacteria and bacteria, resulting in TRF 304, by qPCR
analysis .................................................................................................... 98
Figure 3.24 Total cell counts and DOPE-FISH results during temperature increase (A),
percentage of cells stained with PI during 21 days of fermentation at
65 °C (B) ................................................................................................ 100
Figure 7.1 NMDS of the bacterial TRFLP profiles obtained for samples of the 21-day
fermentation at LBR temperatures of 55 °C (A), 60 °C (B), 65 °C (C),
70 °C (D) and 75 °C (E) ......................................................................... 149
XVI
LISTOFTABLES
Table 2.1 Primers and probes used for the polyphasic analyses of the two-phase
biogas systems ........................................................................................ 46
Table 3.1 Analytical parameters (ODM, pH and COD) of substrates and digestates
................................................................................................................ 48
Table 3.2 Process parameters of the three two-phase biogas systems .................. 51
Table 3.3 Biogas and methane yield and degradation rate of the three two-phase
biogas systems ........................................................................................ 51
Table 3.4 Statistical parameters of the archaeal and bacterial rrs gene libraries .... 55
Table 3.5 Bray-Curtis similarity values of the bacterial TRFLP profiles of four DNA
extraction methods .................................................................................. 56
Table 3.6 Affiliation of archaeal (A) and bacterial (B) TRFs according to the NCBI
and RDP taxonomy ................................................................................. 63
Table 3.7 Similarity indices for the bacterial TRFLP profiles of digestate samples
taken from the three biogas reactors ....................................................... 64
Table 3.8 454-pyrosequencing parameters of four different samples derived from
the two-phase biogas system .................................................................. 77
Table 3.9 Number of metagenomic sequences and the number of sequences
assigned to taxonomic groups applying the RDP classifier and CARMA 78
Table 3.10 Distribution of phylogenetic assignments of rrs sequences and EGTs to
taxonomic groups applying the RDP classifier and CARMA ................... 79
Table 3.11 Pfam analysis of lignin degrading enzymes applying CARMA ................ 84
Table 3.12 Most prevalent GH families of the metagenomic samples as derived from
the Pfam analysis by CARMA ................................................................. 85
Table 3.13 Assignment of EGTs matching with toxin-associated Pfam families ....... 91
Table 3.14 Uncultured or environmental sequences retrieved from the primer BLAST
search against the NCBI nr database (last access November 2012)
showing no mismatch to the primer set 304 ............................................ 97
Table 7.1 Accession numbers of protein families relevant for carbohydrate
degradation (A) and methanogenesis (B) obtained from the Pfam
database ................................................................................................ 150
Table 7.2 Accession numbers of protein families relevant for pathogenicity of
selected pathogens obtained from the Pfam database ......................... 151
XVII
ABBREVIATIONS
16S rRNA 16S ribosomal RNA of the
prokaryotic 30S small subunit
ADP Adenosine diphosphate
AF Anaerobic filter reactor
ARDRA Amplified ribosomal DNA
restriction analysis
ATB Leibniz Institut für Agrartechnik
Potsdam-Bornim e.V. (Leibniz
Institute for Agricultural
Engineering Potsdam-Bornim)
ATP Adenosine triphosphate
BFR Bundesanstalt für
Risikobewertung (Federal
Institute for Risk Assessment)
BLAST Basic Local Alignment Search
Tool
bp Base pair
BSA Bovine serum albumin
C Compost sample
Camp. Genus Campylobacter
Cl. Genus Clostridium
C/N Carbon/nitrogen ratio
CaCl2 Calcium chloride
CARD-FISH Catalyzed reporter deposition
fluorescence in situ
hybridization
CARMA Software pipeline for the
characterization of
metagenomic sequences
CAZY Carbohydrate-active enzymes
database
CeBiTec Center for Biotechnology,
Bielefeld University,
Germany
CO2 Carbon dioxide
CoA Coenzyme A
COD Chemical oxygen demand
CoM Coenzyme M
CrK(SO4)2 Chromium potassium sulfate
CSTR Continuously stirred tank
reactor
Ct Threshold cycle number
CTAB Hexadecyltrimethylammonium
bromide
Cy5 Cyanine 5
D Digestate sample
DAPI 4′,6-diamidino-2-phenylindole
DGGE Denaturing gradient gel
electrophoresis
DNA Deoxyribonucleic acid
dNTP Deoxyribonucleotide
triphosphates
DOPE-FISH Double labeling of
oligonucleotide probes for
fluorescence in situ
hybridization
dsDNA Double strand DNA
DSMZ Deutsche Sammlung von
Mikroorganismen und
Zellkulturen GmbH (German
Collection of Microorganisms
and Cell Cultures)
E. Genus Escherichia
EDTA Ethylenediaminetetraacetic
acid
EEG Erneuerbare-Energien-Gesetz
(Renewable energy law)
XVIII
EGT Environmental gene tag
EHEC Enterohemorrhagic E. coli
strain
FAM 6-carboxyfluorescein
FISH Fluorescence in situ
hybridization
FNR Fachagentur nachwachsende
Rohstoffe (Agency for
Renewable Resources)
FP FastDNA® Spin Kit for Soil
(x) g Acceleration of gravity
Gb Gigabase
GC Guanine and cytosine
GH Glycoside hydrolase
GOLD Genome OnLine Database
GS Genome Sequencer
H2 Hydrogen
H2O Water
H2O2 Hydrogen peroxide
HAc eq. Acetic acid equivalents
HCl Hydrogen chloride
hPa Hectopascal
HPLC-H2O High-performance liquid
chromatography water
IPTG Isopropyl β-D-1-thiogalacto-
pyranoside
KCl Potassium chloride
KEGG Kyoto Encyclopedia of Genes
and Genomes
KH2PO4 Monopotassium phosphate
L 0 - 21 Leachate sample at different
time points of fermentation
L. Genus Listeria
LB Lysogeny broth
LBR Leach-bed reactor
LOD Limit of detection
LOQ Limit of quantification
LR Leachate reservoirNa
M molar (mol L-1)
Mcu. Genus Methanoculleus
Msr. Genus Methanosarcina
Mtb. Genus Methanothermobacter
Mb Megabase
MgCl2 Magnesium chloride
NA Not analyzed
Na2HPO4 Sodium hydrogen phosphate
NaCl Sodium chloride
NaOH Sodium hydroxide
NCBI National Center for
Biotechnology Information
NH3 Free ammonia
NH4- N Total ammonia nitrogen
NMDS Non-metric multidimensional
scaling
ODM Organic dry matter
oS Organic substance
OTU Operational taxonomic unit
P Packing sample of the AF
PBS Phosphate-buffered saline
PCR Polymerase chain reaction
pEGT Prokaryotic environmental
gene tag
XIX
pH Pondus hydrogenii
PI Propidium iodide
PS PowerSoil® DNA Isolation Kit
Ps. Genus Pseudomonas
PVPP Polyvinylpolypyrrolidone
qPCR Quantitative real-time PCR
RDP Ribosomal Database Project
rfu Channel intensity
RNA Ribonucleic acid
rRNA Ribosomal ribonucleic acid
rrs 16S rRNA gene
S Rye silage sample
Salm. Genus Salmonella
SDS Sodium dodecyl sulfate
Sh. Genus Shigella
SOC Super optimal broth with
catabolite repression
sp. Species
R2 Squared correlation coefficient
SSCP Single strand conformation
polymorphism
ssDNA Single strand DNA
ST Step by step DNA extraction
St Straw sample
STB Step by step DNA extraction
with beating step
str. Strain
TAMRA 6-carboxytetramethyl-
rhodamine
TRF Terminal restriction fragment
TRFLP Terminal restriction fragment
length polymorphism
U Enzyme unit
UASSR Upflow anaerobic solid-state
reactor
VFA Volatile fatty acids
WHO World health organization
X-Gal 5-bromo-4-chloro-indolyl-β-D-
galactopyranoside
The terms mesophilic, thermophilic and hyperthermophilic temperatures (or biogas
systems, biogas reactors, conditions etc.) were used to indicate temperature ranges
between 35-45°C, 55-65°C and 70-75°C, although these terms are logically incorrect
due to the fact that e.g. thermophilic (= ”heat-loving”) temperatures do not exist.
1
1 INTRODUCTION
1.1 Renewableenergysources
The enduring use of fossil fuels, such as coal, petroleum and natural gas, as a source
of energy is unsustainable due to supply limitations and greenhouse gas emissions.
The global energy reserves of non-renewable raw material make up approximately
39,375 exajoule (= 1018 joule) focusing on the currently technically and economically
recoverable quantities (DERA, 2011). With an average annual production of
479 exajoule for global consumption, non-renewable energy reserves will last for the
next 82 years (DERA, 2011). However, the total amount of resources is much higher
focusing on the geologically identified quantities of non-renewable raw material, which
are currently not economically or technically recoverable (DERA, 2011).
Nevertheless, the supply of some non-renewable energy sources will run out in the
coming decades. For instance, petroleum will be the first fossil fuel whose rising
demand cannot be met (DERA, 2011). In contrast, the non-renewable raw material
uranium, which is used for nuclear power production, will last for the next decades
(DERA, 2011), but nuclear power bears a substantial risk for the environment during
operation and also afterwards due to the need for nuclear waste storage.
The limitations on fossil fuels, the problems of greenhouse gas emission and the
potential risk of nuclear events, as occurred in the nuclear power plant Daichii
(Fukushima, Japan) in 2011, combined with the demographic and industrial increase
have intensified the urgent need to strengthen the usage and the efficiency of
renewable energy sources in recent years. Since 2000, the German renewable energy
2
law (EEG) favors renewable energy sources, such as wind and water power or biogas,
and hence the reduction of the greenhouse gas emission by prescribing incentives for
renewable energy supply. Furthermore, it schedules the nuclear phase-out for 2022,
supports the expansion and modernization of the electric supply network and supports
the further expansion of all renewable energies in Germany (BMU, 2012).
In 2011, renewable energy sources accounted for 12.5% of final energy consumption in
Germany, which includes the consumption of electricity, heat and motor fuel. This is an
increase by 36% within the last 5 years (BMU, 2012). In detail in 2011, 20.3% of
electricity generation, 11% of heat supply and 5.5% of motor fuel consumption based
on total energy consumption were provided by renewable energy sources (Figure 1.1).
For instance, biogas represents the fourth most prevalent renewable energy source for
the generation of electricity with 14.2% (Figure 1.1). Furthermore, it also contributes to
heating supply with 11.8% (biogas, sewer and landfill gas were combined as biogenic
gaseous fuel; Figure 1.1).
The production of biogas prevents an emission of 549 g CO2 equivalent per kWh in
electricity generation and 171 g CO2 equivalent per kWh in heating supply (BMU,
2012). In comparison, the generation of electricity by water or wind power prevents an
emission of 779 g and 721 g CO2 equivalent per kWh, respectively (BMU, 2012).
Although the production of biogas is not as effective as water or wind power in
preventing CO2 emission, it still contributes to the reduction of greenhouse gas and air
pollutant emission.
Electricity supply
by renewable energy sources
Water
power
14.7%
Wind
power
39.7%
Photo-
voltaic
15.7%
Biogenic solid fuel 9.2%
Biogenic liquid fuel 1.1%
Biogas 14.2%
Sewer and landfill gas 1.4%
Others 4%
Heating supply
by renewable energy sources
Biogenic liquid fuel 5.4%
Biogenic
solid fuel
68.1% Biogenic gaseous fuel 12.8%
Solar heat 3.9%
Geothermal energy 4.4%
Others 5.3%
Motor fuel supply
by renewable energy sources
Bio-
diesel
72.8%
Plant oil 0.6%
Bio-
ethanol
26.6%
20.3% 11.0% 5.5%
Figure 1.1 Proportion of renewable energy based on the total energy consumption in
Germany 2011 (BMU, 2012)
3
Beside the direct generation of electricity and heat supply via a communal heating
power station, biogas can also be used as biomethane after the reduction of impurities,
such as carbon dioxide (CO2), below 5% (Weiland et al., 2010). This can be achieved
by different purification and concentration steps in order to reach the quality
requirements for biomethane injection into the natural gas grid. In 2011, 83 biogas
plants were purifying their biogas to biomethane for the injection into the natural gas
grid (FNR, 2012b) and the tendency is to increase. This biomethane can also be used
as motor fuel and also for electricity and heat generation in many places due to the
distribution via the natural gas grid.
1.2 Agriculturalbiogasproduction
In Germany in 2011, 7,215 biogas plants were in operation with a total installed electric
capacity of 2,904 MW, which is enough to replace four coal-burning power plants or
two large nuclear power plants (FNR, 2012a).
In the most cases agricultural biogas plants were fed with renewable primary products
and animal manure. A survey of operators of German biogas plants in 2010 (n = 622)
indicated that 46% of the total substrates fed (based on weight) were renewable
primary products (energy crops), 45% livestock excrement, 7% biowaste and 2%
industrial and agricultural residues (DBFZ, 2011). Hence, energy crops, mainly maize
silage, play an important role in the production of biogas.
In 2010, approximately 5% of the 11.8 million hectares of German cropland was used
for the cultivation of biogas energy crops, whereas the percentage increased to
approximately 7% in 2011 (FNR, 2012c; Statistisches Bundesamt, 2012). To prevent
an excessive utilization of cropland and the monocropping of maize, several attempts
have been made to analyze and improve the utilization of other substrates, such as
industrial food waste (Kastner et al., 2012; Merlino et al., 2012) or grass silage
(Nizami et al., 2010, 2011; Nizami & Murphy, 2011; Lehtomäki et al., 2008).
The majority of agricultural biogas plants use substrates in continuously stirred tank
reactors (CSTRs) with a dry matter content of below 15% (so-called wet fermentation;
4
FNR, 2012a). Such systems are thoroughly stirred by a stirring device and usually
consist of more than one biogas-producing fermenter, while applying no phase-
separation between microbial conversion steps, such as hydrolysis or acidogenesis
and methanogenesis. These biogas systems are called two (or more) stage biogas
plants. In contrast to that, agricultural biogas plants with phase-separation converting
substrates with a dry matter content of up to 40% (so-called dry fermentation; FNR,
2012a) have been rare up to now. Dry fermentation combined with a separation of
process phases can lead to improvements due to the fact that optimal parameters for
each microbial conversion step can be set. This was supported by the findings of
Demirer and Chen (2005), who have achieved higher biogas yields in experimental
two-phase systems compared to one-phase reactors.
Beside the configuration of biogas plants, the temperature for the fermentation in those
systems also plays an important role. Although the majority of agricultural biogas plants
are operated at mesophilic temperatures (FNR, 2012a), it has been assumed that
thermophilic systems are more productive. Hence, a biogas system operated under
thermophilic conditions can achieve higher methane rates than a comparable
mesophilic system (Dugba & Zhang, 1999).
The biogas produced by agricultural biogas plants is versatile. The majority of
agricultural biogas plants convert the biogas to electricity and heat via communal
heating power stations (FNR, 2010). Prior to that, the biogas consisting of methane,
CO2, water vapor and hydrogen sulfide has to be purified. First of all, the biogas is
desulfurized and dried due to the fact that hydrogen sulfide and water vapor react to
produce sulfuric acid, which can corrode parts of the communal heating power station
(FNR, 2010). Afterwards, the biogas can be converted to electricity and heat.
To inject biomethane into the natural gas grid, further purification steps are needed
(FNR, 2010). Impurities, such as CO2 and traces of oxygen are separated from the
residual biomethane. After the addition of an odor-producing compound to the odorless
biomethane and the adjustment to the gross calorific value of natural gas, the
biomethane can be injected into the natural gas grid.
Beside the energy carrier methane, the digestates produced in agricultural biogas
plants are also utilized and applied on agricultural lands as fertilizers. A survey of
operators of German biogas plants in 2010 (n = 334) indicated that 78% of the
digestates were applied mainly without any treatment to their own fields (DBFZ, 2011).
5
In contrast to a direct application of undigested manure as fertilizer, the digestate has
an altered nutrient composition. The main differences consist in an increase in
phosphor, potassium and total nitrogen, which are mainly fixed as inorganic
compounds. Further, the dry matter content and the C/N ratio in the digestate are
reduced due to the anaerobic digestion of organic compounds (FNR, 2010). Whether
this leads, combined with an increased production of energy crops (need of carbon-rich
soils), to a lack of carbon in soils is still in dispute (Willms et al., 2008). However, any
such effects can be improved by other parameters, such as crop rotation (Willms et al.,
2008).
1.3 Thebiogasproductionprocess
1.3.1 Anaerobicdigestionofplantderivedbiomass
The first step in the anaerobic digestion process is the hydrolysis of plant-derived
carbohydrates, lipids and peptides (Figure 1.2). Energy crops, a widely used substrate
for biogas plants, are rich in carbohydrates, such as the energy store starch or the
plant cell wall polymers pectin, cellulose, hemicellulose and lignin. The degradation of
these polymers is of major importance for the whole biogas-producing process.
Starch consists of a large number of D-glucose molecules forming the linear, helical
amylose and the branched amylopectin. These polymers are easily degradable by
bacteria with the help of amylases. For instance, α-amylase cleaves the α-glycosidic
bonds within the starch polymer.
In contrast to easily degradable carbohydrates, the cell wall polysaccharides are more
recalcitrant to bacterial hydrolysis due to the complex network of different types of
polysaccharides.
Lignin, a complex aromatic heteropolymer, is the most recalcitrant component of plant
cell walls. It can be slowly degraded under aerobic conditions by white-rot and partly
brown-rot fungi (Dashtban et al., 2010). Only a few bacteria were identified with lignin
6
degrading activity, such as species of the Streptomycetes, Sphingomonas,
Pseudomonas and Acinetobacter genera (Ahmad et al., 2010; Bugg et al., 2011).
Different enzymes, such as laccases and peroxidases were essential ligninolytic
enzymes. Laccases are multicopper oxidoreductases, which catalyze the oxidation of
aromatic and non-aromatic compounds by radical-catalyzed reactions (Claus, 2004).
Peroxidases, heme-containing enzymes, catalyze H2O2-dependent oxidations of lignin-
associated aromatic compounds (Reddy & D`Souza, 1994). Furthermore, the enzyme
glyoxal oxidase is also essential for the extracellular lignin degradation (Whittaker et
al., 1999).
The cell wall polysaccharide, hemicellulose, consists of branched heteropolymers, such
as xylan, xyloglucan, arabinoxylan and (gluco-) mannan, which were formed by
different pentoses, such as xylose and arabinose and by hexoses, such as mannose
and glucose. However, the structure of hemicellulose can vary strongly in different
plants (Gaillard, 1965). Several hemicellulolytic enzymes are responsible for the
modification and degradation of hemicellulose. Enzymes, such as xylanase and
β-xylosidase are involved in the breakdown of xylan. The former enzyme cleaves the
β-1,4-xylosidic bond within xylan, whereas the latter enzyme cleaves xylose residues
from the non-reducing end of xylan. Furthermore, enzymes, such as β-mannase or
arabinase, act on the hemicellulolytic polymers mannan and arabinan. Hemicellulolytic
enzymes were summarized in different glycoside hydrolase (GH) families, such as GH
10, 11 and 43 (Henrissat, 1991; Cantarel et al., 2009). In particular, the GH families 10
and 11 compromise a high number of the enzymes acting on xylan (Collins et al.,
2005).
Pectin, another component of the plant cell wall, is mostly composed of galacturonic
acids, such as the α-1,4-linked homogalacturonan and rhamnogalacturonan (O’Neill et
al., 1990). These polymers are linked to other cell wall polysaccharides and hence form
part of the cellulosic, hemicellulosic network. Some of the pectinolytic enzymes, such
as the (exo-) polygalacturonase, exo-polygalacturonosidase and rhamno-
galacturonase are summarized in the GH family 28 (Henrissat, 1991; Cantarel et al.,
2009). For instance, the polygalacturonase catalyzes the hydrolysis of the α-1,4-bond
within the homogalacturonan polymer. This enzyme is mainly studied in the context of
the fruit-ripening process of plants (Giovannoni et al., 1989; Downs et al., 1992), but
also in the context of pectin modification or degradation performed by different
microorganisms (Barnby et al., 1990; McKay, 1990; Karam & Belarbi, 1995). However,
7
further enzymes such as pectin lyase and pectin esterase are also involved in the
degradation and modification of pectin (Jayani et al., 2005).
Furthermore, the homopolymer cellulose consists of D-glucose residues, which are
linked to each other by β-1,4-glycosidic bonds and form long linear chains. Multiple
layers of these linear polymers are linked to each other by hydrogen bonds forming
crystalline microfibrils. For degradation to occur, the β-1,4-glycosidic bonds as well as
the hydrogen bonds between the linear polymers must be cleaved by cellulases, i.e.,
endo- and exoglucanases and β-glucosidases (Lynd et al., 2002). The endoglucanases
act within the homopolymer cellulose in an amorphous region and cleave the
β-1,4-glycosidic bonds. The exoglucanases, such as cellobiohydrolases and
glucanohydrolases, release cellobiose and glucose residues, respectively, from the end
of the (crystalline) cellulosic chains. Finally, glucose residues are released from short
glucose chains, such as cellobiose or cellodextrin by the β-glucosidases (Lynd et al.,
2002). These cellulolytic enzymes are grouped in different GH families, such as GH 5,
6, 7, 9, 12 and 48 (Cantarel et al., 2009). However, the highest number of cellulase
genes was grouped in the GH families 5 and 9 (Schülein, 2000).
A high number of anaerobic bacteria, belonging to the Clostridiaceae,
Syntrophomonadaceae, Lachnospiraceae and Eubacteriaceae families, were identified
to hydrolyze crystalline cellulose (Schwarz, 2001). Particularly, members of the
Clostridiaceae family have a specialized extracellular cellulose-degrading enzyme
complex, the so-called cellulosome. Several studies have been published, which
analyzed the cellulosomal structure of Clostridium species and particularly
Cl. thermocellum (Bayer et al., 1985; Bayer & Lamed, 1986; Lamed et al., 1987).
Furthermore, cellulosomal structures were also detected within other anaerobic
bacteria, such as Acetivibrio or Bacteroides (Lamed et al., 1987; Ponpium et al., 2000;
Xu et al., 2003) and also within anaerobic fungal species, such as Neocallimastix
patriciarum (Wang et al., 2011). Cellulosomal complexes consist of many
polysaccharide degrading enzymes, such as cellulases and hemicellulases. These
enzymes are linked to the non-catalytic subunit, scaffoldin, which performs different
functions, such as the binding to the substrate cellulose and to the surface of the host
cell (Bayer et al., 1994; Shoham et al., 1999).
8
1.3.2 Productionofprecursorstomethanogenesis
The hydrolysis of polymers, such as carbohydrates, produces precursors to
acidogenesis (Figure 1.2). The di- and monomers produced are transported into the
cells and metabolized by various fermentative bacteria into volatile fatty acids (VFA),
alcohols, CO2 and H2. Both, the hydrolysis of polymers and the production of VFA are
performed by hydrolytic and fermentative bacteria. Members of the Clostridia class,
such as Clostridium, Ruminococcus, Caldicellulosiruptor or Anaerobacter, degrade
various carbohydrates and produce VFA and alcohols as fermentation end products
(De Vos et al., 2009). Different studies on biogas-producing communities have already
indicated an important role for these microorganisms (e.g., Klocke et al., 2007;
Krause et al., 2008a; Schlüter et al., 2008; Liu et al., 2009; Krakat et al., 2010a;
Wirth et al., 2012).
In contrast to acetate, the other VFAs have to be converted to acetate by acetogenic
bacteria (Figure 1.2) to make them useable for methanogenesis. The oxidation of
VFAs, such as propionate, is only thermodynamically favorable under reduced H2
partial pressure, which can be achieved by a syntrophic interaction with H2-scavenging
microorganisms, such as hydrogenotrophic methanogens (Scholten & Conrad, 2000;
Ahring, 2003). Furthermore, Siriwongrungson and coworkers (2007) have indicated
that homoacetogenic bacteria could also act as H2 sinks. They showed that the H2
produced after butyrate oxidation was directly used together with CO2 for the
production of acetate via homoacetogenesis (Figure 1.2) at least under thermophilic,
methanogenesis-repressed conditions. However, they supposed that homoacetogenic
microorganisms also occur in normally operated biogas processes.
Different syntrophic VFA-oxidizing bacteria, such as Syntrophomonas wolfei
(McInerney et al., 1981) or Syntrophothermus lipocalidus (Sekiguchi et al., 2000) and
also syntrophic propionate-oxidizing bacteria, such as Syntrophobacter fumaroxidans
(Harmsen et al., 1998) or Pelotomaculum thermopropionicum (Imachi et al., 2002),
were identified. The transfer of H2 between the VFA-oxidizing bacterium and its
syntrophic H2-scavenging methanogenic partner is enabled via direct interspecies
transfer as shown for Pelotomaculum thermopropionicum and Methanothermobacter
thermautotrophicus (Ishii et al., 2005).
9
Figure 1.2 Scheme of the anaerobic digestion process leading to the production of
biogas (modified after Weiland, 2010)
The acetate produced by syntrophic VFA oxidation can be directly converted by
acetoclastic methanogens or can be oxidized to CO2 and H2 by syntrophic acetate-
oxidizing bacteria (Figure 1.2). Hence, syntrophic acetate oxidizers compete with
acetoclastic methanogens for acetate uptake, having an advantage at reduced acetic
acid concentration and thermophilic temperatures (Ahring, 1995). Furthermore, this
syntrophic reaction is also thermodynamically more favorable the lower the H2 partial
pressure (Ahring, 2003). Up to now, only few syntrophic acetate-oxidizing bacteria are
known, which are capable to oxidize acetate in cooperation with H2-scavenging
partners: the rod-shaped strain AOR (Lee & Zinder, 1988), Cl. ultunense (Schnürer et
al., 1996), Thermacetogenium phaeum (Hattori et al., 2000), Thermotoga lettingae
(Balk et al., 2002), Syntrophaceticus schinkii (Westerholm et al., 2010) and
Tepidanaerobacter acetatoxydans (Westerholm et al., 2011).
The breakdown of plant-derived biomass and the subsequent production of VFA,
acetate and CO2 and H2 is a complex interaction of various microorganisms leading to
the production of precursors to methanogenesis (Figure 1.2).
1.3.3 Theproductionofmethane
The main precursors to methanogenesis are acetate and CO2 and H2 (Figure 1.2), but
also other precursors, such as formate, mono-, di- and trimethylamine or methanol can
10
be used for the production of methane by methanogenic archaea (Boone &
Castenholz, 2001). The conversion of these precursors to methane has been well
studied by different authors (Thauer et al., 1993; Blaut, 1994; Deppenmeier et al.,
1996). Various enzymes play an important role in hydrogenotrophic, methylotrophic
and acetoclastic methanogenesis and some of these enzymes are exclusively found in
methanogens (Blaut, 1994).
The hydrogenotrophic pathway, using CO2, H2 or formate, starts with the initial
reduction of CO2 to formyl methanofuran for which H2 or formate serves as electron
donor (Deppenmeier et al., 1996). Then, several enzymes catalyze the subsequent
stepwise reduction to methane. The H2 supply can be achieved via direct interspecies
transfer between the hydrogenotrophic methanogen and a syntrophic, H2-producing
bacterium (Ishii et al., 2005).
The methylotrophic pathway converts methanol or methylamines to methane and CO2.
One main difference to the hydrogenotrophic pathway is the way of electron donor
supply for the reduction to methane. The reducing equivalents are obtained by an
oxidation of a methyl group to CO2, which is subsequently used to reduce other methyl
groups to methane (Blaut, 1994).
Finally, the acetoclastic pathway converts acetate to methane and CO2. The first step
is the activation of acetate to acetyl-CoA, which is subsequently cleaved. The resulting
methyl residue is then reduced to methane (Blaut, 1994).
Despite these changes, the last step of the methanogenesis, the reductive
demethylation of methyl-CoM to methane, is found in the hydrogenotrophic,
methylotrophic and acetoclastic pathways. This reduction is catalyzed by the methyl
coenzyme-M reductase enzyme, whose gene has been often used as target for the
characterization of methanogenic archaea (e.g., Nettmann et al., 2008; Steinberg &
Regan, 2009; Tale et al., 2011; Zhu et al., 2011; Ellis et al., 2012).
Archaea with the capacity for methane production (so-called methanogenic archaea or
methanogens) are found in six orders: Methanobacteriales, Methanococcales,
Methanocellales, Methanomicrobiales, Methanosarcinales and Methanopyrales.
Members of the Methanopyrales do not grow below 80 °C (Boone & Castenholz, 2001)
and therefore play no role in the production of methane in biogas plants. Further, the
Methanocellales and its mesophilic, hydrogenotrophic type strain Methanocella
11
paludicoda, which has been proposed recently by Sakai and coworkers (2008), have
not yet been identified to play a role in biogas-producing plants.
The Methanobacteriales, Methanococcales and Methanomicrobiales orders
encompass hydrogenotrophic methanogens using CO2 and H2 or formate for the
production of methane. Some strains of these orders may also be capable of using
methanol (Boone & Castenholz, 2001). Particularly members of the
Methanobacteriales and Methanomicrobiales are well known as part of the
methanogenic community in mesophilic and thermophilic biogas systems as they were
identified in different studies (e.g., Bauer et al., 2008; Klocke et al., 2008; Schlüter et
al., 2008; Krakat et al., 2010b; Kongjan et al., 2011). Their apparent abundance in
biogas plants might be favored by the fact that they seem to be robust against
environmental factors, such as increased nitrogen concentrations (Koster & Lettinga,
1984).
However, members of the Methanosarcinales were also identified as prevalent within
mesophilic and thermophilic biogas-producing communities (e.g., Godon et al., 1997;
Karakashev et al., 2005; Ziganshin et al., 2011; Ellis et al., 2012; Lerm et al., 2012).
The two families of the Methanosarcinales order differ strongly from each other.
Members of the Methanosarcinaceae are able to use the broadest spectrum of
precursors (e.g., acetate, CO2/H2, methanol) for methane production, whereas
members of the Methanosaetaceae are strict acetoclastic methanogens, exclusively
using acetate for the production of methane (Boone & Castenholz, 2001). In particular,
members of the latter group were identified as reacting more sensitively to high
nitrogen or VFA concentrations (Karakashev et al., 2005; Nettmann et al., 2010), which
can be accumulated during the biogas fermentation process.
1.4 Purportedproliferationofpathogensinbiogassystems
Agricultural biogas plants are still the focus of the public interest due to a potential risk
for unintended proliferation and distribution of pathogens infecting animals, humans
and plants. Pathogens derived from substrates, such as manure or energy crops, are
12
assumed to proliferate during the anaerobic digestion process bearing the risk of
distribution through the application of the digestate as fertilizers on fields. The outbreak
of the enterohemorrhagic E. coli str. O104:H4 (EHEC) in Germany 2011 increased the
attention towards biogas production plants as a potential distributor of pathogens.
During this outbreak, about 3,900 humans were infected by the E. coli str. O104:H4,
suffering from (bloody) diarrhea and also from hemolytic-uremic syndrome (WHO,
2011). After extensive investigations, fenugreek sprouts and not biogas plants were
identified to be the most likely cause of this foodborne infection (BfR, 2011).
Beside the E. coli strain, pathogenic members of the genus Clostridium, such as
Cl. botulinum, Cl. perfringens, Cl. difficile or Cl. tetani, were also alleged to proliferate in
biogas plants leading to an endangering of humans and also animals health. For
instance, Cl. botulinum, which was recently detected in cattle feces (Dahlenborg et al.,
2003), can cause botulism in cattle. This pathogen in addition to other clostridial
pathogens were assumed to find perfect growth conditions in biogas plants due to the
fact that numerous, non-pathogenic Clostridium species, such as Cl. thermocellum or
Cl. stercorarium, play an important role in biomass degradation and the subsequent
acidogenic pathway. However, some studies already indicate that the risk for an
unintended proliferation of pathogenic Clostridium species is rather low (e.g.,
Dohrmann et al., 2011).
Additionally, other pathogens, such as L. monocytogenes, Campylobacter jejuni or
Salmonella enterica, which are also in focus of the public interest as causing foodborne
infections in humans and were also considered in this study.
An unintended proliferation of pathogens may also affect plants due to the application
of digestates as fertilizer. Particularly plant pathogens infecting agricultural crops with
economic importance may raise the concerns not only of farmers. One important plant
pathogen is Clavibacter michiganensis (Actinobacteria), which infects tomatoes and
potatoes (Gartemann et al., 2003). Various attempts have been made to find a control
method of this plant-pathogenic species. Wittmann and coworkers (2010) isolated
proteins from a Clavibacter michiganensis infecting bacteriophage, which showed
Clavibacter-specific bacteriolytic activity leading to a potential biocontrol of this plant
pathogen. However, other plant-pathogenic species are also responsible for infections
in potatoes or tomatoes, such as Synchytrium endobioticum, Rhizoctonia solani,
Helminthosporium solani or Ralstonia solanacearum. Further plant pathogens, such as
Fusarium oxysporum and Sclerotinia sclerotiorum can also lead to infections of grain.
13
1.5 Characterizationandmonitoringofthemicrobialcommunity
A deep understanding of the microbial community within biogas reactor systems is of
major importance for the improvement of such systems. Culture-dependent methods,
although indispensable for the detection of new isolates, can only reflect a marginal
amount of the broad microbial diversity due to limitations in cultivation conditions.
Different authors have indicated that the number of cultivable bacteria varies between
0.1 and 1% in seawater and meso- and oligotrophic lake habitats using viable plate
count analysis (Ferguson et al., 1984; Staley & Konopka, 1985). Further, Wagner and
coworkers (1993) have reported higher recoveries of 14% for activated sludge samples
using specialized media. However, culture-independent methods are important to gain
deeper insights into the microbial structure.
Therefore, a polyphasic approach composed of different culture-independent methods
was performed for the characterization, quantification and monitoring of the microbial
community residing in the two-phase biogas reactor systems. On the one hand, the
culture-independent methods were based on the genetic information of cells, applying
16S rRNA (rrs) gene sequence analysis, terminal restriction fragment length
polymorphism (TRFLP) fingerprinting, microbial metagenome and quantitative
polymerase chain reaction (qPCR) analyses. On the other hand intact, but not
cultivated cells were analyzed by fluorescence in situ hybridization (FISH) as well as
4′,6-diamidino-2-phenylindole (DAPI) and propidium iodide (PI) staining. Both
approaches were extensively used to analyze microbial communities in different
environmental habitats as indicated below.
1.5.1 Analysisofbacterialandarchaealrrsgenesequences
The construction and analysis of gene libraries is a widely applied method for the
determination of the composition of microbial communities in different habitats
(e.g., Großkopf et al., 1998; Vetriani et al., 1999; Danon et al., 2008; Soo et al., 2009)
as well as in biogas reactor systems (e.g., Chouari et al., 2005; Klocke et al., 2007,
2008; Cardinali-Rezende et al., 2009; Liu et al., 2009; Biswas & Turner, 2012). The rrs
14
gene is an often used and well-studied target gene encoding the 16S rRNA, which
forms together with different proteins the 30S small subunit of bacterial and archaeal
ribosomes. The rrs gene is 1.541 kb long, as identified after the analysis of E. coli in
1978 (Brosius et al.) and consists of nine hypervariable regions (van de Peer et al.,
1996). This allows the application of different phylogenetic approaches, such as the
differentiation between bacterial or archaeal groups on a higher or lower taxonomic
level depending on the rrs region used for primer design. In this study, primer sets were
used for the characterization of the bacterial and archaeal community, which spans the
first five hypervariable regions.
However, independent of the target gene, this method is based on a DNA amplification
step using target specific primer sets. Afterwards, the PCR product is cloned into
plasmids, which were subsequently transformed into competent E. coli cells. Cells
transformed were cultivated and the plasmids carrying the gene of interest were
isolated and sequenced. The sequence data obtained yields information about the
composition of the target community.
Furthermore, with the help of different statistical parameters, deeper insights into the
diversity and evenness of the underlying microbial community can be obtained. The
Simpson and Shannon diversity indices (Simpson, 1949; Shannon & Weaver, 1963)
reflect the quantity of operational taxonomic units (OTUs) and the distribution within
these OTUs, assessing the diversity of the target community. The parameter Evenness
(Lloyd & Ghelardi, 1964; Hill, 1973) gives information about the distribution of OTUs
obtained. So a value of 1 means an equal distribution of OTUs and a reduced value
indicates a strong prevalence of a specific OTU. Additionally, to check the quality of the
coverage of the gene libraries, Good’s coverage (Good, 1953) and the Chao-I
estimator (Chao, 1987) can also be calculated. Good’s coverage indicates the
percentage coverage of the underlying microbial community, whereas the Chao-I
estimator values indicate the extrapolated number of species or in this case OTUs.
Both parameters take those species (or OTUs) into account, which have only a rare
abundance.
15
1.5.2 Terminalrestrictionfragmentlengthpolymorphism
The TRFLP method is a powerful, fingerprinting tool for the analysis and comparison of
microbial communities within different habitats. Spatial and temporal changes in the
composition of microbial communities can also be monitored (Lukow et al., 2000;
Marsh, 2005). This method has been introduced by Liu and coworkers in 1997. Up to
now, this valuable method has also been widely applied by different authors in the field
of biogas production (e.g., Feng et al., 2010; Wang et al., 2010; Carballa et al., 2011;
Pycke et al., 2011; Ziganshin et al., 2011).
In comparison to other fingerprinting methods, such as denaturing gradient gel
electrophoresis (DGGE), amplified ribosomal DNA restriction analysis (ARDRA) and
single strand conformation polymorphism (SSCP), the TRFLP method bears two main
advantages. First, this method allows the analysis of a large number of samples at
once and can be seen as high-throughput method (Enwall & Hallin, 2009). Second, the
separation of DNA fragments in a gel-filled capillary as performed by the TRFLP
analysis leads to a higher resolution and a more detailed estimation of the fragment
size. Hence, the immediate digital output is another advantage leading to a better
comparison of multiple profiles (Marsh, 1999).
TRFLP analysis is based, as with many culture-independent methods, on a PCR step
(Figure 1.3) amplifying the ribosomal, but also functional genes. One of the primers
used for the PCR is labeled with a fluorescent dye, such as 6-carboxyfluorescein
(FAM) or cyanine 5 (Cy5). After DNA amplification, the fluorescently labeled PCR
product is digested by a restriction endonuclease (Figure 1.3), which reveals the
evolutionary based differences within the target gene (Marsh, 2005). For a high
resolution of a complex community, a digest with two or even more restriction
endonucleases can reduce the possibility that more than one microbial group results in
the same fragment size (Kitts, 2001). Afterwards, the fluorescently labeled fragments
were separated together with an internal size standard by an automated capillary gel
electrophoresis system after electrokinetic injection (Figure 1.3). The use of an internal
size standard allows a size calculation of the terminal restriction fragments (TRFs) with
an accuracy of one base up to a fragment size of 700 bp (Kitts, 2001; Schütte et al.,
2008).
16
Figure 1.3 Short workflow of the sample preparation for the TRFLP analysis. RFU –
channel intensity
The raw data obtained have to be processed before data analysis and interpretation
can be performed. One point is to separate true fragments from the background noise.
Several methods have been applied by different authors, such as using a manually set
baseline threshold (Osborn et al., 2000; Lueders & Friedrich, 2003), applying a
constant percentage (Lukow et al., 2000; Sait et al., 2003) or a variable percentage
threshold (Osborne et al., 2006).
Furthermore, an analysis of at least duplicate samples, taking only TRFs represented
in both samples into account, could reduce potential “background” fragments as
reported by Dunbar and coworkers (2001). Additionally, these authors have developed
recommendations for the normalization of TRFLP data to solve the issue of different
loads of DNA to the capillary gel electrophoresis. For normalization, the total
fluorescence (based on TRF height) of each sample has to be summed, reflecting the
total DNA injected. Afterwards, the sample with the smallest amount of total
fluorescence is successively divided by the total fluorescence of the other samples in
the dataset resulting in a correction factor for each sample. This factor is then
multiplied by the height of each TRF of the respective sample resulting in the same
total fluorescence as the sample with the smallest total fluorescence. This
normalization method for TRFLP data was used as basis for further data analysis in
this study.
However, another approach also addressing the issues of fragment identification from
background noise as well as the variations in DNA loads has been introduced by Abdo
and coworkers (2006). Their statistical method also produced standardized data ready
for further data analysis. First, each fragment area was divided by the total area of all
fragments of one sample, resulting in relative fragment areas. Then, to distinguish
between true and background fragments, the standard deviation of the sample was
calculated and fragments with a relative area, showing a larger value than three
17
standard deviations, were identified to be true. This iterative procedure was performed
until no true fragment could be identified anymore.
Afterwards, these processed data must be aligned due to slight variations in separation
accuracy in order to compare TRFLP profiles derived from various samples. In earlier
studies, this has been achieved by a manual binning, which is a time-consuming and
error-prone process. An automated and hence more sophisticated approach is the use
of alignment tools, such as the web-based T-Align (Smith et al., 2005) and T-Rex tool
(Culman et al., 2009). Both tools use an algorithm, which clusters all TRFs within a
specific clustering threshold (in this study 0.8 bp), starting with the smallest TRF within
the samples. The clustering threshold can be adjusted by the user in both applications.
However, the T-Rex software is more versatile, allowing the analysis of more than
duplicate samples and the use of different output options.
This multi-level processing of data can be performed on the basis of the TRF height or
area. Different authors have favored the TRF height (Dunbar et al., 2001; Caffaro-Filho
et al., 2007) or the TRF area (Kitts, 2001; Sait et al., 2003), but other show no
preference between the both approaches (Schütte et al., 2008). Hence, both methods
can be used for data processing and the subsequent graphic visualization of the
results, but also for further statistical approaches, such as the calculation of similarity
indices. The calculation of these indices allows an estimation of the similarity of two
different samples. In this study, the Jaccard index improved by Chao (Chao et al.,
2005) and the Bray-Curtis similarity index (Magurran, 1988) were used. Whereas the
Jaccard index is a qualitative index based on the presence and absence of values (in
this case TRFs), the Chao-Jaccard index also takes the potential presence of unseen
shared species of two samples into account (Chao et al., 2005). In contrast, the
Bray-Curtis index (Magurran, 1988) also gives information about the numerical quantity
of values (TRFs) from two samples.
For further interpretation of potential differences of TRFLP results, statistical ordination
methods, such as non-metric multidimensional scaling (NMDS), can be applied. The
NMDS approach tries to reduce the complex correlation between objects (samples) on
the basis of the order of similarity (or dissimilarity) values to resolve ideally in a
two-dimensional space by applying an iterative algorithm (Leyer & Wesche, 2008). In
contrast to other ordination methods, NMDS only requires a monotonous relationship
between the similarity values of the objects in the data matrix. To confirm the quality of
such ordination methods, Kruskal and coworkers (1964) have introduced a factor of
18
goodness, the so-called stress value, which has been reinterpreted by Clarke (1993).
He suggested that a stress value of up to 0.1 shows a good ordination and stress
values up to 0.2 still can lead to usable ordination plots.
In this study, the Bray-Curtis similarity index was used as basis for non-metric
multidimensional scaling. The stress values obtained, which described the quality of
the NMDS analysis, were indicated for each NMDS plot. Furthermore, a scree plot
analysis was constructed for ensuring that the reduction to two dimensions is
appropriate. This test has been proposed by Cattell (1966) and has been used in the
context of the multidimensional scaling by Kruskal and Wish (1978).
1.5.3 AnalysisofmicrobialmetagenomesbyhighthroughputDNA
sequencing
High-throughput DNA sequencing approaches, so-called next generation sequencing
methods, were applied for whole genome sequencing, metagenome or transcriptome
sequencing and also further applications. In contrast to the Sanger dideoxy sequencing
method, which was introduced in 1977 (Sanger et al., 1977), these methods allow an
extremely increased sequence throughput at less cost and in less time (Margulies et
al., 2005).
In 2005, 454 Life Science (now a Roche Company, Branford, USA) introduced the
Genome Sequencer (GS) 20 system applying the 454-pyrosequencing method
(Margulies et al., 2005), which since then has been widely employed (e.g., Leininger
et al., 2006; Turnbaugh et al., 2006; Warnecke et al., 2007; Krause et al., 2008a;
Schlüter et al., 2008). The recent available GS FLXTM Titanium system is a further
development, which when combined with the latest chemistry, allows the analysis of
one million sequences per run with a mean sequence length of 800 bp, accounting for
700 Mb of total sequence information (454 Life Science, 2012).
Recently other high-throughput sequencing technologies have been introduced, such
as the HiSeq and MiSeq systems by Illumina (San Diego, USA) and the SOLiDTM
system by Applied Biosystems (a Life Technologies brand, Carlsbad, USA). These
methods result in shorter sequence length than achieved by 454-pyrosequencing, but
19
are able to sequence more bases per run (up to 600 Gb) (Applied Biosystems, 2011;
Illumina Inc., 2012). In 2011, Pacific Biosciences (Menlo Park, USA) commercialized
their SMRTTM sequencing technology, which enables mean sequence length of up to
3,000 bp (Pacific Biosciences, 2012). These technical developments in the field of the
next generation sequencing have raised the number of (meta-) genomic studies
immensely.
At the end of 2012, the Genome OnLine Database (GOLD; Pagani et al., 2012)
encompassed a total number of 18,896 genomes for worldwide access and
comprehensive analysis of genomic data. Furthermore, a number of 345 metagenomic
studies analyzing 2,145 environmental samples were also deposited in the GOLD
database. A total of 10 studies each focused on waste water and solid waste samples.
However, the greatest effort has been applied to analyze aquatic habitats with
149 metagenomic studies.
For the analysis of metagenomes, the 454-pyrosequencing technology is favorable due
to the fact that a relatively high sequence length can be achieved. Hence, it was also
used in this study. First of all, highly pure and unshared genomic DNA, extracted from
an environmental sample, is needed for the downstream 454-pyrosequencing
(Figure 1.4), which has been described in detail by Margulies and coworkers (2005)
and Droege and Hill (2008). After DNA extraction and purification, adaptors (A and B)
are ligated to the DNA (Figure 1.4), which is nebulized into 300 to 800 bp long
fragments. These fragments are then bound to sepharose beads by their adaptors.
Afterwards, a “water in oil” amplification of these DNA fragments is performed
(Figure 1.4), which results in millions of the same DNA fragments per bead. Each of
these beads is loaded into a well of a 454 PicoTiterPlateTM together with the reaction
chemistry for sequencing. When the plate is loaded onto the GS FLXTM system,
nucleotides are flowed successively across this plate in a fixed order. If a nucleotide is
complementary to the sequence, it will be incorporated in the DNA strand by the DNA
polymerase and a light signal will be generated. The intensity of the light signal is to a
certain extent proportional to the number of nucleotides incorporated (Figure 1.4). This
enzymatic reaction is initialized by the incorporation of the nucleotide into the DNA
strand. The pyrophosphate released can be converted into ATP in the presence of
adenosine 5'-phosphosulfate, which is catalyzed by the ATP sulfurylase enzyme.
Afterwards, the ATP-required conversion of luciferin to oxyluciferin, which is catalyzed
by the luciferase enzyme, produced the light signal detected by a charge-coupled
20
device camera. Due to the fact that the type of nucleotide which is flowed across the
plate is known, the sequence of the DNA strand can be amended. After the light signal
detection, the residual nucleotides per well are degraded, a reaction, which is catalyzed
by the enzyme apyrase. Then, a second type of nucleotide is flowed across the
PicoTiterPlateTM and will be incorporated, if complementary. These steps are repeated
until the decoding of the DNA sequence is completed.
Figure 1.4 Short workflow of the 454-pyrosequencing analysis using the GS FLXTM
Titanium System (454 Life Science, a Roche Company, Branford, USA)
The analysis of such huge datasets is a sophisticated effort for bioinformatics. In the
case of whole genome studies, the assembly of many overlapping metagenomic
sequences is performed, which can also compensate the sequencing error rate
(Margulies et al., 2005). In contrast to that, an assembly of metagenomic sequences is
limited particularly when analyzing very complex microbial communities (Mende et al.,
2012). Further, a taxonomic assignment of metagenomic sequences assembled to
large contigs is only reliable on higher taxonomic levels (Schlüter et al., 2008). To
achieve more detailed phylogenetic information about the metagenomic sequences, no
assembly was performed in this study.
Phylogenetic information can be drawn on the basis of metagenomic sequences
encoding the 16S rRNA. Furthermore, the metagenomic sequences representing
functional genes can improve the phylogenetic insights. Beside these phylogenetic
assignments, such a metagenomic approach also provides information of the genetic
potential and hence the enzymatic capacity of a specific microbial community, which is
in contrast to an exclusive analysis of the rrs gene. These different applications of
sequence interpretation have been applied by various authors analyzing the
metagenome in different habitats (e.g., Bench et al., 2007; Krause et al., 2008a;
Kröber et al., 2009). In this study, these analyses were performed with the MetaSams
platform 0.99, which has been recently introduced by Zakrzewski and coworkers
(2013). This platform, accessible via a web interface provided by the Center for
21
Biotechnology (Bielefeld University, Bielefeld, Germany), has integrated different
approaches for metagenomic data analyses.
1.5.4 QuantificationofbacterialrrsgenesbyquantitativePCR
The qPCR technique is an often applied method for the detection and quantification of
microorganisms. Several authors have quantified methanogenic or biogas communities
on the basis of the rrs gene applying qPCR (e.g., Yu et al., 2005, 2006; Bergmann et
al., 2010; Blume et al., 2010). This method follows the principle of a PCR, but allows a
simultaneous quantification of DNA after each qPCR cycle due to double-strand (ds)
DNA-intercalating fluorescent dyes or group-specific fluorescently labeled probes.
Thereby, the fluorescence signal detected reflects the concentration of the target gene
in real-time (Zhang & Fang, 2006). The dsDNA-intercalating fluorescent dye
(e.g., SYBR Green) can be easily used with each standard PCR protocol, but it also
presents some problems. The fact that these dyes bind unspecifically not only to the
target dsDNA, but also to primer dimers or non-target DNA could distort the qPCR
results (Sharkey et al., 2004; Zhang & Fang, 2006).
More specific is the use of group-specific, fluorescently labeled probes (e.g., TaqMan
probes) and its corresponding primer pairs. The TaqMan probe is labeled by a reporter
(e.g., FAM) and a quencher dye (e.g., TAMRA) and binds to the target gene within the
binding site of the primer pair. Due to the close proximity of the reporter and the
quencher dye, the fluorescent signal is repressed. During DNA amplification the
TaqMan probe is degraded by the exonuclease activity of the DNA polymerase, which
leads to an enlarged distance between both dyes. Now, the detection of the
fluorescence of the reporter dye is possible. These and further general aspects of the
qPCR analyses have been reviewed in detail by different authors, such as Bergmann
(2012) and Zhang and Fang (2006).
In this study, a TaqMan approach, targeting the rrs gene, was used for the absolute
quantification of both Bacteria as a whole and a specific putatively process-relevant
bacterium.
22
1.5.5 Microscopicalanalyses
The use of microscopical analyses, such as FISH or DAPI staining, allows a
visualization of microbial cells and has been widely applied by different authors to gain
information about the number of microbial cells in various biogas reactors (e.g., Burrell
et al., 2004; Chouari et al., 2005; Krakat et al., 2010b; Nettmann et al., 2010; Biswas &
Turner, 2012). DAPI staining allows the detection of all microbial cells due to the fact
that DAPI can pass through intact membranes and intercalates in the minor groove of
the DNA. In contrast to that the fluorescent dye PI intercalates in DNA and RNA, but
only within cells having reduced membrane integrity. Hence, this dye has been often
used for estimations of potential cell damage, for instance, in flow cytometry analyses
(e.g., Belloc et al., 1994; Fröhling, 2010; Khan et al., 2010). However, potential cell
damage, indicated by the PI staining, could also be a reversible process as indicated
by Davey and Hexley (2011). They showed that a certain number of stressed yeast
cells did not incorporate PI after some time of recovery. Therefore, a reduced
membrane integrity detected by PI staining may at least indicate changes in the cell
status.
For a specific differentiation of microbial groups, FISH has been developed by Amann
and coworkers (1995). The functional principle of this method lies in the binding of
fluorescently labeled oligonucleotide probes to the complementary region within the
ribosomal RNA of the target cells. Cells hybridized with a fluorescently labeled probe
could then be visualized by fluorescence microscopy.
Depending on the samples analyzed, the background fluorescence, caused by plant
material, could infer the detection of the probe-labeled cells. Therefore, further
developments of the FISH method, such as the Catalyzed Reporter Deposition FISH
(CARD-FISH; Speel et al., 1999; Pernthaler et al., 2002) and the Double Labeling of
Oligonucleotide Probes for FISH (DOPE-FISH; Stoecker et al., 2010) have been
proposed for improving the signal intensity and compensating the background
fluorescence. The DOPE-FISH method is based on the fluorescent labeling of the
5’ and 3’ end of oligonucleotide probes. In contrast, the CARD-FISH method is based
on the binding of oligonucleotide probes labeled with horseradish peroxidase and a
separated signal amplification using a fluorescently labeled tyramide.
23
Several recent studies use a variety of oligonucleotide probes for targeting different
groups of microorganisms. The online resource probeBase (Loy et al., 2007) gives an
overview and detailed information of probes designed for different microbial groups.
1.6 TheAimsofthisstudy
The main aim of this study was the characterization, quantification and monitoring of
the microbial community in three identically constructed phase-separated leach-bed
biogas systems under thermophilic to hyperthermophilic conditions. A deep
understanding of the microbial community involved in the breakdown of plant-derived
biomass and the subsequent production of methane is of major importance for a further
improvement of biogas reactors.
The biogas systems analyzed in this study consisted of a leach-bed reactor (LBR) for
hydrolysis and acidogenesis, a leachate storage reactor and a downstream anaerobic
filter reactor (AF) for methanogenesis. The LBR was supplied with rye silage and straw
over a period of 21 days and was operated at temperatures from 55 to 75 °C. The
temperature of the AF remained at 55 °C throughout the experiment.
In detail, the various aspects listed below were analyzed to achieve the main aim.
To derive an overview about the microbial community in the two-phase biogas
system, the bacterial and archaeal community in the LBR and in the downstream
AF was characterized.
To check the intended separation of process phases within this phase-separated
biogas system, the spatial distribution of the microbial community was analyzed.
To assess changes in the microbial community as a consequence of biomass
degradation, the bacterial and archaeal community was analyzed at different time
points within the 21-day fermentations of one load of substrate.
To assess temperature-dependent changes, the bacterial and archaeal
community was analyzed during the stepwise temperature increases in the LBR.
24
To gain deeper insights into the anaerobic digestion of plant-derived biomass, the
genetic potential for the expression of glycoside hydrolases, enzymes involved in
the breakdown of glycosidic linkages of carbohydrates, was determined.
To evaluate the putative risk of an unintended proliferation of pathogens,
selected pathogens and enzymes relevant for pathogenicity were analyzed.
To evaluate the effect of a bioaugmentation with compost at hyperthermophilic
temperatures, the microbial community and the biogas reactor performance were
analyzed after bioaugmentation.
To determine the total number of microorganisms, Bacteria, Archaea and a
putatively process-relevant bacterium residing in the two-phase biogas system
were quantified.
These community analyses were performed applying different culture-independent
methods. The polyphasic approach comprised the construction of rrs gene libraries and
the application of the TRFLP fingerprinting method to characterize and monitor the
community. Furthermore, the analysis of microbial metagenomes also revealed insights
into the genetic potential of the microbial community in the biogas system. The
quantification of Bacteria, Archaea and the putatively process-relevant bacterium was
performed by qPCR analyses based on the rrs gene and by microscopical studies of
intact, but not cultivated cells.
25
2 MATERIALANDMETHODS
2.1 Thetwophaseleachbedbiogassystems
2.1.1 Reactorsetupandoperation
Three identically constructed two-phase biogas systems, which have been operated at
the ATB (Potsdam, Germany) since 2006 by M. Schönberg (Schönberg & Linke, 2012),
were analyzed in this study. The majority of the results refer to the first reactor system.
The second and third reactor system was analyzed at specific time points to confirm
the reproducibility of the results.
These two-phase biogas systems consisted each of three stainless steel reactors
(Figure 2.1) with gastight top covers of acryl glass: a leach-bed reactor (LBR; net
volume 100 L), a leachate reservoir (LR; net volume 60 L) and a downstream
anaerobic filter reactor (AF; net volume 30 L) with 390 packings (Bioflow 40; Rauschert
GmbH, Judenbach-Heinersdorf, Germany) each with a surface area of 305 m2 m
-3.
Two internal circulations of leachate were applied using membrane pumps (W100,
Wilden Pump & Engineering LLC, Grand Terrace, CA, USA) to distribute nutrients,
microorganisms and to keep the moisture and temperature level consistent. Each LBR
was supplied discontinuously with 10 kg of rye silage (agricultural farm Damsdorf GbR
H. & T. Wessels, Damsdorf, Germany) with a chaff length of 2 cm (silage period
100 days without an ensiling agent). In addition to the silage, 0.5 to 1.0 kg of winter
barley straw (Lehr- und Versuchsanstalt für Tierzucht und Tierhaltung e.V., Groß
Kreutz, Germany) or wheat straw (agricultural farm Bornim, Potsdam, Germany) were
26
used as bulking material. The rye silage was stored at 4 °C and the straw material at
room temperature until usage. The retention time of the substrates was 21 days. After
that fermentation period, the digestate was removed and each LBR reactor was refilled
with fresh substrate.
All reactors were heated by a water jacket. Each AF reactor was operated throughout
the experiment at a temperature of 55 °C, whereas the temperature of the LBR reactor
was increased stepwise from 55 to 75 °C by 5 °C increments. Three fermentation
periods of 21 days were conducted at each of the mentioned temperature regimes
(LBR 55 - 75 °C) before further temperature increase was applied (Figure 2.2). After
the temperature regime of 70 °C, a bioaugmentation by addition of 4-week-old compost
(5 kg, OTS content 14.55%, pH 8.17; Biowork GmbH, Schmergow, Germany) to the
substrate was conducted while maintaining the LBR temperature at 70 °C. After two
fermentations with compost lasting 21 days, 10 kg of rye silage and 1 kg of straw
material were digested again in the LBR for 21 days at 70 °C as a control. Afterwards,
the temperature regime in each LBR was increased to 75 °C following the previous
scheme. To ensure that the micronutrients were not a limiting factor for microbial
growth, approximately 50 g of a DSMZ 144 micronutrient medium (2.5x concentrated;
DSMZ GmbH, Braunschweig, Germany) was added to each AF before reactor restart
with new substrate.
Figure 2.1 Scheme of the thermophilic two-phase leach-bed biogas reactor system.
Circulation rates of the two internal leachate circulations were indicated.
LBR, leach-bed reactor (net volume 100 L); LR, leachate reservoir (net
volume 60 L); AF, anaerobic filter reactor (net volume 30 L); GB, gas bag
(modified after Schönberg & Linke, 2012)
27
2.1.2 Analysisofprocessparameters
To monitor the biogas process, various process parameters were determined by the
analytical chemistry work group at the ATB. Prior to each 21-day anaerobic digestion,
the pH value (cf. pH measurement below), dry matter and the organic dry matter
(ODM) content of the substrates were measured. The dry matter was determined
through evaporation of the water content in substrates by heating to 105 °C until no
further weight loss occurred. The ODM was determined by ashing the samples at
550 °C. After 21 days of fermentation, the degradation of organic material was
calculated based on the ODM before and after the fermentation. This value was
corrected by the nondegradable lignin fraction of the substrate supplied.
During the first five days of fermentation, pH, chemical oxygen demand (COD) and
concentrations of acetic acid, n-butyric acid, propionic acid, valeric acid, total VFA,
ethanol and propanol in the leachate derived from the LBR and the AF were measured
once a day. Afterwards, the intervals were enlarged to two or three days.
The pH was measured using a calibrated pH meter 340i (WTW GmbH, Weilheim,
Germany).
The COD was determined using potassium dichromate solution for solid samples and
the COD-cuvette-test for liquid samples (Hach Lange GmbH, Düsseldorf, Germany).
The concentrations of acetic acid, n-butyric acid, propionic acid, valeric acid, ethanol
and propanol were analyzed after filtration of a cold water extract with toluol by the
Varian CP-3800 gas chromatograph (Varian Inc., Palo Alto, CA, USA) and a
Permabond®-FFAP capillary column (Macherey-Nagel GmbH & Co. KG, Düren,
Germany). The VFA were calculated as acetic acid equivalents (HAc eq.).
The total ammonia nitrogen (NH4-N) was determined once a week. For this purpose,
the samples were analyzed by a Vapodest 20 distillations system (C. Gerhardt GmbH
& Co. KG, Königswinter, Germany). The concentration of free ammonia (NH3) was
calculated according to Anthonisen and coworkers (1976) using the following formula:
NH3 = (total ammonia nitrogen * 10pH) (Kb/Kw + 10pH)-1;
where the total ammonia nitrogen concentration is measured in g L-1, Kb/Kw is
e(6344/(273+t)) with t equal to the temperature in °C.
28
The biogas generated was collected in TECOBAG gas bags (100 L, Tesseraux
Spezialverpackungen GmbH, Bürstadt, Germany). Every day, an automatic gas
analysis was conducted by a TG05/5 drum-type gas meter (Ritter Apparatebau GmbH
& Co. KG, Bochum, Germany) and a SSM 600 biogas analyzer (Pronova
Analysentechnik GmbH & Co. KG, Berlin, Germany). The biogas obtained was
standardized to the norm temperature 0 °C (273.15 K) and to the norm pressure of
1,013 hPa with the following formula:
Biogasnormed = biogasmeasured * ((pmeasured – pH20) To (Tmeasured * po)-1),
where pmeasured is the air pressure at the point of measurement, pH2O is vapor pressure
of water at the point of measurement, To is the norm temperature (273.15 K), Tmeasured is
the temperature in Kelvin at the point of measurement and po is the norm air pressure
(1,013 hPa).
Then, the biogasnormed was calculated on the basis of the total organic substances used
for biogas production (including ODM, VFA and alcohol). The biogas data were
provided by M. Schönberg (ATB, Potsdam, Germany) for subsequent data analysis.
2.1.3 Reactorsampling
Depending on the method applied, different samples were taken from the biogas
reactor system. Samples were taken at the earliest after one fermentation period of
21 days at each temperature regime (LBR 55 - 75 °C; Figure 2.2) to ensure adaption of
the microbial community to the increased temperature. The one exception was the
experimental run 12 (Figure 2.2) which was analyzed directly after temperature
increase to 70 °C in order to monitor the immediate bacterial response to temperature
increase.
About 50 g of the substrate samples (i.e., rye silage (S), straw (St) and compost (C))
were taken before filling the LBR. The leachate samples (L) of the LBR (approximately
50 mL each) were obtained from the leachate reservoir, whereas the leachate samples
of the AF were derived from the plug of the AF reactor (Figure 2.1). These samples
were taken after draining the plug by 500 mL to prevent sampling of stale leachate
29
within the tube. The digestate samples (D, approximately 50 g each) were obtained
after a fermentation period of 21 days from the LBR at each temperature regime.
Before aliquotation, the digestate was mixed manually.
The packing (P) samples of the AF were taken from the upper half of the AF at each
temperature regime of the LBR. The gastight reactor was opened at the end of the
fermentation period to prevent a disturbance of the biogas production process. The
biofilms of these packings were rinsed with 1x phosphate-buffered saline (1x PBS;
pH 7.4, 80 g L-1 NaCl, 2 g L-1 KCl, 26.8 g L-1 Na2HPO4 x 7 H20, 2.4 g L-1 KH2PO4) in
order to remove the planktonic cells and were afterwards detached using a sterile
scalpel. All samples were stored at -20 °C until further processing.
Experimental runs (each 21 days)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Temperature of LBR (°C)
55
60
65
70
75 Bioaugmentation
Experimental run omitted
Samples taken
Figure 2.2 Experimental runs of the three two-phase biogas systems and sampling of
the first biogas system. Arrows, samples taken for molecular biological
analyses; x, data of these experimental runs were omitted from results
due to technical problems
2.2 ExtractionandquantificationofgenomicDNA
2.2.1 DNAextractionprotocols
To evaluate the bias caused by DNA extraction and to define the standard DNA
extraction protocol for the following analyses, four DNA extraction methods were
compared. Digestate (0.25 g) derived from the LBR was used for extraction and the
efficiency and reproducibility subsequently analyzed using the TRFLP method. The
30
different DNA extraction methods were performed applying the PowerSoil® DNA
Isolation Kit (MO Bio Laboratories, Inc., supplied by Dianova GmbH, Hamburg,
Germany), the FastDNA® Spin Kit for Soil (MP Biomedicals GmbH, Heidelberg,
Germany) as well as a step by step DNA extraction protocol with and without a beating
step. Both DNA extraction kits are based on a mechanical cell lysis and were used
according to the manufacturer’s instructions. The step by step protocol without a
beating step used chemicals (SDS) and enzymes (lysozyme) for cell lysis. The fourth
protocol, the step by step DNA extraction with beating step (cf. protocol below)
combined the chemical and enzymatic lysis with a beating step on a test tube shaker
for 10 min at maximum power, using sterile glass beads. Depending on the applied
molecular analysis, two different DNA extraction protocols were applied as standard
protocols.
The FastDNA® Spin Kit for Soil (MP Biomedicals GmbH, Heidelberg, Germany) was
applied as standard DNA extraction protocol for TRFLP analyses, rrs gene library
construction and qPCR analyses. Cell pellets were collected from 2 to 4 mL of leachate
by centrifugation at 14,000 x g for 10 min (C5417R centrifuge; Eppendorf GmbH,
Hamburg, Germany) and were added directly to the supplied Lysing Matrix E tubes. To
isolate DNA from solid materials, 0.2 g of digestate from the LBR, 0.2 g of biofilm
material from the packing of the AF and 0.05 g of straw material were added directly to
the tubes mentioned above. The subsequent DNA extraction was performed as
described in the manufacturer’s instructions.
For archaeal positive controls of the rrs gene amplification, the DNA of the
methanogenic pure cultures Mtb. thermautotrophicus DSM 1053, Methanosarcina
thermophila DSM 1825 and Mcu. marisnigri DSM 1498 (DSMZ GmbH, Braunschweig,
Germany) were also extracted using the FastDNA® Spin Kit for Soil.
The step by step DNA extraction protocol was performed for metagenomic analyses
according to Rheims and Stackebrandt (1999) and Klocke and coworkers (2008). In
order to yield highly pure and high-molecular-weight DNA for the 454-pyrosequencing,
three digestate samples of the LBR at 55 °C (D 55), 65 °C (D 65) and 70 °C (D 70) as
well as one biofilm sample of the packing derived from the AF (P 55) at an LBR
temperature of 55 °C were analyzed. For each of these four samples, the DNA of ten
subsamples was extracted to reduce a potential DNA extraction bias. The subsamples
of the digestate (0.3 g each) were resuspended in 1 mL saline-EDTA (0.1 M EDTA,
0.15 M NaCl), whereas the packing’s subsamples (0.2 g each) were washed twice with
31
1 mL sodium-phosphate buffer (0.1 M, pH 7.0) to reduce contamination of humins.
These samples were then centrifuged at 3,500 x g for 15 min (C5417R centrifuge;
Eppendorf GmbH, Hamburg, Germany). The cell pellets obtained were also
resuspended in 1 mL saline-EDTA.
In addition to the chemical and enzymatic cell lysis, a mechanical lysis step with a
lower intensity than the FastPrep® Instrument was applied to keep the genomic DNA
unshared. In detail, sterile glass beads (1 x 4 mm and 20 x 1.5 mm in diameter) were
added to the samples, which were homogenized at maximum speed on a test tube
shaker (Heidolph Instruments GmbH & Co. KG, Schwabach, Germany) for 5 min.
Afterwards, 15 to 20 mg of polyvinylpolypyrrolidone (PVPP) and 30 µL of lysozyme
solution (10 mg mL-1 lysozyme, 10 mM Tris/HCl) were added and the samples were
incubated at 37 °C for 60 min. Briefly, 30 µL of proteinase K (10 mg mL-1 proteinase K,
50 mM Tris/HCl, 1.5 mM calcium acetate), 120 µL of SDS (10% w/v) and 120 µL of
CaCl2 (10 mM) were added for lysing the cells. After incubation at 65 °C for 45 min, the
samples were centrifuged at 6,000 x g for 10 min (C5417R centrifuge, Eppendorf
GmbH, Hamburg, Germany) and the supernatants were adjusted to 0.7 M NaCl and
5% (w/v) CTAB (hexadecyltrimethylammonium bromide). The samples were incubated
on a heating block (Eppendorf GmbH, Hamburg, Germany) at 65 °C for 30 min.
Subsequently, the DNA was extracted twice with an equal volume of chloroform-
isoamylalcohol (24:1 v/v) and precipitated from the obtained water phase by adding
0.1 volumes of 3 M sodium acetate (pH 5.2) and an equal volume of isopropanol. The
precipitated DNA was collected by centrifugation at 4 °C with 20,817 x g for 20 min
(C5417R centrifuge, Eppendorf GmbH, Hamburg, Germany). The DNA pellets were
washed twice with ethanol (70% v/v), dried and resuspended in 50 µL of HPLC-H2O.
All chemicals were provided by AppliChem GmbH (Darmstadt, Germany).
The genomic DNAs derived from the subsamples of the respective metagenomic
samples was pooled and then purified with the NucleoBond CB 20 Kit (Macherey-
Nagel, Düren, Germany) according to the manufacturer`s instructions. The DNAs,
extracted from the digestates of the LBR and the biofilm of the AF had concentrations
of 1.04 µg µL-1 (D 55), 0.62 µg µL-1 (D 65), 0.53 µg µL-1 (D 70) and 0.43 µg µL-1 (P 55),
respectively, and were highly pure with an absorption ratio of A260/A280 = 1.82 ± 0.03
and A260/A230 = 2.08 ± 0.29.
This protocol was also used for the DNA extraction from the rye silage as the DNA
extraction using the FastDNA® Spin Kit for Soil was not applicable for this substrate.
32
Chemical lysis was applied for extracting DNA of the pure cultures Cl. tyrobutyricum
DSM 2637, Escherichia coli DSM 1116 and Ps. fluorescens DSM 50090 (DSMZ
GmbH, Braunschweig, Germany), which were used as bacterial positive controls for
the rrs gene amplification. In detail, 2 mL of a liquid culture were centrifuged at
14,000 x g for 5 min (C5417R centrifuge, Eppendorf GmbH, Hamburg, Germany). The
supernatant was removed and the pellet obtained was resuspended in 200 µL of the
residual supernatant. Then, 200 µL of cell lysis solution (2.5% SDS, 25 mM NaOH) was
added and the sample was incubated in a heating block at 95 °C for 15 min. The
obtained lysate was centrifuged at 6,000 x g for 10 min (C5417R centrifuge, Eppendorf
GmbH, Hamburg, Germany) and the supernatant containing the extracted DNA was
collected.
The successful DNA isolation was validated by a 1.2% agarose gel electrophoresis
(Biozyme Scientific GmbH, Hessisch Oldendorf, Germany) stained with ethidium
bromide (0.3 µg mL-1), VWR International GmbH, Darmstadt, Germany). A total of 2 µL
of DNA together with 18 µl loading buffer (0.1 M EDTA, 40% glycerol, 0.1% SDS,
0.025% bromphenol blue) were applied to the gel. Additionally, 6 µl of DNA size
standard (Lambda DNA/EcoRI+HindIII; Fermentas GmbH, part of Thermo Fisher
Scientific, St. Leon-Rot, Germany) was used for size evaluation. The DNA was stored
at 4 °C until further use.
2.2.2 DNApurityandquantification
The purity of genomic DNA was measured using the NanoPhotometer (Implen,
München, Germany) according to the manufacturer’s instructions. The ratios of the
absorbance at 260 nm and 280 nm as well as at 260 nm and 230 nm were used as
quality index for DNA purity.
The DNA quantification for the metagenomic and qPCR analyses was performed using
the NanoDrop 3300 (Thermo Fisher Scientific, Wilmington, USA). The DNA
concentration was determined after applying the DNA intercalating fluorescent dye
PicoGreen (Quant-iT™ PicoGreen® dsDNA; Life Technologies GmbH, Darmstadt,
Germany). In equal parts, the PicoGreen dye (1:200) was incubated together with the
33
DNA of the sample for 5 min before measuring the emission at 525 nm (excitation at
470 nm). The standard curve was designed using calf thymus DNA (Sigma-Aldrich
Chemie Gmbh, München, Germany) with concentrations between 20 and 2,000 ng
per mL following the scheme described above.
2.3 Constructionofrrsgenelibraries
Two bacterial and one archaeal rrs gene libraries were constructed to gain insights into
the composition of the microbial community, to verify the TRFLP data and to identify
the TRFs obtained.
For an optimal amplification of the bacterial and archaeal rrs gene, the 27f forward
primer (bacterial assay, Table 2.1) and the 109f forward primer (archael assay,
Table 2.1) were tested with several reverse primers: 926r (5’ CCGTCAATTCCTTTGAG
TTT 3’) (Moeseneder et al., 1999), 926Rr (5’ CCGTCAATTCCTTTRAGTTT 3’) (Muyzer
et al., 1995), 926MRr (Table 2.1), 1492r (5’ GGYTACCTTGTTACGACTT 3’) (Sait et
al., 2003) and 16Srev (5’ TACGGYTACCTTGTTACGACTT 3’) (Klocke et al., 2007) for
the bacterial assay and Ar912r (Table 2.1), Ar915r (5’ GTGCTCCCCCGCCAATT
CCT 3’) (Großkopf et al., 1998) and Ar958r (5’ YCCGGCGTTGAMTCCAATT 3’)
(DeLong, 1992) for the archaeal assay. The reaction mix was as follows in a total
volume of 25 µL: 1x PCR buffer, 1.75 mM MgCl2 (bacterial assay) or 2 mM MgCl2
(archaeal assay), 0.2 mM dNTPs, 0.4 µM of each primer and 1.2 U (bacterial assay) or
1 U (archaeal assay) of the recombinant Taq DNA polymerase (all: Fermentas GmbH,
part of Thermo Fisher Scientific, St. Leon-Rot, Germany). The thermal profiles of the
PCR reactions were performed as follows: (1) 95 °C for 3 min, (2) 94 °C for 30 s, (3)
50 °C for 40 s (primer 926r, 926Rr, 926MRr) or (3) 51 °C for 30 s (primer 1492r,
16Srev) or (3) 47 to 52 °C (gradient) for 30 s (primer Ar912r, Ar915r, Ar958r), (4)
elongation at 72 °C for 90 s, steps 2 to 4 were repeated 25 times followed by a final
extension at 72 °C for 8 min.
After primer evaluation, the optimal concentration of chemicals and enzymes, such as
the MgCl2 and Taq DNA polymerase, and the optimal annealing temperature were
34
evaluated. Furthermore, to optimize the PCR reaction and to estimate the robustness
of the TRFLP method, the cycle numbers of the PCR reactions were evaluated. The
protocols described below were used as standard protocols for the characterization of
Bacteria and Archaea by rrs gene sequence and TRFLP analyses.
For the construction of the rrs gene library, DNAs extracted from the leachate of the
LBR at 55 and 75 °C on day 7 were used for the bacterial rrs gene libraries, whereas
DNA from the packing’s biofilm of the AF at an LBR temperature of 55 °C was used for
the archaeal rrs gene library construction. For the amplification of the bacterial rrs
gene, the 27f forward primer and the 926MRr reverse primer (Table 2.1) were used.
For the amplification of the archaeal rrs gene, the Ar109f forward primer and the Ar912r
reverse primer (Table 2.1) were applied (all: Biomers, Ulm, Germany). The following
reaction mix was used as standard protocol for the DNA amplification in a total volume
of 25 µL: 1x PCR buffer, 2 mM MgCl2 (bacterial assay) or 2.5 mM MgCl2 (archaeal
assay), 0.2 mM dNTPs, 0.4 µM of each primer and 1 U of the recombinant Taq DNA
polymerase (all: Fermentas GmbH, part of Thermo Fisher Scientific, St. Leon-Rot,
Germany). The thermal profile for the DNA amplification was as follows: (1) initial
denaturation at 95 °C for 3 min, (2) denaturation at 94 °C for 30 s, (3) annealing at
51 °C (bacterial assay) or (3) 52 °C (archaeal assay) for 30 s, (4) elongation at 72 °C
for 90 s, steps 2 to 4 were repeated 25 times for the bacterial assay or 28 times for the
archaeal assay followed by a final extension at 72 °C for 8 min (Thermal Cycler 2720,
Applied Biosystems by Life Technologies, Darmstadt, Germany; TGradient and TProfessional,
Biometra GmbH, Göttingen, Germany).
The PCR amplification was validated by 1.2% agarose gel electrophoresis (AppliChem
GmbH, Darmstadt, Germany). After purification using the QIAquick PCR Purification Kit
(Qiagen, Hilden, Germany), the PCR products were ligated into a pGem®-T vector
(Promega, Mannheim, Germany) at 4 °C over night. Then the ligation mixtures were
transformed into competent E. coli JM 109 cells (Promega GmbH, Mannheim,
Germany) at 42 °C for 45 s and incubated in SOC medium (20 g L-1 tryptone, 5 g L-1
yeast extract, 0.5 g L-1 NaCl, 2.5 mM KCl, pH 7; after sterilization 20 mM sterile glucose
solution, 10 mM sterile MgCl2 solution were added) for 1.5 h at 37 °C on a test tube
shaker. After the incubation, 100 µL transformation mix was plated onto LB plates
containing ampicillin (50 µg mL-1), IPTG (60 µg mL-1) and X-Gal (60 µg mL-1, all
provided by AppliChem GmbH, Darmstadt, Germany). The plates were incubated at
37 °C over night. White colonies were picked after blue-white screening and incubated
35
over night at 37 °C in LB broth (Carl Roth GmbH & Co. KG, Karlsruhe, Germany)
containing ampicillin (50 µg mL-1). Recombinant plasmids were extracted using the
NucleoSpin® Plasmid kit (Macherey-Nagel, Düren, Germany) according to the
manufacturer’s protocol. The cloning was verified by means of DNA restriction in a total
volume of 20 µL using 1 U NcoI and 1 U SalI (Fermentas GmbH, part of Thermo Fisher
Scientific, St. Leon-Rot, Germany) together with double-concentrated Tango buffer for
2 to 3 h. For each of the three rrs gene libraries, 96 plasmids with inserts of expected
length were sequenced by GATC Biotech AG (Konstanz, Germany). The quality of
sequences was checked using the Chromas software tool (version 1.43; Technelysium
Pty Ltd, South Brisbane, QLD, Australia). Chimeric sequences were identified applying
the Mallard 1.02 software (Ashelford et al., 2006). A total of 9% of all sequences were
identified as chimera and were removed from the study. The remaining rrs sequences
were clustered into OTUs with a p-distance of 0.03 to each other applying the Mega
5.05 software (Tamura et al., 2011). The sequences were blasted against the NCBI nr
database applying the megablast algorithm. Sequences from environmental or
uncultured species were excluded from the BLAST analysis.
The statistical analyses of the rrs gene clone libraries were conducted with the PAST
software (Hammer et al., 2001), using the default settings, in the case of the Simpson
and Shannon diversity indices (Simpson, 1949; Shannon & Weaver, 1963) and the
Evenness value (Lloyd & Ghelardi, 1964; Hill, 1973). The Chao-I estimator (Chao,
1987) based on the bias-corrected formula was calculated with the EstimateS software
version 8.0.0 (Colwell, 2009). Furthermore, Good’s coverage was calculated according
to Good (1953). The Bray-Curtis similarity indices (Magurran, 1988) were calculated
using the default settings of the EstimateS software version 8.0.0 (Colwell, 2009).
2.4 Terminalrestrictionfragmentlengthpolymorphismanalysis
Two TRFLP assays were established focusing on the rrs genes of Bacteria and
Archaea, respectively. The first step for TRFLP analysis was the amplification of the rrs
gene. For bacterial rrs gene amplification, the 27f forward primer labeled with Cy5
36
(5’ end) and the 926MRr reverse primer (Table 2.1) were used. For the amplification of
the archaeal rrs gene, the Ar109f forward primer also labeled with Cy5 (5’ end) and the
Ar912r reverse primer (Table 2.1) were applied (all: Biomers, Ulm, Germany). The
amplification of the rrs gene was performed using the amplification protocol as
described above (cf. 2.3). Positive controls were set up using genomic DNA
purified from following cultures as template: E. coli (DSM 1116), Ps. fluorescens
(DSM 50090) and Cl. tyrobutyricum (DSM 2637) for the bacterial TRFLP assay and
Mtb. thermautotrophicus (DSM 1053), Msr. thermophila (DSM 1825) and Mcu.
marisnigri (DSM 1498, DSMZ GmbH, Braunschweig, Germany) for the archaeal assay.
The PCR amplification was validated by 1.2% agarose gel electrophoresis (AppliChem
GmbH, Darmstadt, Germany). The PCR products were purified using the QIAquick
PCR Purification Kit (Qiagen, Hilden, Germany) according to manufacturer’s
instructions and again checked by gel electrophoresis. The concentrations of the PCR
products were determined using the NanoPhotometer (Implen, München, Germany).
For the restriction enzyme digestion of the PCR products, seven different enzymes for
the bacterial and archaeal assay (AluI, BfaI, HaeIII, Hin6I, MspI, TaqI, TasI; Fermentas
GmbH, part of Thermo Fisher Scientific, St. Leon-Rot, Germany) and three enzyme
combinations for the bacterial assay (MspI and Hin6I, HaeIII and MspI, HaeIII and
Hin6I) were tested with preference for enzymes resulting in the most heterogeneous
fingerprint profile.
As standard protocol, a total of 200 to 250 ng DNA was digested enzymatically in a
total volume of 20 µL at 37 °C for 4 h using 10 U of each enzyme, MspI and Hin6I, in
the case of the bacterial assay or AluI for the archaeal assay according to the
manufacturer’s instructions. Complete digestion of the PCR products was checked
using the capillary gel electrophoresis.
The digested PCR products were purified by ethanol precipitation using 0.1 volume of
3 M sodium acetate and 200 µL 75% ethanol (protocol based on A. Ulrich, ZALF,
Müncheberg, Germany and amended by the sodium acetate step). The samples were
incubated in the dark for 30 min and centrifuged at 20,817 x g at 4 °C. The supernatant
was removed and again 200 µL ethanol was loaded into the samples, repeating the
centrifugation step. Then the pellets were dried in a vacuum concentrator (concentrator
5301, Eppendorf GmbH, Hamburg, Germany) for 3 to 5 min and resuspended in 20 µL
(archaeal assay) or 40 µL (bacterial assay) HPLC-H2O.
37
Electrophoretic separation and detection of TRFs, which were labeled with the
fluorescent dye Cy5, were conducted using a GenomeLabTM GeXP Genetic Analysis
System (Beckman Coulter, Krefeld, Germany). Each sample was loaded together with
0.2 to 0.5 µL of DNA size standard labeled with D1 dye (S-600, Beckman Coulter,
Krefeld, Germany) comprising DNA fragments from 60 to 640 nucleotides. The
samples were filled up to a volume of 30 µL with sample loading solution (SLS,
Beckman Coulter, Krefeld, Germany) and denatured at 90 °C for 120 s. The
electrokinetic injection of samples and size standards lasted 10 to 20 s at 2 kV and the
separation lasted 70 to 90 min applying 4.8 kV. A separation time of 90 min was used
to confirm a complete restriction enzyme digestion. After validating the complete
digestion, the separation time was reduced to 70 min.
Parts of the bacterial TRFLP analyses of the 21-day fermentation period at 60, 65 and
70 °C were conducted in the context of the Bachelor thesis of C. Nolte (2011) and
provided the basis for further analysis and interpretation of the data in this study.
Recording and analysis of data were performed with the GeXP analysis software
(version 10.2, Beckman Coulter, Krefeld, Germany). The size of the TRFs was
calculated with the quartic model based on the migration time of the size standard. The
calibration curve was generated arranging the fragment size of the size standard
against the migration time. The size standard fragment with 140 bp was omitted from
the analyses, which improved the squared correlation coefficient of the calibration
curve. The squared correlation coefficient and the standard deviation of the size
standard were above 0.99 and between 0.3 to 0.4 nucleotides, respectively, during all
measurements, indicating a good DNA fragment separation. Due to the size standard,
TRFs between 60 and 650 bp were analyzed in this study. Further, TRFs showing
fluorescence intensities below 370 channel intensity (rfu) were not measurable within a
linear range and were therefore removed from this study. The TRFLP data obtained
were subsequently normalized focusing on the total fluorescence intensity of the peak
height corresponding to Dunbar and coworkers (2001). A comparison between results
based on the peak height and peak area was applied to evaluate potential differences
in the TRFLP results due to the analysis of the peak area. Afterwards, the analysis
based on the TRF heigh was used as standard.
Alignment of TRFs with a clustering threshold of 0.8 was conducted applying the T-Rex
software (Culman et al., 2009). For further downstream analyses, the only TRFs used
were those which were represented in at least two of the conducted triplicates per
38
sample. For the bar charts, TRFs with fluorescence intensities below 3% of the total
detected fluorescence intensity were removed from the analysis. Then the relative
abundance of each TRF was calculated based on the aligned and normalized data set
according to Wang and coworkers (2010) and displayed using SigmaPlot 8.0 (Systat
Software, Erkrath, Germany). The similarity indices, Bray-Curtis (Magurran, 1988) and
Chao-Jaccard (Chao et al., 2005), were calculated for different TRFLP profiles using
the default settings of the EstimateS software version 8.0.0 (Colwell, 2009). The non-
metric multidimensional scaling (NMDS) was constructed on the basis of a Bray-Curtis
similarity matrix. The NMDS was performed applying the Statistica software (StatSoft
GmbH, Hamburg, Germany) using the Standard Guttman-Lingoes option as starting
configuration. Up to 6 dimensions were analyzed for each TRFLP profile for scree plot
construction in order to confirm that a reduction to two dimensions, as shown in the
NMDS plots, was appropriate.
To identify TRFs, the plasmids of interest were amplified using the T7 and SP6
(Table 2.1) promoter primers (Eurofins MWG GmbH, Ebersberg, Germany). The
reaction protocol was as follows using a total volume of 20 µL: 1x PCR buffer, 1.5 mM
MgCl2, 0.2 mM dNTPs, 0.2 µM of each primer and 0.8 U of the recombinant Taq DNA
polymerase (all: Fermentas GmbH, part of Thermo Fisher Scientific, St. Leon-Rot,
Germany). The thermal profile started with (1) an initial denaturation at 94 °C for 2 min,
followed by (2) a denaturation step at 94 °C for 30 s, (3) an annealing step at 47 °C
1 min and (4) an elongation step at 70 °C for 2 min. Steps 2 to 4 were repeated
30 times and were finished by a final extension at 70 °C for 10 min. After the QIAquick
PCR purification, a second PCR was applied focusing on the bacterial or archaeal rrs
gene using the TRFLP primers and the thermal profile mentioned above (cf. 2.4). PCR
products were purified, digested, precipitated by ethanol and finally analyzed on the
GenomeLabTM GeXP Genetic Analysis System as described before (cf. 2.4). The
identification and phylogenetic assignment of TRFs monitored for the reactor samples
were conducted on the basis of the cloned rrs gene sequences resulting in a TRF of
similar size.
39
2.5 Metagenomicanalysisapplying454pyrosequencing
technology
To analyze the microbial metagenome, the genomic DNA, isolated from the digestate
of the LBR at 55 (D 55), 65 (D 65) and 70 °C (D 70) and from the packing’s biofilm of
the AF at an LBR temperature of 55 °C (P 55), was used. The 454-pyrosequencing
analysis was performed at the Center for Biotechnology (CeBiTec, Bielefeld University,
Germany) using the Genome Sequencer (GS) FLXTM platform (Roche Applied Science,
Mannheim, Germany). Altogether, four libraries of the DNA samples were constructed
applying the GS Rapid Library Prep Kit (Roche Applied Science, Mannheim, Germany)
according to the manufacturer’s protocol. These four libraries representing different
microbial communities were sequenced on a PicoTiterPlate on the GS FLXTM system
using the Titanium sequencing chemistry (Roche Applied Science, Mannheim,
Germany). The raw data were processed by CeBiTec using the analysis pipeline for
whole genome shotgun sequence reads applying the GS FLXTM System Software
(version 2.3, Roche Applied Science, Mannheim, Germany).
The subsequent phylogenetic assignment and functional annotation were conducted on
the basis of the unassembled metagenomic sequences, using the MetaSAMS platform
0.99 (Zakrzewski et al., 2013), a tool for metagenomic data analysis. This platform was
also provided by CeBiTec via web interface. The integrated Ribosomal Database
Project classifier (RDP classifier; Wang et al., 2007) was applied for the phylogenetic
assignment based on rrs gene sequences. To identify rrs gene fragments, a search
using the Basic Local Alignment Search Tool (BLAST) (Altschul et al., 1990) versus the
RDP database (Release 10.19) was conducted. To allow the analysis of metagenomic
sequences with low complexity, the sequence complexity filter was disabled
(option ‘-F F’). Reads with an E-value threshold of 10-10 and a confidence value of at
least 80% were extracted and classified using the RDP classifier. The accuracy of RDP
classification ranges between 99.5% to 99.8% at phylum level and 83.2% to 88.7% at
genus level for 200 to 400 nucleotides segments analyzing type strain sequences and
further rRNA coding sequences (Wang et al., 2007).
A further classifier, CARMA version 2 (Krause et al., 2008b) was used for phylogenetic
assignment and functional annotation of metagenomic sequences representing
functional genes (so-called environmental gene tags, EGTs). Using the default settings
40
(Krause et al., 2008b), EGTs were identified by matching metagenomic sequences on
protein family (Pfam) members (Pfam database 24.0), which were represented by
multiple sequence alignments and hidden Markov models (Finn et al., 2008). In a
further step, the functional information based on the Pfam accessions was combined
with the results of the phylogenetic assignment obtained by CARMA. CARMA allows
classification with specificity of 97% at superkingdom and 68% at genus level analyzing
short EGTs of a synthetic dataset produced from 77 complete genomes (Krause et al.,
2008b).
To get an impression of the genetic potential for carbohydrate degradation,
methanogenesis and pathogenicity in the microbial community, the Pfam (Finn et al.,
2008) profiles obtained from CARMA were used. For this purpose, enzymes relevant to
the degradation of biomass and to the methanogenesis were identified with the help of
KEGG (Ogata et al., 1999) and grouped according to Pfam protein families.
Furthermore, enzymes relevant to pathogenicity of selected pathogens were
determined with the help of the Pfam database (Punta et al., 2012).
2.6 QuantitativerealtimePCR
The qPCR is a powerful tool for quantifying microorganisms within various habitats.
The number of Bacteria as well as the number of a specific potentially process-relevant
bacterium in the two-phase biogas reactor (TRF 304) was quantified. This bacterium
was dominant in the TRFLP analysis and rrs gene library analysis after seven days of
fermentation at LBR temperatures of 55 and 60 °C, which indicated a potentially
important role in biomass degradation.
To quantify Bacteria, a Bacteria-specific primer set with Bac338f, Bac516TaqMan and
Bac805r (Table 2.1) was used, which amplifies a 468 bp fragment. Furthermore, a TRF
304-specific primer set, composed of the 304f forward primer, the TaqMan304 probe
and the 304r reverse primer (Table 2.1), was designed, applying the software tools
Primer Express® version 3.0 (Applied Biosystems by Life Technologies, Darmstadt,
Germany) and Primer 3 version 0.4.0 (Rozen & Skaletsky, 2000). This primer set
41
amplifies a 151 bp fragment. Afterwards, the target specificity was analyzed in order to
identify false negative and positive assignments. Therefore, the rrs gene libraries were
checked after a sequence alignment using the software Mega 5.05 (Tamura et al.,
2011). Furthermore, the primer set 304 was checked for false positive and false
negative matches applying the Primer BLAST tool of NCBI (Altschul et al., 1990;
Wheeler et al., 2007) and the probe match tool of RDP (Cole et al., 2009).
For standard curve construction, the corresponding plasmid of TRF 304
(ATB-AR_23384; accession number HE804843) was used for the primer set 304 and
the Pectobacterium carotovorum ssp. carotovorum DSM 30168 strain as a
representative of the Bacteria was used for the primer set Bac. The cultivation, DNA
extraction and amplification, cloning into the pGem vector and isolation of plasmids of
this strain were performed as described in Bergmann (2012). The plasmid of TRF 304
(HE804843) was retransformed into E. coli JM 109 cells and extracted as described
above for the rrs gene library construction (cf. 2.3). The plasmids were linearized in a
total volume of 40 µL, using 2 U ScaI (Fermentas GmbH, part of Thermo Fisher
Scientific, St. Leon-Rot, Germany) for 16 h at 37 °C and 80 °C for 20 min (inactivation
of the enzyme). Linearization was validated by 1.2% agarose gel electrophoresis
followed by a purification of the linearized plasmids using the QIAquick purification kit
(Qiagen, Hilden, Germany). The concentrations of genomic DNAs and plasmid DNAs
were determined using the NanoDrop 3300 (Thermo Fisher Scientific, Wilmington,
USA), following the previously described scheme (cf. 2.2.2). For standard curve design,
the number of plasmid copies was calculated as follows, according to Bergmann
(2012):
NDNA = (CDNA*NA) (lplasmid*mmol,bp)-1,
where NDNA is the number of DNA copies (μL-1), CDNA is the DNA concentration (g μL-1),
NA is the Avogadro constant (6.022 * 1023 mol-1), lplasmid is the plasmid length in base
pairs and mmol,bp is the average molar mass of a base pair (660 g mol-1). Afterwards,
serial dilutions with copy numbers between 101 and 109 were made.
The amplification of the target sequences was performed using the ABI 7300 System
(Applied Biosystems by Life Technologies, Darmstadt, Germany). The qPCR for both
assays was performed in a total volume of 25 µL using 12.5 µL TaqMan universal PCR
master mix (Applied Biosystems by Life Technologies, Darmstadt, Germany), 0.9 µM of
each primer, 0.2 µM TaqMan probe (FAM/TAMRA) and 1 ng template DNA. All DNA
42
samples were measured in triplicates. The thermal profile for the bacterial primer set
(Bergmann, 2012) and the primer set 304 were as follows: 2 min at 50 °C, 10 min at
95 °C, followed by 45 cycles (primer set Bacteria) or 40 cycles (primer set 304) of 15 s
at 95 °C, 30 s at 57 °C (only for primer set Bacteria) and 1 min at 60 °C. At the end of
each cycle, the fluorescence emitted from the samples was detected.
The analysis of the results was conducted by applying the 7300 Real-Time PCR
Sequence Detection Software (version 1.3; Applied Biosystems by Life Technologies,
Darmstadt, Germany). The calibration curve was generated by plotting the plasmid
concentration of the standard against the number of cycles. The squared correlation
coefficient (R2) and the slope of the calibration curve ranged between 0.992 and 0.995
and between -3.28 and -3.37, respectively, for all measurements. The slope of the
standard curve was used to determine the efficiency of the PCR reaction by the
following formula:
efficiency = (10(-1/slope)) - 1.
The calculated PCR efficiency was between 0.980 and 1.018, which indicated an
efficient standard curve calibration (Zhang & Fang, 2006).
The limit of detection (LOD) and the limit of quantification (LOQ) were calculated
according to Bergmann (2012). For the bacterial qPCR analysis, Ct values (threshold
cycle number) of 40.7 and 40.0 were achieved for LOD and LOQ, respectively. For the
analysis, using the TRF 304 primer set, the LOD and LOQ values ranged from 35.0 to
35.6 and from 34.4 to 35.2, respectively. All samples measured showed Ct values
above these calculated limits.
For graph design, the number of rrs gene copies per mL leachate was calculated as
follows:
rrs gene copies (mL-1) = (rrs gene copy number * DNA concentration
(ng mL-1)) (used volume for DNA extraction)-1.
43
2.7 Microscopicalanalyses
Three different methods for staining and counting microbial cells derived from the two-
phase biogas system were applied. To determine the total cell counts in the leachate,
the microbial cells were stained with DAPI. To quantify the bacterial and archaeal cells
in the reactor samples, DOPE-FISH was performed. Further, the percentage of
microbial cells with reduced membrane integrity was estimated by PI staining. This
fluorescent dye intercalates in DNA and RNA, but only in cells having reduced
membrane integrity.
Parts of the total cell count analysis as well as the DOPE-FISH and PI analyses were
performed in the context of the Diploma thesis of C. Krumrei (2010).
For the total cell count and the DOPE-FISH analyses, the leachate of the LBR was
sampled on day 7 of the 21-day fermentation at each temperature regime due to the
fact that an active community should have been established after one week of
fermentation. The samples were fixed immediately. To this end, 500 µL of the leachate
were washed twice with 1x PBS (pH 7.4), centrifuged at 3.500 x g for 15 min and
incubated in 2.8% formaldehyde (Carl Roth GmbH & Co.KG, Karlsruhe, Germany) at
4 °C over night. The samples were centrifuged as described above and the pellets
were washed two times with 1x PBS (pH 7.4). The pellets were resuspended in equal
parts of 1x PBS (pH 7.4) and 96% ethanol (AppliChem GmbH, Darmstadt, Germany).
All samples were stored at -20 °C until further processing.
For both analyses, the fixed samples were washed again with 1x PBS (pH 7.4) and
centrifuged at 15,000 x g for 5 min. Each sample, resuspended in 1x PBS (pH 7.4),
was sonicated (Sonoplus GW2070; Bandelin, Berlin, Germany) two times with 50%
power (about 35 W) at pulse level 1 for 30 s to destroy cell aggregates according to
Nettmann and coworkers (2010). Triplicates of the samples (10 µL each) were
transferred on a Teflon-coated slide (Gerhard Menzel GmbH, Braunschweig,
Germany), which was covered with 0.1% gelatine and 0.01% CrK(SO4)2 (AppliChem
GmbH, Darmstadt, Germany).
To analyze the total cell counts, the samples were stained using 5 µL Citifluor AF1
(Citifluor Ltd, London, UK) antifading reagent and 0.2 µL DAPI (33 µg mL-1, Carl Roth
GmbH & Co.KG, Karlsruhe, Germany). Stained microbial cells were detected on the
Olympus BX51 fluorescent microscope (Olympus, Hamburg, Germany) at a
44
magnification of 600 using the filter set WU for DAPI detection. Images of the samples
were taken with a digital Olympus DP72 camera applying the software cell F (Olympus,
Hamburg, Germany). For each triplicate of the samples, about 1,000 stained cells were
counted from randomly taken images. The calculation of the total cell count was
performed according to Raizada (2004) with the following formula:
total cell count (mL-1) = Awell / Acount * Xm * v,
where Awell is the area of each well, Acount is the area of each counted image, Xm is the
average cell number per image and v is the dilution factor.
For DOPE-FISH analysis, the samples on Teflon-coated slides were permeabilized
with 10 µL freshly prepared lysozyme solution (10 mg mL-1 lysozyme; 50 mM EDTA,
pH 8; 0.3 M Tris/HCl, pH 8) for an improved uptake of hybridization probes. The
samples were incubated at 37 °C for 30 min in a 50 mL reaction tube together with wet
paper to create a humid atmosphere. The slide was rinsed briefly with distilled H2O and
dehydrated by a stepwise incubation (3 min each) in 50%, 80% and 96% ethanol
(AppliChem GmbH, Darmstadt, Germany). Hybridization of the samples was performed
on dried slides. The hybridization buffer was prepared with a stringency of 50% (0.9 M
NaCl; 0.02 M Tris/HCl; 10x sterile Denhardt reagent - containing 1% FicoII, 1% PVPP,
1% BSA; 50% formamide and 0.01% SDS; from AppliChem GmbH, Darmstadt,
Germany and Carl Roth GmbH & Co. KG, Karlsruhe, Germany). To each sample well,
9 µL of the hybridization buffer was applied together with 1 µL Arch915FITC (50 ng L-1,
Table 2.1) for the detection of archaea, 1 µL EUB338Cy3 I (50 ng L-1, Table 2.1) and
1 µL of a 1:1 solution of EUB338 Cy3 II and III (each 50 ng L-1, Table 2.1) for the
detection of bacteria. Furthermore, 1 µL of the nonsense probe NON338Cy3 or FITC (each
50 ng L-1, Table 2.1) having no binding site in the 16S rRNA of Bacteria and Archaea
as well as µL of the probe UNIV1390Cy3 or FITC (each 50 ng L-1, Table 2.1) binding to
both Bacteria and Archaea were analyzed in order to detect false positive hybridization
and to determine the hybridization rate, respectively. Then, the slide was incubated in a
50 mL reaction tube at 46 °C for 2 h in a hybridization oven (HL 2000, UVP Ltd,
Cambridge, UK) together with paper moistened with the residual hybridization buffer.
After hybridization, the slide was washed carefully in a 48 °C preheated washing buffer
(0.02 M NaCl; 0.02 M Tris/HCl; 5 mM EDTA; 0.01% SDS; AppliChem GmbH,
Darmstadt, Germany) for 10 to 15 min to remove unbound probes from the samples.
45
Finally, the slide was dipped into ice cold distilled H2O for 3 s, dried and stored at
-20 °C in the dark until further processing.
As positive controls various pure, actively growing cultures were used: E. coli (DSM
1116), Cl. tyrobutyricum (DSM 2637) for bacterial detection and Methanosaeta concilii
(DSM 6752; DSMZ GmbH, Braunschweig, Germany) and Methanospirillum hungateii
Mh1 (Department of Microbial Ecology, Limnology and General Microbiology,
University of Konstanz) for archaeal detection.
The detection of fluorescently labeled cells was performed on the Olympus BX51
fluorescent microscope (Olympus, Hamburg, Germany) at a 600-fold magnification
using the filter set WU (excitation 330 - 385 nm, emission >420 nm) for DAPI detection,
the filter set NIBA (excitation 470 - 495 nm, emission 510 - 550 nm) for FITC detection
and the filter set Cy3 (excitation 510 - 560 nm, emission 573 - 647 nm) for Cy3
detection. Images of the samples were taken with a digital Olympus DP72 camera
applying the software cell F (Olympus, Hamburg, Germany) for each filter set.
Afterwards, the fluorescently labeled archaeal and bacterial cells were counted from
randomly taken images for about 1,000 DAPI stained cells for each triplicate of
samples as described above.
For PI analysis, fresh leachate samples of the LBR 1, 2 and 3 (LBR 65 °C) were
analyzed in triplicate on day 0, 7, 15 and 21. In detail, 100 µL of leachate was washed
with 1x PBS (pH 7.4) and centrifuged at 15,000 x g for 5 min. The pellet was
resuspended in 1 mL 1x PBS (pH 7.4) and 20 µL of PI solution (1 mg mL-1, Invitrogen,
Darmstadt, Germany) was added. The sample was incubated for 10 min in the dark.
Briefly, the sample was centrifuged at 15,000 x g for 5 min and the pellet was
resuspended in 1 mL 1x PBS (pH 7.4). 10 µL of the sample were applied per well on a
10-well Teflon-coated slide. The slide was dried and also stained with DAPI as
described above. Microscopical detection and counting of cells was performed as
described above.
46
Table 2.1 Primers and probes used for the polyphasic analyses of the two-phase biogas systems
Primer Sequence (5'-3') Target Amplicon
size
Target site
according to
E. coli Method Reference Supplier
27f-Cy5 Cy5-AGAGTTTGATCMTGGCTCAG Bacteria approx. 899 8 - 27 TRFLP Lane, 1991; Sipos et al., 2007 b
926MRr CCGTCAATTCMTTTRAGTTT 926 - 947 Weisburg et al., 1991; Despres et al., 2007 b
Ar109f-Cy5 Cy5-ACKGCTCAGTAACACGT Archaea approx. 803 109 - 125 TRFLP Großkopf et al., 1998 b
Ar912r CTCCCCCGCCAATTCCTTTA 912 - 931 Lueders & Friedrich, 2000 b
27f AGAGTTTGATCMTGGCTCAG Bacteria approx. 899 8 - 27 rrs gene
library
Lane, 1991; Sipos et al., 2007 b
926MRr CCGTCAATTCMTTTRAGTTT 926 - 947 Weisburg et al., 1991; Despres et al., 2007 b
Ar109f ACKGCTCAGTAACACGT Archaea approx. 803 109 - 125 rrs gene
library
Großkopf et al., 1998 b
Ar912r CTCCCCCGCCAATTCCTTTA 912 - 931 Lueders & Friedrich, 2000 b
T7 TAATACGACTCACTATAGGG pGem®-T vector variable - rrs gene
library
Promega GmbH c
SP6 ATTTAGGTGACACTATAG Promega GmbH c
304f GACGCATGTTGGACATATTAAAGC 194 - 219 this study b
304TaqMan 6-Fam-TAAAGGCCCACCAAGGCGACGA-Tamra TRF 304 151 281 - 302 qPCR this study b
304r AGTTTGGGCCGTGTCTCAGT 341 - 360 this study b
Bac338f ACTCCTACGGGAGGC 338 - 354 Yu et al., 2005 c
Bac516TaqMan TGCCAGCAGCCGCGGTAATA Bacteria 468 516 - 536 qPCR Yu et al., 2005 c
Bac805r GACTACCAGGGTATCTAATC 785 - 805 Yu et al., 2005 c
EUB388 (pB-00159)a CY3-GCTGCCTCCCGTAGGAGT-CY3 Bacteria - - DOPE-FISH Amann et al., 1990 d
EUB388 II (pB-00160)a CY3-GCAGCCACCCGTAGGTGT-CY3 Bacteria - - DOPE-FISH Daims et al., 1999 d
EUB388 III (pB-00161)a Cy3-GCTGCCACCCGTAGGTGT-Cy3 Bacteria - - DOPE-FISH Daims et al., 1999 d
Arch915 (pB-00027)a FITC-GTGCTCCCCCGCCAATTCCT-FITC Archaea - - DOPE-FISH Stahl & Amann, 1991 d
UNIV1390 (pB-00327)a FITC-GACGGGCGGTGTGTACAA Bacteria & Archaea - - FISH Zheng et al., 1996 d
UNIV1390 (pB-00327)a Cy3-GACGGGCGGTGTGTACAA Bacteria & Archaea - - FISH Zheng et al., 1996 d
NON338 (pB-000243)a FITC-ACTCCTACGGGAGGCAGC No organism - - FISH Wallner et al., 1993 d
NON338 (pB-000243)a Cy3-ACTCCTACGGGAGGCAGC No organism - - FISH Wallner et al., 1993 d
a, accession number as indicated by probeBase (Loy et al., 2007)
b, Biomers.net GmbH, Ulm, Germany; c, Eurofins MWG GmbH, Ebersberg, Germany; d, Metabion International AG, Martinsried, Germany
-, no amplification
47
3 RESULTS
3.1 Operationofthetwophaseleachbedbiogassystems
The two-phase biogas systems have been operated throughout the whole experiment
by M. Schönberg with the support of C. Joost. The temperature of each LBR of the
three identically constructed biogas systems was increased stepwise from 55 up to
75 °C by 5 °C increments, while the temperature of each AF was kept constant at
55 °C throughout the experiment. Each temperature regime consisted of three 21-day
fermentations to allow the microbial community to adapt to the increased temperatures.
To each system, rye silage with an average pH of 4.20 ± 0.06, an ODM content of
17.55 ± 0.73% and a COD of 204.88 ± 9.75 g per kg (Table 3.1) was supplied. In
contrast to that, the straw material used had higher pH, ODM and COD values
(Table 3.1). After 21 days of fermentation, the digestate showed reduced COD values
(Table 3.1), resulting from the degradation of the substrate. Process parameters, such
as pH, ODM, COD, concentration of intermediates and total ammonia nitrogen, were
provided by the analytical chemistry work group at the ATB. Biogas and methane yield
and degradation rate of biomass was calculated and provided by M. Schönberg for
further data interpretation
During the increase in temperature in the LBR, the pH in the three biogas systems was
nearly constant with values between 7.52 ± 0.36 (LBR) and 7.95 ± 0.11 (AF;
Table 3.2). Furthermore, the concentration of total ammonia nitrogen and free
ammonia in the AFs of the three systems ranged from 0.85 ± 0.05 to 1.00 ± 0.11 g
per L and from 0.21 ± 0.02 to 0.27 ± 0.04 g per L, respectively, during the whole
experiment (Table 3.2), indicating no strong alterations.
48
Table 3.1 Analytical parameters (ODM, pH and COD) of substrates and digestates
pH ODM (%) COD (g kg-1)
Rye silagea 4.20 ± 0.06 17.55 ± 0.73 204.88 ± 9.75
Winter barley straw 7.58 81.39 894.84
Wheat straw 7.42 81.89 875.80
Compost material 8.17 14.55 212.22
Digestateb 9.17 ± 0.16 11.14 ± 1.31 139.14 ± 26.17
a, average of eight rye silage samples
b, average of three biogas reactor systems at each temperature regime
However, several parameters, such as the biogas yield (Table 3.3), differed markedly
during the temperature increase in the three biogas systems. The amount of biogas at
an LBR temperature of 55, 60 and 65 °C was almost constant with values from 605.09
± 2.73 to 618.90 ± 4.87 L per kgoS. At 70 °C, the biogas yield decreased by 30% in the
three biogas systems. After the bioaugmentation with compost (LBR 70 °C), an
increase in the biogas yield of 15 to 21% was obtained for the three reactor systems,
leading to average biogas yields of 516.23 ± 2.93 L per kgoS. However, the positive
effect of the bioaugmentation with compost on the biogas yield did not persist at higher
LBR temperatures of 75 °C. Here, the average biogas yield was reduced to
397.20 ± 16.65 L per kgoS.
The mean methane concentration of the biogas differed between 48.99 ± 6.45% and
61.99 ± 4.47% during temperature increase in the LBRs (Table 3.3). Regarding the two
phases of the subject biogas systems separately, certain differences in the methane
yield and content occurred between LBR and AF (Table 3.3). In all experimental runs,
the methane content obtained for the AF ranged between 70 and 81% and for the LBR
between 21 and 46%.
Simultaneously with the changes in the biogas yield at 70 °C, the degradation rate of
the substrate also decreased in the three reactor systems (Table 3.3). Between 55 and
65 °C, a high degradation of ODM was achieved with values between 80.31 ± 1.39%
and 65.40 ± 5.23%. Then, the degradation rate changed to 39.80 ± 3.60% at 70 °C, to
48.95 ± 2.85% after bioaugmentation at 70 °C and to 32.47 ± 3.05% at 75 °C also
indicating a positive effect of the bioaugmentation with compost, which could not
persist at higher temperatures.
The impact of the temperature increase in the LBRs on the intermediates produced
was shown using the first reactor system as example (Figure 3.1). However, the other
49
two reactor systems revealed similar concentrations of the intermediates. At an LBR
temperature of 55 to 65 °C, the maximal acetic acid concentration in the leachate
samples was relatively high with 6.69 g per L, whereas the other intermediates such as
n-butyric acid (0.96 g L-1), propionic acid (1.32 g L-1) and valeric acid (0.18 g L-1)
showed lower maximum concentrations. Above 65 °C, the maximal concentrations of
acetic acid (Figure 3.1 B) and n-butyric acid (Figure 3.1 B) decreased strongly to
4.65 g per L and 0.25 g per L, respectively. In contrast, the propionic acid concentration
still reached maximum concentrations of 0.93 g per L (Figure 3.1 D), showing less
reduction. The concentration of the valeric acid remained nearly constant with maximal
values of 0.14 g per L (Figure 3.1 E).
However, the highest concentration of the intermediates was measured during the first
nine days of the 21-day fermentation independent of the LBR temperature regime.
Interestingly at 70 °C, the degradation of acetic acid was prolonged in comparison to
lower LBR temperatures. This effect was reduced after the bioaugmentation, but
reappeared at 75 °C. In contrast to the carboxylic acids, the concentration of alcohols,
i.e. ethanol and propanol, was slightly increased by the temperature increase in the
LBR (Figure 3.1 F). At 75 °C, an alcohol concentration with a mean value of 0.15 g
per L was measured during the first five days of the fermentation. At lower
temperatures, the alcohol concentration was below the detection limit of 0.02 g per L in
most cases.
50
Duration of fermentation (d)
0 2 4 6 8 10 12 14 20
CVFA (g L-1)
0
2
4
6
8LBR 55°C
LBR 60°C
LBR 65°C
LBR 70°C
LBR 70°C*
LBR 75°C
Duration of fermentation (d)
0 2 4 6 8 10 12 14 20
Cacetic acid (g L-1)
0
2
4
6
8LBR 55°C
LBR 60°C
LBR 65°C
LBR 70°C
LBR 70°C*
LBR 75°C
Duration of fermentation (d)
0246810121420
Cn-butyric acid (g L-1)
0,0
0,2
0,4
0,6
0,8
1,0
1,2
1,4 LBR 55°C
LBR 60°C
LBR 65°C
LBR 70°C
LBR 70°C*
LBR 75°C
Duration of fermentation (d)
0 2 4 6 8 10 12 14 20
Cpropionic acid (g L-1)
0,0
0,2
0,4
0,6
0,8
1,0
1,2
1,4 LBR 55°C
LBR 60°C
LBR 65°C
LBR 70°C
LBR 70°C*
LBR 75°C
Duration of fermentation (d)
0 2 4 6 8 10 12 14 20
Cvaleric acid (g L-1)
0,0
0,2
0,4
0,6
0,8 LBR 55°C
LBR 60°C
LBR 65°C
LBR 70°C
LBR 70°C*
LBR 75°C
Duration of fermentation (d)
0 2 4 6 8 10 12 14 20
Cethanol, propanol (g L-1)
0,0
0,2
0,4
0,6
0,8 LBR 55°C
LBR 60°C
LBR 65°C
LBR 70°C
LBR 70°C*
LBR 75°C
AB
CD
EF
Figure 3.1 Concentration of total VFA (A), acetic acid (B), n-butyric acid (C),
propionic acid (D), valeric acid (E) and ethanol and propanol as sum (F)
shown for the first leach-bed reactor system during the 21-day
fermentation process at all specified temperature regimes
51
Table 3.2 Process parameters of the three two-phase biogas systems
Temperature
LBR reactor pHa (LBR; AF) NH4-Na (g L-1)
(AF)
NH3a,b
(g L-1) (AF)
Maximum VFAa
(gHAc eq. L-1) (LBR; AF)
Maximum CODa
(g L-1) (LBR; AF)
55 °C 7.52 ± 0.36; 7.80 ± 0.19 1.00 ± 0.11 0.24 ± 0.06 7.73 ± 0.64; 5.03 ± 0.71 17.89 ± 1.41; 14.89 ± 1.29
60 °C 7.56 ± 0.40; 7.95 ± 0.11 0.94 ± 0.09 0.27 ± 0.04 7.23 ± 0.71; 4.20 ± 1.26 17.65 ± 1.32; 14.68 ± 2.19
65 °C 7.43 ± 0.48; 7.89 ± 0.09 0.98 ± 0.06 0.24 ± 0.05 8.87 ± 0.62; 4.07 ± 1.88 19.42 ± 1.34; 12.37 ± 1.79
70 °C 7.54 ± 0.27; 7.88 ± 0.07 0.97 ± 0.11 0.24 ± 0.05 6.38 ± 0.87; 1.34 ± 0.73 18.93 ± 1.13; 9.00 ± 1.54
70 °C* 7.58 ± 0.24; 7.91 ± 0.06 0.99 ± 0.18 0.25 ± 0.06 5.12 ± 0.13; 1.01 ± 0.16 17.49 ± 1.19; 9.50 ± 0.87
75 °C 7.53 ± 0.25; 7.89 ± 0.09 0.85 ± 0.05 0.21 ± 0.02 3.99 ± 0.29; 0.74 ± 0.05 18.73 ± 0.87; 8.69 ± 0.67
a, average values and standard deviation of the three biogas systems at one LBR temperature regime
b
, free ammonia was calculated accoding to Anthonisen and coworkers (1976)
70 °C*, analysis after bioaugmentation with compost; LBR, leach-bed reactor; AF, anaerobic filter reactor
Table 3.3 Biogas and methane yield and degradation rate of the three two-phase biogas systems
Temperature
LBR reactor
Total biogas
yielda,b
(L kgOS-1)
Total
methane
contentb (%)
Methane contentb (%) Methane yielda,b (L kgOS-1) Degradation
rate of ODMb,d
(%)
LBR AF LBR AF
55 °C 618.90 ± 4.87 50.54 ± 0.48 44.78 ± 0.99 69.99 ± 1.67 213.62 ± 11.61 99.17 ± 12.23 80.31 ± 1.39
60 °C 605.09 ± 2.73 54.62 ± 0.80 46.36 ± 0.95 72.89 ± 2.45 193.17 ± 3.07 137.32 ± 3.50 78.10 ± 0.85
65 °C 610.09 ± 1.45 49.70 ± 0.35 40.40 ± 0.44 77.27 ± 0.77 184.28 ± 3.18 118.95 ± 4.17 65.40 ± 5.23
70 °C 427.16 ± 17.71 48.99 ± 6.45 32.32 ± 8.72 79.15 ± 0.92 87.68 ± 19.66 121.20 ± 15.01 39.80 ± 3.60
70 °C* 516.23 ± 2.93 49.25 ± 1.17 34.80 ± 1.95 77.92 ± 0.85 119.42 ± 6.28 134.79 ± 3.89 48.95 ± 2.85
75 °C 397.20 ± 16.65 61.99 ± 4.47 21.00 ± 9.90 80.80 ± 1.21 28.55 ± 6.95 202.14 ± 19.51 32.47 ± 3.05
a, gas yield normalized to 0 °C and 1,013 hPa
b, average values and standard deviation of the three biogas systems at one LBR temperature regime
d
, degradation rate of ODM was corrected by the non-degradable lignin fraction
70 °C*, analysis after bioaugmentation with compost; LBR, leach-bed reactor; AF, anaerobic filter reactor
52
3.2 Analysisofbacterialandarchaealrrsgenesequences
3.2.1 BacterialcommunitycompositionintheLBR
The bacterial community composition was analyzed by two rrs gene libraries
constructed from samples derived from the LBR at 55 and 75 °C. A total number of 76
(LBR 55°C) and 74 (LBR 75°C) rrs gene sequences were obtained, representing 30
and 17 bacterial OTUs, respectively.
Not only the number of OTUs, but also the composition of the bacterial community
changed strongly during the temperature increase in the LBR. A pairwise comparison
of the rrs gene libraries results revealed Bray-Curtis similarity values of 0.57 calculated
for the OTUs and 0.21 calculated for the rrs sequences. This supports the findings of
strong alterations within the bacterial community analyzed at an LBR temperature of
55 and 75 °C.
OTUs (%)
55°C
75°C
rrs sequences (%)
OTUs (%) rrs sequences (%)
Clostridia
(67%)
Bacilli (3%)
Cytophagia (3%)
Bacteroidia (7%)
Unclassified bacteria (7%)
Synergistia (3%)
Gammaproteobacteria (7%)
Actinobacteria (3%)
Clostridia
(83%)
Bacilli (1%)
Cytophagia (1%)
Bacteroidia (3%)
Unclassified bacteria (3%)
Synergistia (1%)
Gammaproteobacteria (7%)
Actinobacteria (1%)
Clostridia
(59%)
Bacilli (6%)
Cytophagia (6%)
Bacteroidia (6%)
Unclassified bacteria (6%)
Thermotogae (12%)
Betaproteobacteria (6%)
Clostridia
(16%)
Bacilli (1%)
Cytophagia (3%)
Bacteroidia (62%)
Unclassified bacteria (7%)
Thermotogae (9%)
Betaproteobacteria (1%)
Figure 3.2 Results of the bacterial rrs gene sequence analyses of the LBR at 55 and
75 °C. The distribution of OTUs and rrs sequences obtained is indicated
at class level.
53
During temperature increase in the LBR, the bacterial community changed from being
Clostridia-dominated towards a more even distribution of Bacteroidia, Clostridia and
Thermotogae. More specifically, members of the Clostridia were prevalent with 67% of
OTUs and 83% of rrs sequences at 55 °C (Figure 3.2). At 75 °C, members of the
Clostridia were strongly reduced accounting for 59% of OTUs and 16% of rrs
sequences. Simultaneously, the number of rrs sequences assigned to the Bacteroidia
class increased strongly from 3% to 62% after the temperature increase to 75 °C in the
LBR.
Additionally, the results also showed that rrs sequences assigned to the Actinobacteria,
Gammaproteobacteria and Synergistia classes, which were detected at lower levels at
55 °C, were absent at 75 °C (Figure 3.2). Contrarily, rrs sequences related to the
Betaproteobacteria and particularly to the Thermotogae, which were absent at 55 °C,
showed a relative abundance at 1% and 9% at 75 °C, respectively (Figure 3.2).
3.2.2 ArchaealcommunitycompositionintheAF
The archaeal community composition was determined by the construction of an rrs
gene library analyzing the packing’s biofilm derived from the AF at an LBR temperature
of 55 °C. A total number of 88 rrs sequences were analyzed, representing 12 archaeal
OTUs.
About 50% of OTUs were assigned to the Methanobacteriales order and within this
order to the Methanobacterium and Methanothermobacter genera (Figure 3.3). Further,
Methanomicrobiales and Methanosarcinales were also identified with 33% and 17% of
OTUs, respectively. The analysis of the rrs sequences revealed Methanobacteriales
(52% of rrs sequences), followed by Methanosarcinales (30%) and
Methanomicrobiales (18%) as prevalent archaeal orders (Figure 3.3). At the genus
level, Methanothermobacter (38% of rrs sequences) and Methanoculleus (17%)
followed by Methanobacterium, Methanosarcina and Methanosaeta (each 15%) were
detected.
54
OTUs (%) OTUs (%)
rrs sequences (%)
Order Genus
rrs sequences (%)
Methanobacteriales
(52%)
Methano-
microbiales
(18%)
Methano-
sarcinales
(30%)
Methano-
bacterium
(15%)
Methano-
thermo-
bacter
(38%) Methano-
culleus
(17%) Methanospirillum (1%)
Methanosaeta (15%)
Methanosarcina (15%)
Methano-
bacterium
(25%)
Methano-
thermo-
bacter
(25%)
Methano-
culleus (25%) Methanospirillum (8%)
Methanosaeta (8%)
Methanosarcina (8%)
Methanobacteriales
(50%)
Methano-
microbiales
(33%)
Methano-
sarcinales
(17%)
Figure 3.3 Results of the archaeal rrs gene sequence analyses of the packing’s
biofilm derived from the AF at an LBR temperature of 55 °C. The
distribution of OTUs and rrs sequences is indicated at order and genus
level.
3.2.3 Statisticalanalysesoftherrsgenelibraries
The three rrs gene libraries were statistically analyzed calculating different parameters
(Table 3.4). Two parameters, Good’s coverage and the Chao-I estimator, were
calculated in order to evaluate the coverage of microbial diversity. The results showed
high coverage values (Good, 1953) for the bacterial rrs gene library at 75 °C and for
the archaeal rrs gene library accounting for 85 and 95%, respectively (Table 3.4). Only
the bacterial rrs gene library, constructed for the sample derived from the LBR at 55 °C,
showed a lower coverage with 68%.
The Chao-I estimator values (Chao, 1987), which indicate the extrapolated number of
OTUs, supported the findings of the coverage results (Table 3.4). The difference
between the extrapolated OTU values and the identified OTUs for the bacterial rrs
gene library at 55 °C differed strongly, supporting the lower coverage of the underlying
55
microbial community. The Chao I values calculated for the two other rrs gene libraries
(Bacteria, 75 °C and Archaea, 55 °C) were more similar to the OTU values obtained.
Table 3.4 Statistical parameters of the archaeal and bacterial rrs gene libraries
OTUs
identified
Good's
coveragea
(%)
Shannon
diversity
indexa
Simpson
diversity
index
Evenness Chao-I estimatora
Bacteria LBR 55 °C 30 68.42 2.74 0.89 0.51 168 (73 - 476)b
Bacteria LBR 75 °C 17 85.14 1.63 0.60 0.30 31 (20 - 80)b
Archaea AF 55 °C 12 95.45 1.94 0.81 0.58 14 (12 - 28)b
a, based on the OTUs obtained
b
, confidence values (95% bootstrap)
To estimate the diversity of the bacterial and archaeal community, the Shannon
(Shannon & Weaver, 1963) and Simpson indices (Simpson, 1949) were calculated
(Table 3.4). Both indices showed that the highest diversity could be expected for the
bacterial community at an LBR temperature of 55 °C. Furthermore, the parameter
Evenness that was also calculated gave information about the distribution of the OTUs
obtained (Lloyd & Ghelardi, 1964; Hill, 1973). A value of 1 indicates an equal
distribution of OTUs. A low value indicates a strong prevalence of a specific OTU. In
the case of the bacterial rrs gene library at 75 °C, one OTU (Bacteroidia) was strongly
increased as indicated by the low Evenness value (Table 3.4).
56
3.3 Characterizationandmonitoringofthemicrobialcommunity
byTRFLPanalyses
3.3.1 EstablishmentandoptimizationofTRFLPassaysforbacterial
andarchaealcommunityanalyses
Application of different DNA extraction methods
Four different methods were applied to evaluate the bias caused by DNA extraction.
Two DNA extraction kits were used, which both induced cell lysis by mechanical force
(sample FP and PS), whereas the third protocol, a step by step protocol, was based on
chemical (SDS) and enzymatic (lysozyme) cell lysis (sample ST). The fourth protocol,
step by step DNA extraction with an additional beating step, combined the chemical
and enzymatic lysis with a mechanical treatment (sample STB).
All methods resulted in DNA, which was applicable for rrs gene amplification and
downstream bacterial TRFLP analysis. A pairwise comparison of the TRFLP profiles
obtained from the duplicates of each DNA extraction revealed rather high Bray-Curtis
similarity values of 0.91 to 0.94 indicating reproducible results. Only the duplicate
samples extracted by the PowerSoil® DNA Isolation Kit showed a slightly lower
Bray-Curtis similarity with 0.84 (Table 3.5).
Table 3.5 Bray-Curtis similarity values of the bacterial TRFLP profiles of four DNA
extraction methods. All samples were analyzed in duplicate.
ST STB FP PS
ST 0.94 0.73 ± <0.01 0.81 ± <0.01 0.75 ± 0.05
STB 0.73 ± <0.01 0.94 0.79 ± 0.03 0.82 ± 0.01
FP 0.81 ± <0.01 0.79 ± 0.03 0.91 0.78 ± 0.01
PS 0.75 ± 0.05 0.82 ± 0.01 0.78 ± 0.01 0.84
ST, step by step DNA extraction protocol; STB, step by step DNA
extraction protocol with beating; FP, FastDNA® Spin Kit for Soil; PS,
PowerSoil® DNA Isolation Kit
57
Furthermore, the comparison of TRFLP profiles derived from the four DNA extraction
methods showed only slight differences, resulting in rather high Bray-Curtis similarity
values between 0.73 and 0.82 (Table 3.5).
These findings were also supported by analyzing the stacked bar diagram of the
bacterial TRFLP profiles obtained from the different DNA extraction protocols
(Figure 3.4). Here differences, particularly between the PS duplicates, were also
detected. Furthermore, the relative abundance of TRFs varied between the four DNA
extraction methods. For instance, TRF 1 showed a higher relative abundance when
extracted with a DNA protocol, including a mechanical treatment step (Figure 3.4).
Despite these differences in the relative abundance of individual TRFs, all four DNA
extraction methods resulted in the same 16 prevalent TRFs.
DNA extraction protocol
ST STB FP PS
Relative abundance of TRFs (%)
0
20
40
60
80
100
TRF1TRF1 TRF1 TRF1TRF1TRF1 TRF1 TRF1
Figure 3.4 Bacterial TRFLP profiles obtained by four different DNA extraction
protocols using digestate of the two-phase biogas reactor. TRFs above
2% of the total fluorescence intensity are shown. All samples were
analyzed in duplicate. ST, step by step DNA extraction protocol; STB,
step by step DNA extraction protocol with beating step; FP, FastDNA®
Spin Kit for Soil; PS, PowerSoil® DNA Isolation Kit
Finally, the STB protocol was used for metagenomic analysis, resulting in high
molecular weight DNA with high concentrations. The FastDNA® Spin Kit for Soil was
applied for the other methodical approaches, resulting in reproducible TRFLP profiles
and also in higher DNA concentrations than the PowerSoil® DNA Isolation Kit.
58
Optimization of PCR protocols and evaluation of primer sets
Different bacterial and archaeal reverse primer sets were tested by applying the TRFLP
approach. This resulted in rather similar bacterial and archaeal TRFLP profiles
independent of the primer sets applied. For instance, a comparison of TRFLP profiles
obtained from two biogas samples amplified with each of the five bacterial primer sets
(27f and 926r, 926Rr, 926RMr, 16S, 1492r) revealed high Bray-Curtis similarities of
0.88 ± 0.02 and 0.86 ± 0.04, respectively. Furthermore, the prevalent TRFs above 1%
of the total fluorescence intensity were detected with all bacterial primer sets
mentioned. However, the primer-set (27f, 926MRr) with two degenerated bases in the
reverse primer was used as standard due to the higher number of potential targets in
comparison to e.g. the reverse primer 926r without degenerated basis as assumed
after applying the RDP probe match tool (Cole et al., 2009).
The comparison of the TRFLP results obtained after applying three archaeal primer
sets (Ar109f and Ar912, Ar915, Ar958) also showed similar results (Bray-Curtis value
of 0.80 ± 0.04) independent of the primer set used. The number of TRFs with
fluorescence intensities above 1% varied slightly between 12 and 14. The primer set
used in this study (Ar109f, Ar912r) resulted in the highest number of archaeal TRFs.
Additionally, the impact of cycle numbers was also evaluated, applying the bacterial
TRFLP assay. The results showed an increase in TRFs of 29%, comparing 15 and 35
PCR cycles. Due to the fact that an increased cycle number may increase the risk of
PCR biases, such as raised chimeric sequences (v. Wintzingerode et al., 1997), 25 and
28 PCR cycles were chosen for the bacterial and archaeal DNA amplification,
respectively.
Evaluation of restriction enzymes
For an in-depth characterization of the microbial community, a good resolution of the
inherent diversity resulting in a diverse and heterogenic TRFLP profile is essential. To
evaluate the optimal restriction enzyme(s) for the biogas samples analyzed in this
study, different enzymes and their combinations were tested.
For example, the results of three different enzymes (HaeIII, Hin6I and MspI) and their
combinations are shown representing the bacterial community (Figure 3.5). The single
enzyme digests resulted in a lower resolution of the bacterial community (total number
59
of TRFs: 42 ± 8.5) than the approach with two enzymes (54 ± 3.6 TRFs). The
combination of the restriction enzymes MspI and Hin6I were identified as most valuable
showing a good resolution of the TRFLP profile and a high number of TRFs. Although
the combination of HaeIII and MspI resulted in slightly more TRFs, the resolution
obtained by using these enzymes was less due to the fact that some fragments are
separated close to each other. This impedes the interpretation and identification of
TRFs.
The optimization steps of the archaeal TRFLP assay were performed according to the
steps described for the bacterial TRFLP assay, showing the best resolution after AluI
digestion.
HaeIII and Hin6I
Size (bp)
0 100 200 300 400 500
0
10
20
30 MspI
0 100 200 300 400 500 600
HaeIII and MspI
Relative abundance of TRFs (%)
0
10
20
30 Hin6I
HaeIII
MspI and Hin6I
0
10
20
30
Figure 3.5 TRFLP profiles obtained by the use of different restriction enzymes and
their combinations analyzing the bacterial community derived from the
digestate. The average TRFLP profiles of duplicate samples are shown.
Reproducibility of the TRFLP analyses
All TRFLP analyses were conducted in triplicate samples based on three independent
DNA extractions of the same biogas reactor sample. To give an idea of the
reproducibility within the triplicates measured, the prevalent TRFs with at least 2% of
60
the total fluorescence intensity are shown with standard deviations as example
(Figure 3.6).
The standard deviation of the triplicate samples varied between 0.1 and 5.8%, which
was not obviously affected by the intensity or by the size of TRFs. Due to the fact that
triplicate samples consisted of independent DNA extractions, PCR, digestion and
separation steps, these standard deviations were rather low and proved reliability and
reproducibility of the method applied. The triplicate TRFLP profiles are shown as
average values without standard deviation in the result section of this study to keep the
clarity of the diagrams.
TRF1
TRF2
TRF3
TRF4
TRF5
TRF6
TRF7
TRF8
TRF9
TRF10
TRF11
TRF12
TRF13
TRF14
TRF15
Relative abundance of TRFs (%)
0
5
10
15
20
40
60
LBR 1
LBR 2
LBR 3
Figure 3.6 Triplicate TRFLP samples indicated with standard deviation. Digestate
samples of the three identically constructed LBRs were analyzed by the
TRFLP method applying the bacterial assay. TRFs above 2% of the total
fluorescence intensity are shown.
Comparison of TRFLP profiles based on TRF height and area
In the literature, the analysis of TRFLP data is based on the TRF height or area. To
evaluate the potential difference between these two data processing methods, a
comparison between TRFLP results based on the TRF height and area was conducted.
The TRFLP profiles obtained were similar, when focusing on the prevalent TRFs with
relative abundances of more than 2% of the total fluorescence intensity (Figure 3.7).
Only slight variations occurred in the relative abundance of the TRFs identified,
indicating no obvious preference for the analysis of TRF by height or area. This is also
supported by the high Bray-Curtis similarity values (0.89 - 0.94) obtained for samples
61
focused on the TRF height and area. However, due to the fact that the threshold for
linear measurement is based on the TRF height on the GeXP Genetic Analysis
System, the analysis of the TRFLP data was also conducted on the basis of the TRF
height.
sample 1 sample 2 sample 3 sample 4
Area Height Area Height Area Height Area Height
Relative abundance of TRFs (%)
0
20
40
60
80
100
Figure 3.7 Comparison of TRF profiles based on TRF area and height. Biogas
samples were analyzed by the bacterial TRFLP assay. TRFs above 2% of
the total fluorescence intensity are shown.
3.3.2 Identificationofterminalrestrictionfragments
The affiliation of TRFs (Table 3.6) was identified by constructing and analyzing
bacterial and archaeal rrs gene sequences. The bacterial TRFs, which could be
assigned to taxonomic groups, represented approximately 60 to 70% of the TRFLP
profiles (TRFs >3% of the total fluorescence intensity) depending on the sample
analyzed. Several of these TRFs showed the highest sequence similarity to members
of the Clostridia class, as indicated by NCBI BLAST search against the NCBI nr
database and by RDP classification. Members of other taxonomic groups, such as
Bacteroidia or Thermotogae, were also identified.
62
In contrast to that, the assigned archaeal TRFs represented an even higher percentage
of the TRFLP profiles accounting for up to 95%. The phylogenetic assignment of these
archaeal TRFs revealed members of the Methanobacteriales, Methanosarcinales and
Methanomicrobiales orders. In one case, the assignment of the archaeal TRFs was
ambiguous. The rrs sequences of Methanobacterium, Methanoculleus and
Methanosaeta resulted in a TRF with 107 bp length (Table 3.6).
Further, it must be noted that not all TRFs could be assigned to taxonomic groups due
to the absence of corresponding sequences in the rrs gene libraries. Furthermore, the
phylogenetic assignments of the rrs gene sequences to sequences retrieved from the
NCBI nr database showed maximum identities between 85 and 100% indicating at
least in part a distant relationship. Particularly, the assignment of the bacterial rrs
sequences to reference sequences in the NCBI nr database showed rather low
sequence identities, whereas the archaeal sequences revealed sequence identities to
database sequences of at least 97%. This is also true for the assignment of rrs
sequences applying the RDP classifier. Sequences with a sequence identity below
90% to reference sequences of the NCBI nr database, could only be assigned to higher
taxonomic levels, such as family or order, using the RDP classifier. In contrast to that,
the assignment of archaeal sequences also showed higher sequence similarities by
means of the RDP classification. However, due to the differences in the taxonomy of
NCBI and RDP, differences in the assignment of sequences occurred, notably at lower
taxonomic levels (Table 3.6).
In the following descriptions of the TRFLP results, the affiliation of TRFs is based on
the NCBI taxonomy due to the fact that the results of the metagenomic analyses
obtained from CARMA (phylogenetic assignment and functional analysis) are also
based on the NCBI taxonomy.
63
Table 3.6 Affiliation of archaeal (A) and bacterial (B) TRFs according to the NCBI and RDP taxonomy
NCBI taxonomy RDP taxonomy
TRF NCBI nr database entry with the highest similaritya Accession
number
Query
coverage
(%)b,c
Maximum
identity (%)b,c Sequence assignment using the RDP Naive Bayesian rRNA
Classifierc,d
A 107 Methanobacterium beijingense str. 4-1 AY552778 99 - 100 97 - 98 Methanobacterium 100%
Methanoculleus bourgensis str. MS2 NR_042786 99 97 - 98 Methanoculleus 100%
Methanosaeta concilii str. GP-6 CP002565 99 - 100 99 - 100 Methanosaeta 100%
A 337 Methanothermobacter wolfeii str. DSM 2970 NR_040964 100 100 Methanothermobacter 100%
A 339 Methanothermobacter crinale str. Tm2 HQ283273 99 - 100 98 - 99 unclassified Methanobacteriaceae (Methanobacterium 72%)
A 430 Methanoculleus bourgensis AB065298 99 100 Methanoculleus 100%
Methanoculleus chikugoensis str. MG62 NR_028152 99 98
Methanoculleus bourgensis str. MS2 NR_042786 99 100
A 628 Methanosarcina str. 2214B AB300208 99 99 Methanosarcina 100%
B 76 Thermoanaerobacter inferii str. AK15 EU262599 99 87 unclassified Firmicutes (Fervidicola 39%)
B 95 Bacteroidales str. 28bM GU129116 57 - 61 87 - 96 unclassified Bacteroidetes (Cesiribacter 20%)
B 139 Bacillus str. TP-84 AJ002154 99 89 unclassified Bacillales (Ureibacillus 50%)
B 141 Dethiobacter alkaliphilus str. AHT 1 NR_044205 96 88 Dethiobacter 92%
B 151 Clostridium str. 6-31 FJ808611 86 - 88 87 - 89 unclassified Clostridiales (Desulfitibacter 24 - 26%)
B 161 Clostridium orbiscindens str. 17 GU968170 93 85 unclassified Ruminococcaceae (Anaerotruncus 36%)
B 167 Clostridium str. 6-16 FJ808609 93 - 94 96 - 99 Clostridium III 96 - 100%
B 194 Fervidobacterium str. CBS-2 EF222229 99 - 100 97 - 98 Fervidobacterium 100%
B 214 Acetomicrobium faecale type str. DSM 20678 FR749980 94 - 99 94 - 98 Acetomicrobium 100%
B 217 Clostridium thermocellum str. ATCC 27405 CP000568 99 - 100 91 - 94 Clostridium III 76 - 100%
B 221 Acetomicrobium faecale type str. DSM 20678 FR749980 96 - 99 95 - 98 Acetomicrobium 100%
B 228 Haloplasma contractile str. SSD-17B NR_044362 97 86 unclassified Bacteria (Haloplasma 58%)
B 291 Clostridiales str. 24-4c HQ452852 99 85 unclassified Ruminococcacae (Saccharofermentans 19%)
B 304 Defluviitalea saccharophila str. LIND6LT2 HQ020487 98 89 - 91 unclassified Lachnospiraceae (Sporobacterium 38%)
a, nucleotide BLAST search against the NCBI non-redundant (nr) database without uncultured environmental sequences using the megablast algorithm (last access July 2012)
b, values indicated by BLAST analysis against NCBI nr database
c, ranged values are obtained by analyzing mutiple rrs gene sequences
d, RDP Classifier version 2.5 (May 2012), a confidence value of at least 80% was used for phylogenetic assignment, genera indicated in brackets had a confidence value of ≤ 80%
64
3.3.3 Similarityofthebacterialcommunityinthethreebiogassystems
The similarity of the bacterial community within the three biogas systems was
compared at the beginning of the experiment (LBR 55 °C) by analyzing digestate
samples with the TRFLP method.
The comparison of the TRFLP profiles from samples derived from the three biogas
systems showed high similarity values of 0.92 to 0.95 (Chao-Jaccard) and 0.73 to 0.83
(Bray-Curtis; Table 3.7), indicating a similar bacterial community. Beside the quantity of
TRFs, the Bray-Curtis similarity index also includes the relative abundance of TRFs,
whereas the Chao-Jaccard similarity also includes the unseen shared species of two
samples. However, a reduced Bray-Curtis value indicates more variations within the
relative abundance of TRFs than within the number of TRFs.
Both similarity indices indicated a similar bacterial community at the beginning of the
whole experiment. Further analyses at higher LBR temperatures (LBR 70 °C) revealed
that the bacterial community changed in the same manner in all three biogas systems
during the temperature increase in the LBR (cf. 3.3.6).
Table 3.7 Similarity indices for the bacterial TRFLP profiles of digestate samples
taken from the three biogas reactors
Bray-
Curtis LBR1 LBR2 LBR3 Chao-
Jaccard LBR1 LBR2 LBR3
LBR1 0.90 ± 0.05 0.83 ± 0.04 0.73 ± 0.04 LBR1 0.97 ± 0.02 0.93 ± 0.01 0.92 ± 0.01
LBR2 0.83 ± 0.04 0.91 ± 0.03 0.75 ± 0.03 LBR2 0.93 ± 0.01 0.98 ± 0.00 0.95 ± 0.01
LBR3 0.73 ± 0.04 0.75 ± 0.03 0.82 ± 0.05 LBR3 0.92 ± 0.01 0.95 ± 0.01 0.97 ± 0.01
LBR1 - 3, leach-bed reactors of the three biogas reactor systems
65
3.3.4 Impactofthesubstrateattachedbacterialcommunityonthe
bacterialcommunityinthebiogasreactor
The bacterial community attached to the surface of the substrates (silage and straw)
was analyzed and compared with samples derived from the 21-day fermentation
process at LBR temperatures of 55, 60 and 65 °C. This analysis was performed in the
context of the Bachelor thesis of C. Nolte (2011).
The results showed that the bacterial community in the two-phase biogas reactor was
not affected by the substrate-attached bacterial community (Figure 3.8). Hence, during
the ongoing digestion process, most TRFs which were dominantly detected in
substrate samples were not identified in the leachate samples derived from different
time points of the fermentation process. After 21 days of fermentation, the bacterial
community of the digestate showed almost no similarity to the original substrate-
attached community. Only two TRFs were detected after 21 days of fermentation,
which had also been identified in the community profile of the substrate supplied.
S 0 St 0 L 0 L 2 L 4 L 7 L 14 L 21 D 21
Relative abundance of TRFs (%)
0
20
40
60
80
100
TRF 60
TRF 64
TRF 87
TRF 100
TRF 140
TRF 205
TRF 207
Figure 3.8 Impact of the substrate-attached bacterial community on the bacterial
community in the biogas system. TRFs derived from the 21-day
fermentation process at 65 °C with florescence intensities above 3% are
shown. Colored bars represent major TRFs in the substrates. S 0, silage
on day 0; St 0, straw material on day 0; L 0 - 21, leachate at different time
points of fermentation; D 21, digestate at the end of the fermentation (data
based on the Bachelor thesis of Nolte, 2011)
A pairwise comparison of the TRFLP profiles obtained from the substrate (silage and
straw) and the leachate and digestate samples based on the Bray-Curtis and
66
Chao-Jaccard indices confirmed these findings. The similarity values were extremely
low with 0.02 ± 0.01 to 0.24 ± 0.06, indicating a different bacterial community attached
to the substrate in comparison to the bacterial community in the biogas reactor system
during the 21 days of fermentation.
3.3.5 ChangesinthebacterialcommunityintheLBRduringthe
fermentationofoneloadofsubstrate
The bacterial community dynamics were monitored by analyzing the leachate and
digestate of the LBR at different time points during the 21-day fermentation process
and at the end of the fermentation for each temperature regime of the LBR, applying
the TRFLP method. Parts of this analysis were performed in the context of the
Bachelor thesis of C. Nolte (2011).
LBR temperature of 55 and 60 °C
A similar bacterial community composition was observed at the beginning (day 0) and
the end (day 21) of the fermentation process, whereas in the middle phase, strong
changes occurred, particularly after 7 days of fermentation (Figure 3.9 A, B; Appendix
Figure 7.1 A, B). This indicated that the bacterial community was subject to cyclic
alterations during the digestion of one load of substrate (silage and straw) at LBR
temperatures of 55 and 60 °C. The strongest alterations occurred slightly after the
detection of the highest VFA concentration in the leachate.
These findings of cyclic changes in the bacterial community were supported by the
results of the Bray-Curtis similarity analyses, which revealed high similarity values of
0.79 ± 0.02 (55 °C) and 0.75 ± 0.05 (60 °C) after a pairwise comparison of the leachate
derived from day 0 and 21. Furthermore, a pairwise comparison of the leachate derived
from day 0 and 7 showed low Bray-Curtis similarity values with 0.28 ± 0.03 (55 °C) and
0.48 ± 0.04 (60 °C), respectively.
All prevalent TRFs of these TRFLP profiles were assigned to members of the Clostridia
class, independent of the alterations within the bacterial community (Figure 3.9 A, B).
67
More specifically, at the beginning and at the end of the fermentation process at 55 and
60 °C, one TRF (TRF 151) dominated the TRFLP profiles. In between, the bacterial
community composition varied strongly during both the fermentation at 55 and 60 °C,
particularly after 7 days. It was here that the TRF 151 (Clostridium) decreased strongly
and TRF 217 (Clostridium) and TRF 304 (Defluviitalea) at 55 °C and TRF 167
(Clostridium) and TRF 304 (Defluviitalea) at 60 °C prevailed.
LBR temperature of 65 °C
The changes in the bacterial community in the LBR operated at 65 °C were also
tracked by TRFLP analyses. In contrast to the former analysis, the fermentation
process was also analyzed on days 4, 9 and 11 to obtain further insights into the
alterations in the bacterial community.
Here, the first strong alterations within the bacterial community occurred on day 4
(Figure 3.9 C, Appendix Figure 7.1 C). It was here that the highest concentration of
VFA was measured. A comparison of the TRFLP profiles derived from days 0 and 2
with those from day 4 showed Bray-Curtis values of 0.43 ± 0.04 and 0.55 ± 0.06,
indicating alterations within the bacterial community.
Between days 4 and 14, the bacterial community showed lower alterations. This is
supported by higher Bray-Curtis similarity values between 0.73 ± 0.10 and 0.79 ± 0.07.
Here, TRF 151 (Clostridium), which had been prevalent before, was strongly reduced
and suppressed by the emergence of TRF 167, followed by TRFs 84, 170 and 228.
Apart from TRFs 84 and 170, which were of unknown phylogenetic affiliation, TRF 167
revealed the highest sequence similarity to a member of the Clostridium genus
(Figure 3.9 C). In contrast to that, the assignment of TRF 228 showed the highest
sequences similarity to the Haloplasma genus (unclassified bacteria).
At the end of the 21-day digestion process, slight alterations within the bacterial
community reappeared. Accordingly, the pairwise comparison of samples from day 21
with those from day 14 showed slightly reduced Bray-Curtis values of 0.63 to
0.65 ± 0.05.
68
D
L 0 L 2 L 4 L 7 L 11 L 14 L 21 D 21
VFA (g L-1)
0
2
4
6
C
Relative abundance
of TRFs (%)
0
20
40
60
80
100
B
VFA (g L-1)
0
2
4
6
A
Relative abundance
of TRFs (%)
0
20
40
60
80
100
NA
NA NA NA
TRF 217, Clostridium (CP000568, 91-94%)
TRF 76, Thermoanaerobacter (EU262599, 87%)
TRF 84, unkown phylogenetic affiliation
TRF 139, Bacillus (AJ002154, 89%)
TRF 141, Dethiobacter (NR_044205, 88%)
TRF 151, Clostridium (FJ808611, 87-89%)
TRF 167, Clostridium (FJ808609, 96-99%)
TRF 228, Haloplasma (NR_044362, 86%)
TRF 291, Clostridiales (HQ452852, 85%)
TRF 304, Defluviitalea (HQ020487, 89-91%)
TRF 374, unkown phylogenetic affiliation
TRF 170, unknown phylogenetic affiliation
E
L 0 L 2 L 4 L 7 L 11 L 14 L 21 D 21
Relative abundance
of TRFs (%)
0
20
40
60
80
100
0
2
4
6
NA NA
TRF 194, Fervidobacterium (EF222229, 97-98%)
TRF 208, unknown phylogenetic affiliation
TRF 214, Acetomicrobium (FR749980, 94-98%)
TRF 221, Acetomicrobium (FR749980, 95-98%)
VFA
Figure 3.9 Bacterial community dynamics and VFA concentration during the 21-day
fermentation of one load of substrate at LBR temperatures of 55 °C (A),
60 °C (B), 65 °C (C), 70 °C (D) and 75 °C (E). TRFs above 3% of the total
fluorescence intensity are shown. Sequence accession number and
identity of the TRF affiliations are indicated in brackets (cf. Table 3.6).
L 0 - 21, leachate at different time points of fermentation; D, digestate at
the end of the fermentation; NA, not analyzed (data of B, C and D based
on the Bachelor thesis of Nolte, 2011)
69
LBR temperature of 70 and 75 °C
The bacterial community was also subject to alterations at LBR temperatures of 70 and
75 °C (Figure 3.9 D, E; Appendix Figure 7.1 D, E). The first changes already occurred
after two days of fermentation. On days 2 and 4, the highest VFA concentrations were
detected in both fermentations. The pairwise comparison of the leachate samples
derived from days 0 and 2, based on the Bray-Curtis similarity, resulted in values of
0.52 ± 0.04 (70 °C) and 0.60 ± 0.02 (75 °C). This indicates differences within the
bacterial community already at the start of the digestion process. Afterwards, the
composition of the bacterial community changed successively until the end of the
21-day fermentation.
Despite these changes, the TRF 214, followed by TRF 221, dominated the TRFLP
profiles during the whole digestion process at 70 and 75 °C (Figure 3.9 D, E). The
TRFs 214 and 221 showed the highest sequence similarity to a member of the
Bacteroidia class (Acetomicrobium). At 70 °C, TRFs, which had been prevalent at
lower LBR temperatures, were also detected in higher abundances, such as TRF 151
(Clostridium) and TRF 84 (unknown phylogenetic affiliation). At 75 °C, TRF 194, which
was assigned to the Fervidobacterium genus (Thermotogae), was detected for the first
time (Figure 3.9 D, E).
3.3.6 Impactoftemperatureincreaseonthebacterialcommunityin
theLBR
To determine the bacterial community changes in the LBR as a consequence of the
temperature increase, the bacterial TRFLP profiles were compared for two different
time points of the fermentation at each temperature regime of the LBR.
Substantial alterations in the bacterial community were observed after temperature
increase from 65 to 70 °C (Figure 3.10). Up to 65 °C, the bacterial community
consisted mostly of members belonging to the Clostridia class. Above 65 °C, members
of the Bacteroidia, Clostridia and Thermotogae classes dominated the bacterial
community within the LBR. These changes in the bacterial community were supported
by the results of a pairwise comparison based on the Bray-Curtis similarity. The
70
similarity values calculated for TRFLP profiles derived from samples at 65 and 70 °C
ranged between 0.38 ± 0.01 and 0.45 ± 0.02.
In detail, this TRFLP analysis resulted in a total number of 31 TRFs. At 55 and 60 °C,
the community profiles showed only slight variations (Figure 3.10). The prevalent TRFs
at these temperatures, i.e., TRFs 151, 167, 217 and 304 showed the highest sequence
similarity to members of the Clostridia class. At 65 °C, slight changes in the bacterial
community were detected (Figure 3.10). For instance, the relative abundance of the
TRFs 84 and 167 were increased, whereas other TRFs decreased (TRF 304) or
disappeared (TRF 217).
A
Temperature of LBR (°C)
55 60 65 70 70* 75
Relative abundance of TRFs (%)
0
20
40
60
80
100 B
Temperature of LBR (°C)
55 60 65 70 70* 75
TRF 76
TRF 84
TRF 139
TRF 141
TRF 151
TRF 167
TRF 170
TRF 194
TRF 208
TRF 214
TRF 217
TRF 221
TRF 228
TRF 291
TRF 304
TRF 374
Figure 3.10 Impact of temperature increases in the LBR (55 - 75 °C) on the bacterial
community composition in the LBR. Bacterial TRFLP profiles of leachate
samples from the LBR on day 7 (A) and 21 (B) are shown. Phylogenetic
affiliation of TRFs as indicated in Table 3.6. TRFs above 3% of the total
fluorescence intensity are shown. *, analysis after bioaugmentation with
compost
After temperature increase to 70 °C, fundamental alterations in the bacterial community
were observed (Figure 3.10). Some TRFs, such as TRFs 167 and 304 decreased
strongly or disappeared, whereas three fragments (TRFs 208, 214 and 221) were
detected for the first time. TRF 214 (Acetomicrobium) was the prevalent fragment at 70
and 75 °C with relative abundances up to 44%. Additionally at 75 °C, another TRF
appeared (TRF 194), which was assigned to the Fervidobacterium genus
(Thermotogae).
71
The TRFLP profiles of samples derived from the fermentation after the
bioaugmentation with compost at 70 °C showed two new TRFs (TRFs 147, 183) as
well as already known TRFs at higher abundances (e.g., TRFs 76, 79, 84, 86, 152, 172
and 177). Some of these TRFs were also detected in the TRFLP profiles derived from
the compost sample. However, the major TRFs of the compost sample (e.g., TRFs
151, 181, 206) were not enriched in the TRFLP profiles derived from the biogas
samples after bioaugmentation.
Impact of temperature increase on the bacterial community in the three biogas systems
The second and third two-phase biogas systems were also analyzed at LBR
temperatures of 65 and 70 °C (each on day 7) to confirm the bacterial community
changes in the LBR of the first biogas system (Figure 3.11).
Reactor and LBR temperature
1st 2nd 3rd 1st 2nd 3rd
Relative abundance of TRFs (%)
0
20
40
60
80
100
TRF 76
TRF 84
TRF 141
TRF 151
TRF 167
TRF 170
TRF 214
TRF 221
TRF 228
TRF 291
TRF 304
TRF 374
A
65°C 70°C
Figure 3.11 TRFLP profiles (A) and NMDS plot of TRFLP profiles (B) obtained from
the three biogas reactors at LBR temperatures of 65 and 70 °C (day 7).
Phylogenetic affiliation of TRFs as indicated in Table 3.6. TRFs above 3%
of the total fluorescence intensity are shown. Stress value of NMDS
analysis is 0.032.
In the two identically constructed systems, the same alterations in the bacterial
community from being Clostridia-dominated towards being dominated by members of
the Bacteroidia and Clostridia classes were detected. These alterations were also
supported by the pairwise comparison based on the Bray-Curtis similarity, showing
B
72
rather low similarity values of 0.41 ± 0.05 between TRFLP profiles derived from 65 and
70 °C. These Bray-Curtis similarity values were the basis for the NMDS plot
(Figure 3.11.) also supporting the strong change of the bacterial community between
65 and 70 °C in all three reactor systems (Figure 3.11).
Direct response of the bacterial community in the LBR to the temperature increase to
70 °C
Samples of the two-phase biogas reactor were normally taken at the earliest after one
fermentation period of 21 days to enable the adaption of the bacterial community to the
increased temperature. Here, to track the direct response of the bacterial community to
the temperature increase to 70 °C, the fermentation was directly analyzed after this
temperature increase.
L 21 L 0 L 7 L 14 L 21
Relative abundance of TRFs (%)
0
20
40
60
80
100
TRF 76, Thermoanaerobacter (EU262599, 87%)
TRF 84, unknown phylogenetic affiliation
TRF 139, Bacillus (AJ002154, 89%)
TRF 141, Dethiobacter (NR_044205, 88%)
TRF 151, Clostridium (FJ808611, 87-89%)
TRF 167, Clostridium (FJ808609, 96-99%)
TRF 170, unknown phylogenetic affiliation
TRF 214, Acetomicrobium (FR749980, 94-98%)
TRF 221, Acetomicrobium (FR749980, 95-98%)
TRF 291, Clostridiales (HQ452852, 85%)
TRF 374, unknown phylogenetic affiliation
65°C 70°C
Figure 3.12 Direct response of the bacterial community in the LBR to the temperature
increase of 70 °C. TRFLP profiles of leachate samples are shown.
Sequence accession number and identity of the TRF affiliations are
indicated in brackets (cf. Table 3.6). TRFs above 3% of the total
fluorescence intensity are shown. L 0 - 21, leachate at different time
points of fermentation
A clear change in the bacterial community in the LBR was already observed after
7 days of fermentation at an LBR temperature of 70 °C (Figure 3.12). In detail, the
TRFs 214 and 221 (Acetomicrobium), which were dominantly detected at 70 °C, were
already prevalent after 7 days of fermentation at 70 °C (Figure 3.12). The relative
73
abundance of these TRFs increased slightly until the end of the fermentation on
day 21. The emergence of these TRFs suppressed the detection of TRFs 76, 151 and
167 (all assigned to Clostridia), which had dominated at the start of the fermentation
directly after temperature increase to 70 °C (Figure 3.12).
3.3.7 Impactoftemperatureincreaseonthebacterialcommunityin
theAF
Besides the obvious role of the bacterial community in biomass degradation in the
LBR, the bacterial community may also play a major role in the AF interacting with
hydrogenotrophic archaea in syntrophy. To evaluate the impact of the temperature
increase on the bacterial community in the AF, this community was also analyzed
(Figure 3.13).
Temperature of LBR (°C)
55 60 65 70 75
Relative abundance of TRFs (%)
0
20
40
60
80
100
TRF 76, Thermoanaerobacter (EU262599, 87%)
TRF 95, Bacteroidales (GU129116, 87-96%)
TRF 141, Dethiobacter (NR_044205, 88%)
TRF 151, Clostridium (FJ808611, 87-89%)
TRF 161, Clostridium (GU968170, 85%)
TRF 217, Clostridium (CP000568, 91-94%)
TRF 374, unkown phylogenetic affiliation
Figure 3.13 Bacterial community attached to the AF packings during temperature
increases in the LBR from 55 to 75 °C. Sequence accession number and
identity of the TRF affiliations are indicated in brackets (cf. Table 3.6).
TRFs above 3% of the total fluorescence intensity are shown.
Altogether, only slight variations were detectable in the bacterial community in the AF
during the temperature increase in the LBR. This is supported by the Bray-Curtis
74
similarity values for these TRFLP profiles, ranging between 0.60 ± 0.09 and
0.77 ± 0.05.
During the whole experiment, the TRF 151 (Clostridium), followed by TRF 161
(Clostridium), TRF 141 (Dethiobacter), TRF 374 (unknown phylogenetic affiliation) and
TRF 95 (Bacteroidia) were the most prevalent bacterial TRFs in the AF (Figure 3.13).
Most of these TRFs were also detected as prevailing in the LBR except for TRF 161,
which was detected for the first time. The TRFs 151, 161 and 141 were assigned to
members of the Clostridia class, indicating a prevalence of Clostridia in the AF. Only
TRF 95 showed the highest sequence similarity to a member of the Bacteroidia class.
3.3.8 ChangesinthearchaealcommunityintheAFduringthe
fermentationofoneloadofsubstrate
In contrast to the dynamics of the bacterial community within the 21-day fermentation
of one load of substrate, the archaeal community in the AF altered less as indicated for
the LBR temperature regime of 55°C (Figure 3.14 A). The pairwise comparison
(Bray-Curtis index) of the TRFLP profiles from AF samples at different time points of
the fermentation showed rather high similarities between 0.70 ± 0.06 and 0.83 ± 0.01.
At all time points, the TRF 339, showing the highest sequence similarity to the
hydrogenotrophic Methanothermobacter genus, was prevalently detected
(Figure 3.14 A). Beside this strain, another member of the strict hydrogenotrophic
Methanothermobacter genus (TRF 337) was identified (Figure 3.14 A). In contrast,
strict acetoclastic (Methanosaeta sp., TRF 107) or mixotrophic methanogens
(Methanosarcina sp., TRF 628) formed the minority within the TRFLP profiles of
leachate samples derived from the AF at an LBR temperature of 55 °C (Figure 3.14 A).
75
AF samples during a
21-day fermentation (LBR 55°C)
L 0 L 2 L 7 L 14 L 21
Relative abundance of TRFs (%)
0
20
40
60
80
100
AF samples at different
temperature regimes of the LBR (°C)
L 55 P 55 P 60 P 65 P 70 P 75 L 75
TRF 107, Methanobacterium (AY552778, 97-98%)
Methanoculleus (NR_042786, 97-98%)
Methanosaeta (CP002565, 99-100%)
TRF 337, Methanothermobacter (NR_040964, 100%)
TRF 339, Methanothermobacter (HQ283273, 98-99%)
TRF 430, Methanoculleus (AB065298, 100%;
NR_028152, 98%; NR_042786, 100%)
TRF 628, Methanosarcina (AB300208, 99%)
BA C
Digestate samples at
different temperature regimes
of the LBR (°C)
D 55 D 65 D 70
Figure 3.14 Changes in the archaeal community composition during the experimental
runs. Archaeal TRFLP profiles were obtained for AF samples during the
21-day fermentation of one load of substrate at an LBR temperature of
55 °C (A), for AF samples (day 21) during temperature increases in the
LBR (55 - 75 °C, B) and for digestate samples at different temperature
regimes of the LBR (C). Sequence accession number and identity of the
TRF affiliations are indicated in brackets (cf. Table 3.6). TRFs above 3%
of the total fluorescence intensity are shown. L 0 - 21, leachate at different
time points of fermentation; L 55 & 75, leachate at LBR temperatures of
55 and 75 °C; P 55 - 75, packing at different temperature regimes of the
LBR; D 21, digestate at the end of fermentation; D 55, 65 & 70, digestate
at different temperature regimes of the LBR
3.3.9 Impactoftemperatureincreaseonthearchaealcommunityin
theAF
During the increase in the LBR operation temperature, the temperature of the AF
remained constant at 55 °C. To monitor the archaeal community during temperature
increase in the LBR, the biofilm on the packing’s surface derived from the AF was
76
analyzed at the end of each LBR temperature regime. Further, leachate samples of the
AF at an LBR temperature of 55 and 75 °C were also analyzed for comparative
analysis.
The archaeal TRF 339, assigned to Methanothermobacter (Methanobacteriales), and
TRF 628, assigned to Methanosarcina (Methanosarcinales), were most prevalent
during the whole experiment, showing an uneven progress (Figure 3.14 B). A pairwise
comparison between packing samples derived from 55 to 75 °C based on the
Bray-Curtis similarity revealed values equal or higher than 0.66 ± 0.03, indicating slight
changes in the archaeal community.
At 55 °C, the archaeal community of the packing’s biofilm consisted mostly of members
of the strict hydrogenotrophic Methanobacteriales, whereas species belonging to the
mixotrophic Methanosarcinales dominated the archaeal community at 75 °C
(Figure 3.14 B). This is also true for the leachate samples analyzed at LBR
temperatures of 55 and 75 °C.
In addition to these TRFs, further fragments were identified in the archaeal TRFLP
profiles. For instance, TRF 107, which was also prevalent, but only in the AF packing at
an LBR temperature of 55 °C (Figure 3.14 B), was assigned to members of the
Methanobacteriales, Methanomicrobiales and Methanosarcinales (Figure 3.14 B).
Additionally, TRFs 337 and 430, assigned to Methanobacteriales and
Methanomicrobiales, respectively, were less frequently detected or not detected in all
samples (Figure 3.14 B).
3.3.10Impactoftemperatureincreaseonthearchaealcommunityin
theLBR
The archaeal community in the LBR was also monitored during temperature increase
(Figure 3.14 C). As shown for the bacterial community, the temperature increase had
also an impact on the archaeal community in the LBR. Here, a shift from mixotrophic to
hydrogenotrophic methanogens was detected during temperature increase.
77
In detail, at an LBR temperature of 55 °C, the archaeal community profile was
dominated by Methanosarcina (TRF 628). Strict hydrogenotrophic methanogens, such
as Methanothermobacter (TRFs 337 and 339), were detected at lower level. At LBR
temperatures of 65 and 70 °C, the TRFLP profiles revealed a strong reduction of
Methanosarcina. Instead, TRFs assigned to the Methanothermobacter genus
(TRF 337) were identified at higher abundances. Interestingly, this TRF was detected
only rarely in the TRFLP profiles from the AF.
3.4 454pyrosequencingofmicrobialmetagenomes
Four different biogas reactor samples were analyzed by 454-pyrosequencing in order
to determine the microbial community and the genetic potential for e.g., carbohydrate
degradation. Digestate samples were taken from the LBR at 55, 65 and 70 °C (D 55,
D 60, D 70). The fourth sample was derived from the packing’s biofilm of the AF (P 55)
at an LBR temperature of 55 °C.
Table 3.8 454-pyrosequencing parameters of four different samples derived from
the two-phase biogas system
Packing of AF
LBR/AF 55 °C
(P 55)
Digestate of LBR
LBR 55 °C
(D 55)
Digestate of LBR
LBR 65 °C
(D 65)
Digestate of LBR
LBR 70 °C
(D 70)
Metagenomic sequences 248,775 303,493 309,589 315,387
Sequenced bases (bp) 97,884,221 120,496,674 124,456,333 127,974,741
Average read length (bp) 393 397 402 406
GC content (%) 8.0 - 80.9 0.0 - 80.0 11.6 - 84.2 12.8 - 81.4
bp, base pairs
The 454-pyrosequencing run resulted in 1,177,244 metagenomic sequences with an
average read length of 400 bases (Table 3.8). The sequence information obtained
comprised 470,811,969 bases for all samples. The metagenomic sequences were
phylogenetically and functionally characterized, using the MetaSams platform 0.99
(Zakrzewski et al., 2013).
78
Phylogenetic assignment of metagenomic sequences by RDP classifier and CARMA
The phylogenetic assignment of metagenomic sequences was conducted using the
RDP classifier and the CARMA tool. A total of 3,473 metagenomic sequences were
identified encoding parts of the 16S rRNA (Table 3.9), which represented only
0.29 ± 0.09% of the total metagenomic sequences. In contrast to the analysis of the rrs
sequences, the analysis of metagenomic sequences representing functional genes
(so-called environmental gene tags, EGTs) encompassed 303,514 EGTs, applying
CARMA. This represented 25.67 ± 1.83% of the total metagenomic sequences
(Table 3.9).
Table 3.9 Number of metagenomic sequences and the number of sequences
assigned to taxonomic groups applying the RDP classifier and CARMA.
For description of samples refer to Table 3.8.
P 55 D 55 D 65 D 70 Distribution (%)
Metagenomic sequences 248,775 303,493 309,589 315,387 100
Identified rrs sequences using RDP 495 771 946 1,261 0.29 ± 0.09
Identified EGTs using CARMA 59,492 75,131 80,124 88,767 25.67 ± 1.83
Identified pEGTs using CARMA 56,532 72,048 76,537 85,579 24.58 ± 1.89
EGT, environmental gene tag; pEGT, prokaryotic environmental gene tag
Altogether, approximately 70% of metagenomic sequences remained unassigned,
which indicated that a certain number of additional species are involved in the biogas
fermentation process. However, the majority of assignable metagenomic sequences
(rrs sequences and EGTs) belonged to the Bacteria domain independent of the
phylogenetic analysis or the metagenomic sample (Table 3.10).
A total ranging from 88.9% (P 55) to nearly 100% (D 70) of rrs sequences (RDP
classifier), encoding parts of the 16S rRNA, were assigned to Bacteria. The CARMA
results showed that 69.8% (P 55) to 95.0% (D 70) of EGTs, encoding functional genes,
were assigned to Bacteria. Independent of the applied phylogenetic assignment
method, the sample derived from the packing of the AF showed a reduced number of
Bacteria, whereas the number of metagenomic sequences assigned to Archaea was
clearly enriched with 11% (RDP) and 25% (CARMA). Simultaneously, the digestate
samples showed slightly increasing assignments to Bacteria and a slight decrease in
Archaea during LBR temperature increase.
79
A minor percentage of EGTs was assigned to Eukaryota using the CARMA tool
(Table 3.10). At a reactor temperature of 55 °C, EGTs belonging to Eukaryota were
identified with an average percentage of 3.9% (P 55, D 55). At higher LBR
temperatures, Eukaryota were detected with 1.4 to 2.0% of EGTs.
Table 3.10 Distribution of phylogenetic assignments of rrs sequences and EGTs to
taxonomic groups applying the RDP classifier and CARMA. For
description of samples refer to Table 3.8.
Distribution of phylogenetic assignments (%)
Domain P 55 D 55 D 65 D 70
RDP CARMA RDP CARMA RDP CARMA RDP CARMA
Bacteria 88.9 69.8 97.9 91.6 99.0 93.6 99.9 95.0
Archaea 11.1 25.2 2.1 4.3 1.0 3.8 0.1 3.3
Eukaryota - 4.1 - 3.6 - 2.0 - 1.4
Others - 0.9 - 0.5 - 0.6 - 0.3
3.4.1 PhylogeneticassignmentofmetagenomicsequencestoBacteria
Phylogenetic assignment of metagenomic sequences to Bacteria by RDP classifier and
CARMA
The number of identified bacterial classes differed depending on the applied
phylogenetic assignment methods. Whereas the CARMA tool revealed approximately
25 bacterial classes within all samples, the RDP classifier showed a descending
diversity of Bacteria from 11 (P 55) to 3 (D 70) classes during LBR temperature
increase.
The majority of metagenomic sequences were assigned to Bacteria and herein most
prevalently to the Clostridia class, followed by the Bacilli, Thermotogae and
Gammaproteobacteria classes independent of the metagenomic sample or the applied
phylogenetic assignment method (Figure 3.15). During the LBR temperature increase,
a slight reduction of members belonging to the Clostridia was only detected for the
results derived from the RDP classifier. Simultaneously, the number of metagenomic
sequences assigned to the Bacilli class increased, indicating alterations induced by the
temperature increase.
80
More specifically, Clostridia, the prevalent class, was detected in 16.2% of rrs
sequences by the RDP classifier and 19.4% of prokaryotic EGTs (pEGTs) by the
CARMA tool in the methanogenic biofilm sample (P 55; Figure 3.15). In contrast, the
digestate samples revealed higher percentages of Clostridia between 24.4 to 31.9% of
rrs sequences by RDP and 36.2 to 44.7% of pEGTs by CARMA.
Bacilli were the next prevalent class. The digestate samples revealed an increasing
occurrence from 0.1 to 4.0% of rrs sequences (RDP) and 6.8 to 9.5% of pEGTs
(CARMA) during the LBR temperature increase from 55 to 70 °C (Figure 3.15). In the
sample derived from the AF, Bacilli were less detected with 0.4% (RDP) and 4.5%
(CARMA).
Members of the Thermotogae class were most frequently identified at an LBR
temperature of 55 °C (P 55, D55) accounting for 1.2 to 1.8% by both, the RDP and
CARMA analysis (Figure 3.15). At higher temperatures, members of the Thermotogae
were less abundant or not detected.
D 70
2 8 14 20 36
RDP
P 55
rrs sequences (%, RDP), pEGTs (%, CARMA)
2 8 14 20 36
Class
Thermotogae
Mollicutes
Spirochaetes
Gammaproteobacteria
Epsilonproteobacteria
Deltaproteobacteria
Betaproteobacteria
Alphaproteobacteria
Planctomycetacia
Fusobacteria
Clostridia
Bacilli
Deinococci
Dehalococcoidetes
Chloroflexi
Anaerolineae
Chlorobia
Flavobacteria
Cytophagia
Bacteroidia
Actinobacteria
D 55
2 8 14 20 36
D 65
2 8 14 20 36
CARMA
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
Figure 3.15 Phylogenetic assignment of metagenomic sequences to Bacteria using
the RDP classifier (grey) and CARMA (black) showing the prevalent
bacterial classes. For description of samples refer to Table 3.8. pEGTs,
prokaryotic environmental gene tags; *, abundance ≤ 0.2% of rrs
sequences or pEGTs
81
Furthermore, the CARMA analysis revealed a presence of Gammaproteobacteria from
2.5 to 3.7% of pEGTs in all samples. In contrast, the RDP classifier found members of
the Proteobacteria, such as Betaproteobacteria, to a minor extent and only at 55 °C
(Figure 3.15). In addition, members of other classes, such as Actinobacteria,
Bacteroidia, Flavobacteria and Cytophaga occurred with relative abundances of 2.2%,
1.3%, 1.3% and 0.3%, respectively (Figure 3.15).
According to the assignment of metagenomic sequences to classes, Clostridium and
Bacillus were the most prevalent bacterial genera in all metagenomic samples as
revealed by CARMA (Figure 3.16 A). In the digestate samples, a slight reduction of
members of the Clostridium genus occurred (20.4 - 17.3%) along with an increase in
members of the Bacillus genus (2.5 - 4.7%), particularly after a temperature increase to
70 °C. In addition, also Caldicellulosiruptor (0.3 - 1.7%) and Alkaliphilus (1.1 - 1.6%)
were detected with increasing proportions in the digestate samples during the LBR
temperature increase.
In the sample derived from the packing’s biofilm from the AF, 4.1% of pEGTs were
assigned to the Clostridium genus, followed by Pelotomaculum (1.1%), Petrotoga
(0.9%) and Syntrophomonas (0.8%; Figure 3.16 A).
P 55
2816
(A) CARMA
Thermotoga
Thermosinus
Thermoanaerobacter
Syntrophomonas
Synechococcus
Symbiobacterium
Streptococcus
Ruminococcus
Petrotoga
Pelotomaculum
Moorella
Halothermothrix
Desulfotomaculum
Desulfitobacterium
Clostridium
Carboxydothermus
Caldicellulosiruptor
Caldanaerobacter
Bacteroides
Bacillus
Alkaliphilus D 55
pEGTs (%)
2816
D 65
2816
D 70
2816
P 55
210
(B) RDP
Ureibacillus
Thermacetogenium
Syntrophomonas
Symbiobacterium
Sporanaerobacter
Ruminococcaceae
Petrotoga
Pelotomaculum
Parabacteroides
Halocella
Ethanoligenens
Corynebacterium
Comamonas
Bacillus
Anaerobaculum
Aminobacterium
Acetivibrio D 55
rrs sequences (%)
210
D 65
210
D 70
210
Figure 3.16 Phylogenetic assignment of metagenomic sequences to Bacteria using
CARMA (A) and the RDP classifier (B) showing the prevalent bacterial
genera. The rrs sequences assigned to Ruminococcaceae could not be
classified at genus level. For description of samples refer to Table 3.8.
pEGTs, prokaryotic environmental gene tags
82
In contrast to that, the RDP classifier revealed slightly different results (Figure 3.16 B).
In the sample derived from the AF, Anaerobaculum (2.8%), Thermacetogenium (2.4%)
and Petrotoga (1.8%) were detected prevalently. In the digestate samples,
Ruminococcaceae, which could not be assigned to a genus, were classified as
dominant with 6.4% (D 55), 11.7% (D 65) and 1.7% (D 70) of rrs sequences, indicating
a strong decrease at an LBR temperature of 70 °C. Notably, members of the
Clostridium genus were not identified in all samples by the RDP classifier, which may
be due to different applied taxonomies. Further genera such as Bacillus, Petrotoga and
Thermacetogenium were less frequently detected (≤ 1% of rrs sequences) in the
digestate samples by the RDP classifier (Figure 3.16 B).
Phylogenetic assignment of EGTs having a predicted function in carbohydrate
degradation
The Pfam database provides a broad range of protein families. Using the Pfam
analysis, carbohydrate degrading enzymes were identified in all samples and
subsequently assigned to taxonomic groups. In total 27 protein families representing
glycoside hydrolases (GH), e.g., ‘GH family 5’ (e.g., cellulase, PF00150) and
‘GH family 10’ (e.g., xylanase, PF00331) were selected for this analysis (Appendix
Table 7.1 A).
The assignment of EGTs having a predicted function in carbohydrate degradation to
taxonomic groups (Figure 3.17) also revealed strong differences between the sample
from the AF and the digestate samples from the LBR. This supports the findings of the
previous phylogenetic assignment based on rrs sequences (RDP classifier) and all
identified EGTs (CARMA). Furthermore, an alteration of the bacterial community during
temperature increase from 65 to 70 °C was detectable. The results of the digestate
samples revealed Clostridia, Bacilli, Gammaproteobacteria and Bacteroidia as
prevalent bacterial classes but with varying abundances (Figure 3.17). Whereas the
number of EGTs assigned to Clostridia remained relatively constant within the three
digestate samples, Gammaproteobacteria were reduced by 40% after the temperature
increase to 70 °C. On the other hand at 70 °C, the assignment of EGTs to Bacilli and
Chloroflexi was increased twofold, whereas Bacteroidia increased threefold, supporting
the strong emergence of Bacteroidia as revealed by the TRFLP analysis.
83
P 55
Number of EGTs
100 250 400 700
Genus Class
Thermotoga
Thermoanaerobacter
Ruminococcus
Clostridium
Caldicellulosiruptor
Bacteroides
Bacillus
Thermotogae
Gammaproteobacteria
Flavobacteriia
Deinococci
Clostridia
Chloroflexi
Bacteroidia
Bacilli
Actinobacteria D 55
100 250 400 700
D 65
100 250 400 700
D 70
100 250 400 700
*
*
*
*
*
*
*
Bacillus
Bacteroides
Caldicellulosiruptor
Clostridium
Ruminococcus
Thermoanaerobacter
Thermotoga
Figure 3.17 Phylogenetic assignment of EGTs encoding carbohydrate degrading
enzymes as revealed by CARMA. Only prevalent taxonomic groups of
class and genus level are shown. EGTs of D 55, D 65 and D 70 are
normalized to an equal amount of total metagenomic sequences. For
description of samples refer to Table 3.8. EGTs, environmental gene tags;
*, detected with ≤ 3 EGTs
At genus level, Clostridium was also prevalently detected in the digestate samples at
LBR temperatures of 55 to 70 °C (Figure 3.17). In addition, the Bacillus,
Caldicellulosiruptor and Bacteroides genera also showed carbohydrate degrading
potential. At 70 °C, Clostridium was slightly reduced, whereas Caldicellulosiruptor,
Bacteroides and Bacillus strongly increased by a factor of up to 5 (Figure 3.17).
Furthermore, Thermoanaerobacter and Ruminococcus genera were also identified as
showing carbohydrate degrading potential.
In contrast to the digestate samples, the number of EGTs with a predicted function in
carbohydrate degradation was strongly decreased in the methanogenic biofilm sample
of the AF (Figure 3.17). This small number of EGTs was assigned to the Clostridia,
Gammaproteobacteria, Bacteroidia and Actinobacteria classes. At the genus level,
Bacteroides, followed by Clostridium and Thermotoga were identified.
84
3.4.2 Geneticpotentialforthedegradationofplantderivedbiomass
Plant-derived biomass is composed of cellulose, hemicellulose and lignin. The
degradation of lignin, the most recalcitrant component, is realized principally by white-
rot fungi under aerobic conditions. Only few bacterial strains were identified showing a
lignin degrading potential (Bugg et al., 2011). To determine the genetic potential of the
microbial community for lignin degradation, Pfam protein families encompassing
enzymes with lignin degrading activity were analyzed (Table 3.11).
Laccases and peroxidases are essential ligninolytic enzymes. Laccase enzymes,
catalyzing the oxidation of aromatic and non-aromatic compounds (Claus, 2004),
belong to the multicopper oxidases protein family (PF00394, PF07731, PF07732 and
PF02578). Only the protein family PF02578 was identified in the metagenomic samples
with increasing tendency during temperature increase in the LBR (Table 3.11). A
taxonomic assignment of EGTs encoding the Pfam PF02578 resulted in bacterial
groups, such as Alkaliphilus, Syntrophobacter and Clostridium.
Furthermore, peroxidases, heme-containing enzymes, are also essential for lignin
degradation (Reddy & D`Souza, 1994). The corresponding protein family PF00141 was
not detected in the metagenomic samples (Table 3.11). Further, the protein family
glyoxal oxidase (PF07250), essential for the extracellular lignin degradation (Whittaker
et al., 1999), was identified in marginal ranges only at higher LBR temperatures of 65
and 70 °C (Table 3.11), indicating that lignin degrading enzymes were less abundant in
this biogas system.
Table 3.11 Pfam analysis of lignin degrading enzymes applying CARMA. For
description of samples refer to Table 3.8.
Pfam
accession Protein family P 55
(EGTs)
D 55a
(EGTs)
D 65a
(EGTs)
D 70a
(EGTs)
PF00394 Multicopper oxidase 0 0 0 0
PF07731 Multicopper oxidase 0 0 0 0
PF07732 Multicopper oxidase 0 0 0 0
PF02578 Multicopper polyphenol oxidoreductase laccase 39.00 40.99 53.04 66.26
PF00141 Peroxidase 0 0 0 0
PF07250 Glyoxal oxidase N-terminus 0 0 2.41 1.58
a, identified EGTs are normalized to an equal amount of total metagenomic sequences
85
In contrast, EGTs encoding protein families, which encompass enzymes with
degradation activity for other plant-derived polysaccharides (e.g., cellulose) and for
oligo- and disaccharides, were identified in higher amounts in the metagenomic
samples (Table 3.12). The analysis of the genetic potential for these carbohydrate
degrading enzymes suggests a defined carbohydrate degrading potential for each
metagenomic sample, representing different microbial communities.
Table 3.12 Most prevalent GH families of the metagenomic samples as derived from
the Pfam analysis by CARMA. For description of samples refer to Table
3.8.
Samples Pfam
accession
GH
family Selection of GH activitiesa EGTsb
P 55
PF03065 GH 57 Α-amylase, 4-α-glucanotransferase, α-galactosidase 79
PF02056 GH 4 Α-glucosidase, α-galactosidase, 6-phospho-β-glucosidase 72
PF04616 GH 43 Β-xylosidase, arabinanase, xylanase 55
PF02837 GH 2 Β-galactosidase, β-mannosidase, β-glucuronidase 51
PF01915 GH 3 Β-glucosidase, xylan 1,4-β-xylosidase 43
PF03636 GH 65 Α,α-trehalase, maltose phosphorylase 34
D 55
PF00759 GH 9 Endoglucanase, cellobiohydrolase, β-glucosidase. 221
PF00331 GH 10 Endo-1,4-β-xylanase, endo-1,3-β-xylanase 199
PF01915 GH 3 Β-glucosidase, xylan 1,4-β-xylosidase 187
PF04616 GH 43 Β-xylosidase, arabinanase, xylanase 161
PF00150 GH 5 Cellulase, glucan β-1,3-glucosidase, endo-β-1,4-xylanase,
cellulose β-1,4-cellobiosidase 116
PF02056 GH 4 Α-glucosidase, α-galactosidase, 6-phospho-b-glucosidase 96
D 65
PF00331 GH 10 Endo-1,4-β-xylanase, endo-1,3-β-xylanase 221
PF01915 GH 3 Β-glucosidase, xylan 1,4-β-xylosidase 191
PF04616 GH 43 Β-xylosidase, arabinanase, xylanase 184
PF01055 GH 31 Α-glucosidase, α-xylosidase 166
PF00150 GH 5 Cellulase, glucan β-1,3-glucosidase, endo-β-1,4-xylanase,
cellulose β-1,4-cellobiosidase 129
PF00759 GH 9 Endoglucanase, cellobiohydrolase, β-glucosidase 112
D 70
PF01055 GH 31 Α-glucosidase, α-xylosidase 438
PF04616 GH 43 Β-xylosidase, arabinanase, xylanase 360
PF01915 GH 3 Β-glucosidase, xylan 1,4-β-xylosidase 335
PF00331 GH 10 Endo-1,4-β-xylanase, endo-1,3-β-xylanase 328
PF01229 GH 39 Β-xylosidase 219
PF02837 GH 2 Β-galactosidase, β-mannosidase, β-glucuronidase 177
a, activities of GH families were retrieved from the CAZY (carbohydrate-active enzymes) database
(Cantarel et al., 2009)
b
, EGTs are normalized to an equal amount of total metagenomic sequences
GH, glycoside hydrolase
86
In the methanogenic sample P 55, the prevalent protein families, such as the GH
families 4 (PF03065), 43 (PF04616) and 3 (PF01915), comprising e.g.,
α-galactosidase, xylanase and β-glucosidase, showed a lower abundance in
comparison to GH families of the digestate samples derived from the LBR (Table 3.12).
Nevertheless, a genetic potential for enzymes having polysaccharide degrading activity
(e.g., cellulose or xylan degradation) was also detected in the microbial community in
the AF.
In contrast, the metagenomic digestate samples revealed GH families with higher
abundances (Table 3.12). GH families, such as 10 (PF00331), 3 (PF01915), 43
(PF04616) and 5 (PF00150) were identified prevalently in the digestate samples
derived from the LBR at 55 and 65 °C. These GH families encompass enzymes, such
as β-xylanase, β-glucosidase, β-xylosidase and cellulase, which are mainly involved in
the degradation and processing of polysaccharides, such as xylan and cellulose. At
70 °C, the GH families 31 (PF01055), 43 (PF04616), 3 (PF01915), 10 (PF00331) and
39 (PF01229) were most prevalent with even increased abundances. These most
prevalent GH families mainly comprise enzymes, such as α- and β-xylosidase,
β-glucosidase and xylanase, which are also relevant for the breakdown of
polysaccharides, such as xylan and cellulose.
Furthermore, to assess differences in the genetic potential of the microbial
communities in the different reactor compartments and at different LBR temperatures,
a comparison of the abundances of specific GH families was performed (Figure 3.18).
To do this, GH protein families obtained through Pfam analysis, representing enzymes
involved in plant-derived biomass degradation were compared.
The comparison of the most prevalent GH families revealed the strongest difference for
the sample of the AF in comparison to the digestate samples of the LBR (Figure 3.18).
Almost all of the GH families analyzed were decreased in the metagenomic sample
from the AF. In the digestate samples of the LBR, the strongest changes were
observed in the LBR sample after the temperature increase to 70 °C, which is in
accordance with the previous TRFLP and phylogenetic assignment results.
87
B
Changing factor to sample D 55
-70 -60 -50 -40 -30 -20 -10 0 10 20 130
Pfam accession number (GH family)
PF07748 (GH 38)
PF02056 (GH 4)
PF07470 (GH 88)
PF00331 (GH 10)
PF01915 (GH 3)
PF07488 (GH 67)
PF04616 (GH 43)
PF00703 (GH 2)
PF02836 (GH 2)
PF02837 (GH 2)
PF07477 (GH 67)
PF00723 (GH 15)
PF02449 (GH 42)
PF08532 (GH 42)
PF01055 (GH 31)
PF02156 (GH 26)
PF01229 (GH 39)
PF00295 (GH 28)
P 55
D 65
D 70
A
PF00150 (GH 5)
PF02011 (GH 48)
PF03636 (GH 65)
PF02055 (GH 30)
PF03065 (GH 57)
PF03632 (GH 65)
PF01074 (GH 38)
PF00722 (GH 16)
PF00759 (GH 9)
PF00728 (GH 20)
PF00457 (GH 11)
PF02446 (GH 77)
PF01270 (GH 8)
Figure 3.18 Differences in GH family abundances between the metagenomic samples
P 55, D 65 and D 70 and the digestate sample derived from the LBR at
55 °C (D 55). Changing factor shows the decreased GH family
abundances in the digestate samples (A) and the increased GH family
abundances in digestate samples (B) during temperature increase. The
most prevalent GH families (number of EGTs identified >50) are shown as
derived from the Pfam analysis by CARMA. For description of samples
refer to Table 3.8. GH, glycoside hydrolase
88
In detail, the digestate samples derived from the LBR at 70 °C, but also at 65 °C,
revealed decreasing abundances of some GH families with polysaccharide degradation
activity in comparison to the digestate sample taken at an LBR temperature of 55°C
(Figure 3.18 A). For instance, the GH protein families 8 (PF01270) and 11 (PF00457),
involved in the degradation of xylan, were strongly decreased particularly at LBR
temperatures of 70 °C. Further GH families, representing enzymes with poly-, di- or
monosaccharide degradation and processing activity, were also less frequently
detected at higher LBR temperatures. For instance, the abundance of the protein
families GH 9 (PF00759) and 38 (PF01074) was strongly reduced.
The possibility of assigning the EGTs, which encode these protein families, to
taxonomic groups gives insights into the changes in the microbial community.
Interestingly, the assignment of EGTs, encoding the GH families 8, 9, 11 and 38
mentioned above, revealed a decrease in members of the Firmicutes and particularly a
decrease of members belonging to the Clostridia class.
In contrast to these findings, the digestate samples derived from the LBR at 65 °C and
particularly from the LBR at 70 °C also revealed higher abundances of specific GH
families in comparison to the digestate sample at 55°C (Figure 3.18 B). Protein
families, such as the GH families 28 (PF00295), 39 (PF01229), 26 (PF02156) and 31
(PF01055) were particularly more abundant in the sample derived from the LBR at
70 °C. Some of these protein families encompass pectinolytic enzymes (GH 28),
enzymes, which are involved in the degradation and the processing of xylan (GH 26,
31, 39) as well as further enzymes for the breakdown of polysaccharides (GH 31).
The taxonomic assignment of EGTs, encoding GH 28, revealed that the strong
increase in the GH family 28 at an LBR temperature of 70 °C was mainly caused by the
emergence of members of the Bacteroidetes phylum. In contrast to that, the
assignment of EGTs, encoding the GH family 39, revealed an emergence of
Proteobacteria (mainly Alphaproteobacteria) and Firmicutes (mainly Clostridia).
Further, an increase of Firmicutes (Clostridia), Bacteroidetes, Proteobacteria and
Chloroflexi was also detected for the GH families 26 and 31 during temperature
increase through the phylogenetic assignment of the corresponding EGTs.
89
3.4.3 Phylogeneticassignmentofmetagenomicsequencestoselected
pathogens
Phylogenetic assignment of metagenomic sequences to selected pathogens by
CARMA
Pathogens residing in agricultural biogas reactors may present risks for plants, animals
and humans due to the usage of digestate as fertilizer. Selected plant pathogens, such
as Agrobacterium tumefaciens, Clavibacter michiganensis, Ps. syringae, Ralstonia
solanacearum and Xanthomonas campestris, were identified in marginal amounts with
less than 0.02% of pEGTs each (Figure 3.19). Furthermore, pathogens responsible for
infections in potatoes, such as Synchytrium endobioticum, Rhizoctonia solani,
Helminthosporium solani but also other plant pathogens, such as Fusarium oxysporum
and Sclerotinia sclerotiorum were not identified in the metagenomic samples.
P 55
0.1 0.3 0.6
Pathogens
Salmonella enterica
Listeria monocytogenes
Escherichia coli
Clostridium tetani
Clostridium perfringens
Clostridium difficile
Clostridium botulinum
Campylobacter jejuni
Xanthomonas campestris
Ralstonia solanacearum
Pseudomonas syringae
Clavibacter michiganensis
Agrobacterium tumefaciens D 55
0.1 0.3 0.6
D 65
pEGTs (%)
0.1 0.3 0.6
D 70
0.1 0.3 0.6
1
2
*
*
*
*
*
*
Figure 3.19 Phylogenetic assignment of pEGTs to selected plant (1) and human or
animal pathogens (2) as revealed by CARMA. For description of samples
refer to Table 3.8. *, pEGTs with an relative abundance <0.05‰
Animal and human pathogens belonging to the Campylobacter, Clostridium,
Escherichia, Listeria and Salmonella genera were also analyzed using CARMA (Figure
3.19). The assignment of pEGTs to these pathogens was higher with up to 0.6%
90
(Cl. difficile, Figure 3.19) in comparison to the plant pathogens. However, this still
indicated a minor abundance of putatively pathogen-associated pEGTs. The
occurrence of some species, such as E. coli, L. monocytogenes and Salm. enterica,
was reduced at an LBR temperature of 70 °C. On the other hand, the detection of
pathogenic Clostridium species or Camp. jejuni was slightly increased in the digestate
sample at an LBR temperature of 70 °C.
Genetic potential for pathogenicity as revealed by Pfam analysis
The analysis of EGTs matching Pfam protein families representing toxins or virulence
factors was performed focusing only on selected animal and human pathogens.
Altogether, only 0.02% of the 1,177,244 metagenomic sequences (Table 3.13) showed
similarities to selected protein families (Appendix Table 7.2), possessing pathogenicity
in animals and humans. This means that the majority of these pathogen-associated
protein families were represented by less than 6 EGTs.
For instance, the β-1,4-N-acetyl-galactosaminyltransferase family (PF06306) of Camp.
jejuni, which is required for the GT1a ganglioside mimic synthesis and therefore
associated with the Guillain-Barré syndrome (Gilbert et al., 2000), was only marginally
detected at a temperature of 55 °C (Table 3.13). This is also true for the heat-labile
enterotoxin of E. coli (PF01375, PF01376) and for the Clostridium neurotoxin
(PF07953), which is composed of the tetanus neurotoxin and different serotypes of the
botulinum neurotoxin (Punta et al., 2012).
Other protein families were also detected to a marginal extent at higher reactor
temperatures of 65 and 70 °C (Table 3.13). For instance, the Campylobacter major
outer membrane protein family (PF05538), which may be involved in the adaption to
host environments (Zhang et al., 2000), but also some Clostridium-associated protein
families were detected. The protein families of the Cl. botulinum HA-17 protein
(PF05588), a hemagglutinin subcomponent, which is part of the L toxin, a progenitor
toxin of the Cl. botulinum type D str. 4947 (Kouguchi et al., 2002) were also identified
marginally at 70 °C. The same is true for the ADP-ribosyltransferase exoenzyme
protein family (PF03496). This enzyme, found in clostridial species, such as
Cl. perfringens, acts on actin, leading to lethal and dermonecrotic reactions in
mammals (Tsuge et al., 2003). Furthermore, EGTs assigned to the heat-stable
91
enterotoxin ST (PF02048) of E. coli were also identified at temperatures of 70 °C
(Table 3.13).
Whereas these pathogen-associated protein families were detected with less than
6 EGTs, the Holin protein family (PF05105) was identified with slightly higher
abundances in all samples (Table 3.13). The Holin protein family includes the protein
TcdE/UtxA, which is involved in toxin secretion in Cl. difficile (Tan et al., 2001). Further,
this family also includes other proteins, which are involved in bacterial lysis and virus
dissemination (Punta et al., 2012).
Table 3.13 Assignment of EGTs matching with toxin-associated Pfam families. For
description of samples refer to Table 3.8.
Putative
pathogen
Pfam
accession Protein family P 55 D 55a D 65a D 70a
Campylobacter
jejuni PF05538 Campylobacter major outer
membrane protein 0.0 0.0 0.8 0.8
Campylobacter
jejuni PF06002 Α-2,3-sialyltransferase (CST-I) 0.0 0.0 0.8 0.0
Campylobacter
jejuni PF06306
Β-1,4-N-
acetylgalactosaminyltransferase
(CgtA)
0.0 2.5 0.0 0.0
Clostridium
botulinum PF05588 Clostridium botulinum HA-17
protein 0.0 0.0 0.8 2.4
Clostridium sp. PF07953 Clostridium neurotoxin, N-terminal
receptor binding 0.0 2.5 0.0 0.0
Clostridium sp. PF08470 Nontoxic nonhaemagglutinin C-
terminal 1.0 0.0 0.0 0.0
Clostridium sp. PF03495 Clostridial binary toxin B/anthrax
toxin PA 0.0 1.6 0.8 0.0
Clostridium
difficile PF05105 Holin family 39.0 33.6 44.2 52.8
Clostridium
perfringens PF03496 ADP-ribosyltransferase exoenzyme 0.0 3.3 0.0 3.9
Escherichia coli PF01375 Heat-labile enterotoxin α chain 1.0 0.0 0.0 0.0
Escherichia coli PF01376 Heat-labile enterotoxin β chain 0.0 5.7 0.0 0.0
Escherichia coli PF02048 Heat-stable enterotoxin ST 0.0 0.0 0.0 2.4
Escherichia coli,
Shigella flexneri PF06109 Haemolysin E (HlyE) 0.0 0.8 1.6 0.0
a, identified EGTs are normalized to an equal amount of total metagenomic sequences
92
P 55
0246
Methanosarcinales
Methanomicrobiales
Methanococcales
Methanobacteriales D 55
rrs sequences (%, RDP), pEGTs (%, CARMA)
0246
D 65
0246
D 70
0246
CARMA
RDP
3.4.4 PhylogeneticassignmentofmetagenomicsequencestoArchaea
Phylogenetic assignment of metagenomic sequences to Archaea by RDP classifier and
CARMA
In the metagenomic samples, the methanogenic archaeal orders Methanosarcinales,
Methanobacteriales, Methanomicrobiales and Methanococcales were identified (Figure
3.20).
Most assignments to these Archaea were obtained in the packing’s biofilm from the AF
at an LBR temperature of 55 °C. Here, Methanosarcinales prevailed with 7.3% of
pEGTs as revealed by CARMA and 5.3% of rrs sequences by the RDP classifier. In
addition, members of the Methanobacteriales (6.9% CARMA, 4.2% RDP) and
Methanomicrobiales (2.5% CARMA, 1.4% RDP) were also detected. Members of the
Methanococcales were less frequently detected with 0.85% of pEGTs (CARMA).
In contrast to these results, the TRFLP analysis and the rrs gene library results
revealed a prevalence of the Methanobacteriales order at an LBR temperature of
55 °C.
Figure 3.20 Phylogenetic assignment of metagenomic sequences to Archaea using
CARMA (black) and the RDP classifier (grey) showing the prevalent
archaeal orders. For description of samples refer to Table 3.8. pEGTs,
prokaryotic environmental gene tags
At genus level, Methanosarcina, Methanothermobacter and Methanobacterium were
most prevalent in the sample of the AF (Figure 3.21). Interestingly, the CARMA tool
revealed Methanosarcina and Methanothermobacter as most prevalent, whereas the
RDP classifier revealed Methanosarcina and Methanobacterium as dominant genera.
93
These differences are probably caused by the use of different taxonomies. Whereas
the RDP classifier is based on the taxonomy proposed by Garrity and coworkers
(2007), CARMA is based on the NCBI taxonomy.
However, in contrast to the AF sample, the digestate samples showed a strongly
reduced number of archaeal assignments (Figure 3.20), particularly at higher LBR
temperatures. At an LBR temperature of 55 °C, members of the Methanosarcinales
were identified with 2.1% of pEGTs (CARMA) and 1.3% of rrs sequences (RDP). At
higher LBR temperatures of 65 and 70 °C, less than 1% pEGTs or rrs sequences
(CARMA and RDP) were classified to methanogenic archaea in the digestate samples.
However, Methanobacteriales was the major archaeal order at these higher
temperatures.
P 55
26
(A) CARMA
Methanothermobacter
Methanospirillum
Methanosphaera
Methanosarcina
Methanosaeta
Methanoculleus
Methanococcus
Methanobrevibacter
Methanobacterium D 55
pEGTs (%)
26
D 65
26
D 70
26
P 55
26
(B) RDP
Methanothermobacter
Methanospirillum
Methanosphaera
Methanosarcina
Methanosaeta
Methanoculleus
Methanococcus
Methanobrevibacter
Methanobacterium D 55
rrs sequences (%)
26
D 65
26
D 70
26
Figure 3.21 Phylogenetic assignment of metagenomic sequences to Archaea using
CARMA (A) and the RDP classifier (B) showing the prevalent archaeal
genera. For description of samples refer to Table 3.8. pEGTs, prokaryotic
environmental gene tags
Phylogenetic assignment of EGTs having a predicted function in methanogenesis
Protein families relevant for methanogenesis were identified and subsequently
assigned to taxonomic groups (Figure 3.22). In total 27 protein families, e.g. ‘methyl-
coenzyme M reductase subunits’ (PF02249, PF02745, PF02241, PF02783) were
selected for this analysis (Appendix Table 7.1 B).
The assignment of methanogenic Pfam protein families to taxonomic groups differed
slightly from the previous phylogenetic assignment of metagenomic sequences by RDP
and CARMA. Here, in the biofilm sample derived from the AF, Methanobacteriales
(182 EGTs), followed by Methanosarcinales (126 EGTs), were prevalently identified
94
(Figure 3.22), as indicated by the TRFLP and rrs gene library analyses. According to
this, the genera Methanothermobacter and Methanosarcina were dominantly detected
in the packing sample derived from the AF at 55 °C.
P 55
Number of EGTs
20 60 100 180
Genus Order
Methanothermobacter
Methanospirillum
Methanosphaera
Methanosarcina
Methanoculleus
Methanococcus
Methanosarcinales
Methanomicrobiales
Methanococcales
Methanobacteriales
D 55
20 60 100 180
D 65
20 60 100 180
D 70
20 60 100 180
Methanobacteriales
Methanococcales
Methanomicrobiales
Methanosarcinales
Figure 3.22 Phylogenetic assignment of EGTs encoding methanogenic enzymes as
revealed by CARMA. Only major groups of the class and genus level are
shown. EGTs of D 55, D65 and D 70 are normalized to an equal amount
of total metagenomic sequences. For description of samples refer to
Table 3.8. EGTs, environmental gene tags
In contrast to that, the digestate samples showed similar results to the phylogenetic
assignment by RDP and CARMA (Figure 3.22). In the digestate sample at an LBR
temperature of 55 °C, Methanosarcinales as well as Methanobacteriales were
identified, but at higher temperatures hardly any methanogenic protein family was
detected.
3.5 Quantificationofbacterialrrsgenes
QPCR analysis allows the detection and quantification of specific microorganisms in
various habitats. In this study, the quantification of Bacteria and of a putatively process-
relevant bacterium was performed. This bacterium was detected as prevalent species
95
after one week of fermentation in the LBR at 55 to 60 °C as shown by TRFLP and rrs
gene library analyses (cf. Figure 3.9 or 3.10). Hence, it can be assumed that this
bacterium plays an important role in the anaerobic degradation process at thermophilic
temperatures. Its corresponding rrs sequence has the accession number HE804843
and resulted in the TRF 304 after TRFLP analysis. The NCBI BLAST analysis of this
sequence showed the highest sequence similarity to Defluviitalea saccharophila str.
LIND6LT2 (Defluviitaleaceae, Clostridiales, Jabari et al., 2012) with 89 to 91%. This
indicates a more distant taxonomical relation to the target microorganism. In contrast,
the RDP classifier assigned the target sequence HE804843 to the Lachnospiraceae
with a confidence level of 93 to 94%, which represents another family within the
Clostridiales order.
To monitor this putatively process-relevant bacterium, a specific TaqMan primer set
(primer set 304), comprising the forward and reverse primer as well as the TaqMan
probe, was developed.
3.5.1 Specificityoftheprimerset304
The TaqMan primer set 304 developed in this study was analyzed for target-strain
specificity and potential false positive or negative results. To identify potential false
negative results, the rrs gene libraries constructed were analyzed for mismatches to
the primer set 304. Only one sequence (out of 16), which resulted in a TRF 304 after
TRFLP analysis, revealed a mismatch to the 304r reverse primer. However, one
mismatch of the target sequence to the reverse primer should not seriously affect the
binding of the whole primer set 304.
To identify potential false positive results, further sequences of the rrs gene libraries
were screened for potential binding sites of the primer set 304. In the rrs gene libraries
(excluding rrs sequences resulting in TRF 304), two sequences with a potential false
positive binding site were found for the 304r primer and the TaqMan304 probe.
However, a false positive amplification is rather unlikely due to the fact that the
corresponding forward primer of the primer set 304 did not bind to these sequences. All
other sequences in the bacterial rrs gene libraries showed at least 8 mismatches to the
96
primer set 304, as confirmed after sequence alignment. Particularly the forward primer
seemed to be highly specific, whereas the TaqMan probe and particularly the reverse
primer were less specific.
To gain more detailed information on potential false positive results, the NCBI Primer
BLAST tool and the RDP probe match tool were used for comparison (last access
November 2012). The primer BLAST analysis of the primer set 304 against the NCBI nr
database (limited to Bacteria) resulted in 46 sequences matching the primer set 304
(Table 3.14). However, most of these sequences from uncultured bacteria showed a
high sequence similarity of 99 to 100% to the target sequence (HE804843), indicating a
very close relation. Furthermore, some of these highly similar sequences also stemmed
from thermophilic anaerobic digesters. Additionally, a primer BLAST search of the
primer set 304 against the archaeal NCBI nr database showed only potential binding
sites within one sequence with five mismatches each to the 304f and 304r primer.
When applying the RDP probe match tool, similar results to those described above
were obtained. When analyzing the target specificity of the forward and reverse primer
304, 24 sequences plus 15 sequences derived from this study (resulting in TRF 304)
were identified to have no mismatch to the primers. These 39 sequences clustered
within the unclassified Lachnospiraceae group, as assumed for the target sequence
HE804843 by RDP classification. Hence, these sequences showed high sequence
similarities of up to 100% to the target sequence also indicating a very close relation.
These sequences partially agree with the sequences shown in Table 3.14.
Furthermore, 48 matches of the 304f and 304r primers to RDP database sequences
were obtained when using the RDP probe match tool enabled for three mismatches.
Only five out of 293,554 sequences belonging to the Bacteroidetes group (accession
numbers: FJ930387, GU230430, HQ478279, FJ745267, GU584663) and two out of
123,596 sequences (accession numbers: EU645078, EU645198) belonging to the
unclassified Bacteria were identified as having only three mismatches to the target
sequence. However, these sequences showed additionally 2 to 4 mismatches to the
TaqMan probe, which indicated that the risk of a false positive amplification is rather
low.
Hence, the primer set 304 seemed to be specific for a small group of unclassified
Lachnospiraceae, closely related to the target sequence HE804843, which was
prevalently identified during the thermophilic fermentation process.
97
Table 3.14 Uncultured or environmental sequences retrieved from the primer BLAST
search against the NCBI nr database (last access November 2012)
showing no mismatch to the primer set 304. Retrieved sequences showed
high sequence similarities to the target sequence (TRF 304; HE804843)
of primer set 304.
Hit Accession number Microorganism Source
Sequence
similarity to
target
sequence
1 DQ887970 Uncultured bacterium
clone B55_K_B_F05
Thermophilic
anaerobic solid
waste bioreactor
99%
2 HQ183761 Uncultured Bacillus sp.
clone De247
Leachate
sediment 99%
3 FN994058 Uncultured bacterium
str. MS14339-B088
Long-term biogas
completely stirred
tank reactor
99%
4 AM947511 Uncultured bacterium
clone 2d_1FB3
Anaerobically
digested sludge 99%
5 JF795000 Uncultured bacterium
clone T4 Anoxic bulk soil 99%
6 - 21
JN708647, -717, -724, -805, -820, -847,
-871, -879, -884, -898, -921, -924, -935,
-950, -979, -997
Uncultured bacterium
clones
Thermophilic
anaerobic
digester
99 - 100%
22 - 46
JQ100075, -0621, -6376, -6551, -7804;
JQ110262, -0839, -1744, -6707;
JQ120796, -3515, -6358, -6445, -9279;
JQ130054, -6180, -8776; JQ148181;
JQ161845, -5666, -6328, -7478, -8540, -
8577, -9911
Uncultured bacterium
clones
Anaerobic sludge
digester
99 - 100%,
2x97%,
1x95%
3.5.2 QuantificationofBacteriaandaputativelyprocessrelevant
bacterium
Parameters suited to assess the efficiency of the qPCR analysis (PCR efficiency, R2
and slope of the calibration curve), were within the limits stated by Zhang and Fang
(2006), indicating reliable results. Furthermore, all results of the analyzed samples
were within the calculated limits of detection and quantification (cf. 2.6).
During temperature increase in the LBR from 55 to 75 °C, the number of bacterial rrs
copies was reduced by almost one log (Figure 3.23 A). More specifically between 55
and 65 °C, the number of the bacterial rrs copies per mL leachate was consistent with
an average number of 1.3 ± 0.5 x 1011. At 70 °C, the number of the bacterial rrs copies
was reduced to 3.6 ± 0.1 x 1010 per mL leachate. After the bioaugmentation with
98
compost, which introduced new thermophilic microorganisms, the bacterial rrs copy
number was slightly increased to 4.3 ± 0.8 x 1010 per mL leachate. This effect did not
persist at 75 °C. Here, a further reduction of the bacterial rrs copy number was
detected, accounting for 2.8 ± 0.5 x 1010 copies per mL leachate.
LBR of the two-phase biogas reactor
55°C 60°C 65°C 70°C 70°C* 75°C
rrs gene copies mL-1 (log)
1e+1
1e+4
1e+5
1e+6
1e+7
1e+8
1e+9
1e+10
1e+11
1e+12 Bacteria
TRF 304
Upflow reactor
of an UASSR
55°C
AB
Figure 3.23 Quantification of Bacteria and bacteria, resulting in TRF 304, by qPCR
analysis. Leachate samples of the LBR of the two-phase biogas system at
temperatures from 55 to 75 °C were analyzed (A). Leachate samples
derived from the upflow reactor of a thermophilic upflow anaerobic solid-
state reactor (UASSR, 55 °C) were employed for comparison (B). The
standard deviation represents the results of three measurements. 70 °C*,
analysis after bioaugmentation with compost
The analysis of bacteria, resulting in TRF 304 after TRFLP analysis and representing a
small group of the unclassified Lachnospiraceae, was performed using the primer set
304. The number of the rrs copies was reduced by 4 logs during temperature increase
in the LBR (Figure 3.23 A), indicating a sensitivity of these bacteria to temperatures
above 65 °C. This supported the findings of the TRFLP analysis, where the TRF 304
was absent in TRFLP profiles derived from samples at 70 and 75 °C.
However, up to an LBR temperature of 60 °C, these bacteria were identified
prevalently. At temperatures of 55 and 60 °C, they were detected with 2% (55 °C) and
3% (60 °C) of the total bacterial rrs copies, accounting for 2.4 ± 1.5 x 109 and
2.0 ± 0.5 x 109 rrs copies per mL leachate, respectively. At 65 °C, the number of rrs
copies was reduced by one log to 1.5 ± 0.9 x 108, representing now only 0.1% of the
99
total bacterial rrs copies. From 70 °C on, the number of rrs copies was strongly
reduced by 4 orders of magnitude to 1.1 ± 0.4 x 105 to 3.0 ± 0.2 x 105 copies per mL. It
was here that these bacteria accounted for less than 0.001% of the total bacterial rrs
copies. Furthermore, an effect of bioaugmentation on the abundance of these bacteria
was not detected.
To prove that bacteria, resulting in TRF 304, also inhabit other thermophilic biogas
reactor systems, a thermophilic upflow anaerobic solid-state reactor (UASSR, 55 °C)
was also analyzed (Figure 3.23 B). The results showed a number of 6.6 ± 0.03 x 108 rrs
copies per mL leachate for these bacteria, representing 0.1% of the total bacterial rrs
copies within the UASSR. Although the relative abundance of these bacteria, was
lower in the thermophilic UASSR, these results proved that they also inhabit other
thermophilic biogas systems.
3.6 Microscopicalanalysesofthemicrobialbiogascommunity
Microscopical analyses were commonly used to quantify the total number of cells,
Bacteria and Archaea. Therefore, DAPI staining of microbial cells and DOPE-FISH
analysis were conducted. Furthermore, fresh leachate samples were also analyzed by
PI staining identifying cells with reduced membrane integrity. Parts of this chapter were
performed in the context of the Diploma thesis of C. Krumrei (2010).
3.6.1 Fluorescenceinsituhybridizationandtotalcellcountanalyses
Total cell count analysis was carried out after sampling the LBR leachate at all LBR
temperature regimes (55 - 75 °C). This increase in operation temperature in the LBR
led to a reduction in cell densities by almost one log as determined by DAPI staining
(Figure 3.24 A), which is in accordance with the qPCR results.
100
Day of fermentation
0 4 8 12 16 20
Cell stained with PI (%)
0
20
40
60
80
100
LBR1
LBR2
LBR3
At 55 °C, an average of 2.7 ± 0.6 x 1010 cells per mL was detected. A slight reduction of
the total cell count was observed after an LBR temperature increase to 60 °C. At 60
and 65 °C, a total of 1.6 to 1.7 ± 0.2 x 1010 cells per mL was determined (Krumrei,
2010). The further increase in the LBR temperature to 70 °C led to a reduced cell count
of 6.2 ± 1.3 x 109 cells per mL. The following bioaugmentation showed no positive
effect on the cell densities. After the bioaugmentation at 70 and 75 °C, the average
total cell count was lower than before with 4.6 ± 1.1 x 109 cells per mL.
The DOPE-FISH analysis was performed for the first three temperature regimes of the
LBR (55 - 65 °C, Figure 3.24 A; Krumrei, 2010). The hybridization rate achieved for this
analysis was very low at less than 30%, although different optimization steps were
performed in the context of a Diploma thesis of Krumrei (2010). On the other hand, the
hybridization rate of the pure cultures, such as E. coli (DSM 1116), Cl. tyrobutyricum
(DSM 2637) and Methanospirillum hungatii str. Mh1, was high between 80 and 100%.
Temperature of LBR
55°C 60°C 65°C 70°C 70°C* 75°C
Cells mL-1 (log)
1e+1
1e+2
1e+6
1e+7
1e+8
1e+9
1e+10
1e+11
1e+12 Total cells
Bacteria
Archaea
NA NA NA
Figure 3.24 Total cell counts and DOPE-FISH results during temperature increase (A),
percentage of cells stained with PI during 21 days of fermentation at
65 °C (B). NA, not analyzed; 70*, analysis after compost bioaugmentation
(data based on the Diploma thesis of Krumrei, 2010)
However, no strong alterations of bacterial and archaeal cells were detected between
LBR temperatures of 55 and 65 °C. The number of bacterial cells as determined by
DOPE-FISH remained almost stable representing 20.4 to 28.2% of the total cell count.
In contrast, the archaeal cells represented only 0.3 to 0.6% of the total cell count, which
indicated that archaeal cells represented only a small part of the microbial community.
A
B
101
3.6.2 Propidiumiodideanalysis
PI is a fluorescent dye that intercalates in RNA and DNA. It can only enter in microbial
cells with reduced membrane integrity. PI staining was performed analyzing fresh
leachate samples of the three LBRs at 65 °C without a previous treatment.
During the 21-day fermentation at 65 °C, the number of microbial cells that were
stained was nearly constant and ranged between 53.7 ± 4.9% on day 0 and
63.4 ± 10.2% on day 15 (Figure 3.24 B; Krumrei, 2010), indicating a high number of
cells with reduced membrane integrity.
102
4 DISCUSSION
4.1 Methodicalaspects
As the majority of microorganisms are not yet cultivable, culture-independent methods
are of major importance for gaining insights into different microbial habitats due to the
fact that these methods overcome the strong limitations concerning cultivation
efficiency. In the following sections, the advantages of the key methods applied as well
as potential drawbacks are discussed.
4.1.1 DNAextraction
An effective DNA extraction is of major importance for culture-independent, DNA-based
molecular analyses. A well known problem is the variable lysis efficiency of microbial
cells. Different lysis procedures are known for extracting DNA from microbial cells,
such as the application of chemical compounds (e.g., SDS), enzymes (lysozyme) and
mechanical (e.g., beating step with silica or glass beads) or physical (e.g., freeze and
thaw cycles) treatment.
Four different DNA extraction methods were compared with respect to their efficiency
and reproducibility. Two DNA extraction kits (FastDNA® Spin Kit for Soil, PowerSoil®
DNA Isolation Kit), including a beating step, and a step-by-step protocol, including
lysozyme and SDS for lysis, with and without an additional beating step were applied.
103
The snapshot of the bacterial diversity obtained after using the four DNA extraction
protocols was the same, focusing on TRFs with more than 2% of the total fluorescence
intensity. Only the relative abundance of TRFs varied between the four DNA extraction
methods. Changes within the relative abundance of the microbial community after
different DNA extractions have been reported by different authors also analyzing
fingerprint profiles (e.g., Krsek & Wellington, 1999; Sun et al., 2012). However, to
secure a high lysis efficiency of microbial cells, the DNA extraction protocols with a
mechanical treatment step (FastDNA® Spin Kit for Soil and step-by-step protocol) were
used in this study, resulting in highly reproducible results. DNA extraction protocols
with a mechanical treatment step were shown to be efficient methods as indicated by
Bergmann and coworkers (2010) and Vanysacker and coworkers (2010).
A further challenge is the DNA extraction from biofilms attached to heterogenic crop
material, such as the digestate. Several studies have applied chemical (e.g. Chen &
Stewart, 2000), enzymatic (e.g. Johansen et al., 1997; Böckelmann et al., 2003) or
physical (e.g. Mott et al., 1998; Rochex et al., 2009) treatment for detaching biofilms.
However, using the whole plant material for DNA extraction, as performed in this study,
minimizes the potential loss of microbial cells, which is in accordance with other studies
(e.g., McEniry et al., 2008; Wang et al., 2010).
4.1.2 TRFLPanalysis
The TRFLP method is a valuable tool for the monitoring of microbial communities. It is
a high-through-put method, which allows a higher resolution of DNA fragments (Marsh,
1999) and therefore an improved comparison of samples in contrast to other
fingerprinting methods, such as the DGGE analysis. Hence, this method has been
widely used for gaining insights into different environmental habitats, such as soil (e.g.,
Lueders & Friedrich, 2000; Singh et al., 2006; Blackwood & Buyer, 2007; Ulrich et al.,
2008; Enwall & Halin, 2009; Cao et al., 2012), marine (e.g., Denaro et al., 2005;
Pereira et al., 2006; Dang et al., 2009; Opatkiewicz et al., 2009; Kim et al., 2011) or
lake habitats (e.g., Chan et al., 2005; Kim et al., 2011; Bai et al., 2012; Bhattarai et al.,
2012).
104
In agreement with all PCR-based techniques, which include the construction of gene
libraries, the TRFLP method can be affected by PCR biases, such as sequence errors
through misincorporation or the formation of chimeric sequences (v. Wintzingerode et
al., 1997). To optimize the PCR step, the influence of the number of PCR cycles on the
TRFLP profile was analyzed. The number of TRFs was increased by 29% comparing
the bacterial TRFLP assay after 15 and 35 PCR cycles, respectively. This increase in
the TRF number might be influenced by the emergence of chimeric sequences, which
increased by higher cycle numbers (v. Wintzingerode et al., 1997). Chimeric
sequences can be generated between two different targets with high sequence
similarity during DNA amplification. To lower the risk of chimera production during
PCR, the number of PCR cycles for rrs gene amplification was kept to 25 (for bacterial
analyses) and 28 (for archaeal analyses), respectively.
In addition, the impact of different bacterial and archaeal reverse primers on the
amplification of the rrs gene was analyzed by the TRFLP method. Each bacterial and
archaeal primer set reflected almost the same community. However, primers designed
for the amplification of Bacteria and Archaea do not equally comply with the two most
important criteria: to be specific for the designated target and also general enough to
amplify all target sequences (Schütte et al., 2008). This may lead to a distorted
representation of the microbial diversity; an important fact of all PCR-based methods.
Another important issue is the occurrence of pseudo-TRFs after TRFLP analysis.
Pseudo-TRFs can be produced by PCR and the subsequent restriction enzyme digest.
The formation of single strand DNA (ssDNA) sequences during PCR can reproducibly
lead to pseudo-TRFs. These ssDNA sequences can form secondary dsDNA structures,
which are recognized as target by restriction enzymes in the digestion step leading to
false fragments (Egert & Friedrich, 2003). To identify such pseudo-TRFs, a
construction of a gene library is required, as performed in this study. Pseudo-TRFs
showed no corresponding sequences in the rrs gene libraries constructed from the
same environmental samples. In this study, no corresponding rrs sequences were
identified for the bacterial TRFs 84, 208 and 374, which indicated putative pseudo-
TRFs.
The construction and analysis of such rrs gene libraries can also be used for the
assignment of TRFs to taxonomic groups and has an advantage over a comparison of
TRFs with in silico databases. In silico comparison bears the risk of false phylogenetic
assignments of TRFs due to the discrepancy of in silico cut TRFs and TRFs measured
105
in a capillary gel electrophoresis. Variations of up to several base pairs can occur as
reported by different authors (e.g., Liu et al., 1997; Kitts, 2001; Kaplan and Kitts, 2003;
Bukovska et al., 2010). In this study, the measured TRF length varied up to 6.7 bp in
comparison to the TRF length obtained after in silico cut of the respective rrs
sequence. These differences result from separation with capillary gel electrophoresis,
which is not only affected by the length of the DNA fragment, but also by its GC content
and the fluorescent dye attached to the DNA. To solve this issue, the rrs sequences
were also analyzed with the capillary gel electrophoresis in this study, which proved a
reliable assignment of TRFs obtained from environmental samples. In only one case,
the phylogenetic assignment of archaeal TRFs was ambiguous. Three different
archaeal rrs sequences, which were most similar to Methanobacterium,
Methanoculleus and Methanosaeta, resulted in the archaeal TRF 107. However, other
archaeal TRFs, showing a higher abundance than TRF 107 and also the bacterial
TRFs could be assigned unambiguously.
Another important point is the application of appropriate enzymes. Further, the use of
more than one enzyme can increase the resolution of the TRFLP approach. In this
study, two restriction enzymes Hin6I and MspI were used, showing the best resolution
for the diverse bacterial community. Engebretson and Moyer (2003) tested 18
restriction enzymes, showing a resolution of up to 70% of OTUs obtained in a
community modeled with more than 50 OTUs. Hence, the TRFLP method can reflect
the majority of a microbial community and therefore it is a valuable tool for comparative
community analyses. Nevertheless, it must be kept in mind that the most dominant
TRF of a TRFLP profile does not necessarily represent the most dominant
microorganism in the microbial community. This is particularly true for analyses based
on genes, which are present as multiple copies in the genome, such as the rrs gene
(Farrelly et al., 1995; Osborn et al., 2000) used in this study.
Although this method was introduced in 1997, the processing of TRFLP data is still not
consistent in scientific studies. For instance, the analysis of the TRFLP data can be
conducted on the basis of the TRF height or area. Different studies have been
performed favoring the TRF height (Dunbar et al., 2001; Caffaro-Filho et al., 2007) or
TRF area (Kitts, 2001; Sait et al., 2003). However, there are also some studies having
no preference or showing positive and negative aspects for both, the TRF height and
area based analysis (Lueders & Friedrich, 2003; Schütte et al., 2008). This was
supported by the findings in this study showing no strong differences between the
106
TRFLP profiles analyzed on the basis of the TRF height or area. However, after data
processing, TRFLP results are suitable for comparative analyses and therefore for the
tracking of microbial community dynamics.
4.1.3 Metagenomicandbioinformaticanalysis
Metagenomic studies, the so-called high-throughput DNA sequencing approaches,
have been widely applied to study different habitats, such as the guts of mice and
termites (e.g., Turnbaugh et al., 2006; Warnecke et al., 2007), soils (e.g., Leininger et
al., 2006), but also mesophilic biogas systems (Krause et al., 2008a; Schlüter et al.,
2008). In contrast to conventional DNA sequencing technologies, such as the Sanger
dideoxy sequencing method, the 454-pyrosequencing technology allows a markedly
increased sequence throughput at less cost and time (Margulies et al., 2005).
Furthermore, the analysis of microbial metagenomes allows not only the
characterization of microbial communities, but also the determination of the genetic
potential for the expression of enzymes, such as carbohydrate degrading enzymes.
Another advantage is that this sequencing approach is not PCR-based and therefore is
not affected by PCR biases. Nevertheless, 454-pyrosequencing can also lead to
sequence errors. Huse and coworkers (2007) and Gilles and coworkers (2011) have
analyzed the older GS20 system and the new GS FLXTM Titanium system and obtained
a sequencing error rate of 0.49% and 1.07%, respectively, which is to a greater extent
linked to the presence of homopolymers (repetitive bases) in the sequence.
The phylogenetic assignment of hundreds of thousands metagenomic sequences to
taxonomic groups remains a sophisticated bioinformatic effort. This assignment is
based on a comparison of metagenomic sequences with entries in different databases
and is therefore influenced by the content of such databases. Due to database
limitations, only approximately 30% of the metagenomic sequences could be assigned
to taxonomic groups with the phylogenetic characterization methods applied.
Therefore, many metagenomic sequences remained unclassified indicating that a
considerable number of additional species might be involved in the production of
biogas.
107
The phylogenetic assignments of the RDP classifier and CARMA are based on two
different taxonomies. The RDP classifier relies on the taxonomy proposed by Garrity
and coworkers (2007), whereas CARMA based on the NCBI taxonomy (Wheeler et al.,
2007; Finn et al., 2008; Krause et al., 2008b). For instance, Cl. thermocellum was
classified to “Ruminococcaceae” by means of the RDP classifier, although the NCBI
taxonomy assigned this species to Clostridiaceae. Therefore, the unclassified members
of the family Ruminococcaceae, identified by means of the RDP classifier, correspond
most probably to Clostridium species as revealed by CARMA, using the NCBI
taxonomy.
Further differences between the results obtained from the RDP classifier and CARMA,
particularly at genus level, could also be explained by the different taxonomies. For
instance, the assignment of metagenomic sequences to archaeal genera revealed a
high number of Methanothermobacter as determined by CARMA, but when analyzed
by the RDP classifier Methanobacterium was more abundant.
Additionally, the assignment of metagenomic sequences by CARMA can also lead to
an overestimation of specific taxa. Krause and coworkers (2008b) have reported that
Proteobacteria were incorrectly assigned by a rate of 3.8%, using the CARMA
software. The comparison of CARMA results with results of the rrs gene library analysis
revealed Proteobacteria at 11% and 7%, respectively, supporting the results described
by Krause and coworkers (2008b).
Providing these concerns are kept in mind, the study of microbial metagenomes by
high-throughput sequencing is a promising tool for gaining detailed information about
the microbial community and its genetic potential.
4.1.4 Quantificationofmicroorganisms
The qPCR and the FISH method are commonly used for the quantification of
microorganisms in different habitats, such as biogas reactors (e.g., Burrell et al., 2004;
Yu et al., 2005, 2006; Blume et al., 2010; Krakat et al., 2010b; Nettmann et al., 2010).
108
In this study, two different TaqMan approaches were used for qPCR analyses. The use
of such target specific approaches has clear advantages over dsDNA-intercalating
fluorescent dyes, such as SYBR green (Sharkey et al., 2004; Zhang & Fang, 2006).
The primer set 304 (TaqMan assay) developed in this study specifically targets a group
of the unclassified Lachnospiraceae as classified by RDP. In contrast, the classification
of the target sequence by NCBI revealed the highest sequence similarity to a member
of the Defluviitaleaceae family. This is another example for a different assignment of
the same rrs sequence using the RDP and NCBI taxonomies.
The DOPE-FISH method allows the quantification of microorganisms based on the
microscopical detection of cells. However, the application of the FISH method for
environmental samples is difficult. Studies on coastal water showed that less than 50%
of total cells were visualized by FISH applying the EUB 338 probe (Pernthaler et al.,
2002). In this study, the hybridization rate of probes by DOPE-FISH analysis was low
(<30%) despite various optimization methods performed in the context of the Diploma
thesis of Krumrei (2010). Impurities in environmental samples can infer the
visualization of cells due to high background fluorescence as indicated for biogas
samples by Nettmann and coworkers (2010). Further, it might be possible that some
microbial groups, which were currently uncharacterized, were not targeted by the
bacterial or archaeal probes used in this study. This was already detected for the EUB
338 probe a few years ago, which was then amended by the EUB II and III probes
(Daims et al., 1999).
In addition, a low ribosomal content as suggested for marine Actinobacteria by
Pernthaler and coworkers (2002) can also reduce the efficiency of probe hybridization.
A fact that can also influence the hybridization rate of other FISH analyses. In this
study, 50% of total cells were stained by PI indicating at least changes in the cell
status. This might affect the hybridization of probes and therefore might be also an
explanation for the low hybridization rate in this study.
109
4.2 Operationoftwophaseleachbedbiogassystems
The leach-bed biogas systems analyzed in this study were phase-separated and
optimized for the conversion of high-fiber substrate. During the whole experimental run,
various parameters, such as biogas and methane yield, VFA concentration and
degradation rate of ODM, were analyzed to monitor the performance of the systems.
4.2.1 Performanceofthebiogassystemsduringtemperatureincrease
The three identically constructed two-phase biogas systems showed similar reactor
performances. The parameters analyzed, such as the biogas yield, the pH, the
concentration of VFA and ammonia, resulted in similar values, indicating a reproducible
reactor performance of the three systems during the whole experimental run.
During temperature increase, no irreversible accumulation of VFA was determined,
which otherwise could lead to an inhibition of microorganisms and particularly to a
restriction of methanogenic archaea. Accordingly, the pH ranged between 6.6 and 8,
which means favorable conditions for the degradation of plant-derived biomass. Hu and
coworkers (2004) showed that the degradation efficiency of crystalline cellulose was
the highest between pH values of 6.8 to 7.3.
In the range of 55 to 65 °C, the overall reactor performance was efficient with high total
biogas and methane yields and high degradation rates. At temperatures of 70 °C, an
abrupt change in the reactor performance, including a decrease in the biogas and
methane yield and a decrease in the degradation rate, occurred. Studies on anaerobic
digestion processes have revealed a positive influence of temperature on the
production of methane from 20 to 60 °C (Ahring, 2003). Above a temperature of 60 °C,
a reduced reactor performance has been observed (Ahring, 2003). Although the overall
performance in the biogas system at 65 °C was still good and the total biogas yield
remained nearly the same, the total methane yield already decreased slightly by 8% at
this temperature. However, a strong reduction of the total biogas and methane yield
110
occurred at 70 °C with 29% and 37%, respectively, in comparison to the values
obtained at 60 °C.
In order to stabilize the microbial community at this high temperature, a
bioaugmentation with compost was performed at 70 °C. The 21-day fermentation after
the bioaugmentation at 70 °C led to improvements in ODM degradation by 9% and
biogas yield by 15 to 21%. A study of Neumann and Scherer (2011) also achieved an
increase in biogas yield of 6% after bioaugmentation compost during continuous
mesophilic fermentations.
The comparison of the bacterial community in this study before and after
bioaugmentation revealed only slight changes, which did not completely explain the
improved reactor performance in the two-phase biogas system. However, changes in
the bacterial community might be occurred during the fermentation with compost, which
still took effect on the 21-day fermentation process thereafter. However, the
improvements achieved by the bioaugmentation were not maintained during a further
temperature increase to 75 °C.
Nevertheless, repeated bioaugmentation with compost might be a valuable tool for the
stabilization of thermophilic fermentative bacterial communities. Furthermore, addition
of specialized bacteria, which are capable to degrade carbohydrates at higher
temperatures, could lead to further improvements. The first studies in this direction
have already been published by different authors (e.g., Bagi et al., 2007; Nielsen et al.,
2007), who have inoculated several reactor types with Caldicellulosiruptor.
4.2.2 Phaseseparationofthetwophasebiogassystem
Phase-separated biogas systems intend to separate process phases, such as
hydrolysis and acidogenesis from methanogenesis. The analysis of the metagenomes
derived from both the LBR and AF give first insights into the spatial distribution of the
microbial community and hence the phase-separation of the biogas system analyzed.
The metagenomes studied here revealed strong differences in the microbial community
composition. The genetic potential of bacteria for the expression of carbohydrate
111
degrading enzymes was strongly increased in the LBR in comparison to the AF. This is
in accordance with their role in the hydrolysis of plant-derived biomass, which is
supplied to the LBR. This finding supports the results obtained by Muha and coworkers
(2011). Their mathematical model of the same two-phase biogas system indicated that
the hydrolysis only takes place in the LBR when analyzing an LBR at 55 °C and an AF
at 38 °C. Additionally, the authors assumed that acidogenesis and acetogenesis most
probably take place in both reactor compartments (LBR and AF) in this two-phase
biogas system.
Furthermore, a spatial distribution of methanogens responsible for methane production
also occurred. The genetic potential for the expression of methanogenic enzymes was
increased in the AF in comparison to the LBR. This led to a high methane content of
70 to 81% in the biogas obtained from the AF in comparison to the LBR (21 - 46%
methane content). Hence, the intended spatial distribution of both the bacterial and
archaeal community was achieved at least in parts in this phase-separated biogas
system.
These findings also confirmed the results obtained by Talbot and coworkers (2010),
who also identified a spatial distribution of the microbial consortium analyzing a plug-
flow-type anaerobic bioreactor. Their study showed that acid producing bacteria as well
as acetoclastic methanogens are mainly located in the hydrolysis/acidification stage of
the reactor system.
4.3 Bacterialcommunityinthetwophasebiogassystem
The composition of the bacterial community was analyzed to achieve an overview
about the bacteria involved in the anaerobic digestion of plant-derived biomass and the
subsequent conversion pathways. Furthermore, the changes in the bacterial
community were monitored in detail during the fermentation of one load of substrate
and at increasing temperatures in the LBR.
112
4.3.1 DynamicchangesinthebacterialcommunityintheLBRduring
the21dayfermentation
The bacterial community residing in the biogas system shifted in a cyclic pattern during
the fermentation of one load of substrate at all LBR temperature regimes (55 - 75 °C).
Thereby, the substrate-attached microbial community introduced by the batch-feeding
every 21 days did not affect this dynamic bacterial community.
Changes in the bacterial community as a consequence of the anaerobic digestion
process are already known for other biogas reactor systems. Some studies have
shown that the bacterial community was highly dynamic in biogas systems, although
the whole reactor performance remained stable (Fernandez et al., 1999; Chackhiani et
al., 2004). Those studies analyzed small CSTRs (working volume <2 L), which were
continuously fed with glucose at mesophilic or with cattle manure at thermophilic
temperatures. An analysis of a continuous mesophilic biogas reactor with an increased
working volume of 16 L, converting organic household waste, also showed alterations
in the bacterial community after 44 and 90 days of fermentation (Cardinali-Rezende
et al., 2009). Furthermore, Kampmann and coworkers (2012) have shown that the
bacterial community composition in a mesophilic biogas system (working volume of
200 L) shifted after changes in the pH.
This is also true for the bacterial community in this study. The maximal concentration of
VFA was achieved during the first nine days of the fermentation. Simultaneously, the
pH decreased slightly in the LBR due to the VFA produced, but not below 6.6. Within
these nine days, the strongest alterations in the bacterial community were detected,
particularly at LBR temperatures from 55 to 65 °C. In this temperature range, different
members of the Clostridia class prevailed as consequence of the anaerobic digestion
process. For instance, the Defluviitalea genus (unclassified Lachnospiraceae as
determined by RDP classifier, TRF 304) emerged at 55 and 60 °C simultaneously to
the increase in VFA concentration and therefore may has a prominent role in the
anaerobic digestion process. For this group, a qPCR primer set was developed in this
study. Nevertheless, also other clostridial groups prevailed during the anaerobic
digestion process (e.g., TRF 167, Clostridium), for which a development of biomarkers
may also be of interest.
113
However, dynamic changes within the microbial community, as observed in this study,
may be the most important feature of a well-functioning anaerobic digestion process,
allowing coherent pathways for converting different intermediates into precursors to
methanogenesis as proposed by Pycke and coworkers (2011). They analyzed
full-scale mesophilic and thermophilic CSTRs and a mesophilic UASB and found
differences in the bacterial composition by 20 to 40% within 15 days, indicating a highly
dynamic system. The hypothesis of Pycke and coworkers (2011) was supported by the
findings of this study that revealed the most prominent changes in the bacterial
community at LBR temperatures of 55 to 60 °C, where the best reactor performances
were achieved.
4.3.2 Compositionanddynamicsofthebacterialcommunityinthe
LBRduringtemperatureincrease
The bacterial community was subject to alterations during temperature increases in the
LBR from 55 to 75 °C as indicated by TRFLP, rrs gene library and metagenomic
analyses.
Between 55 and 65 °C, members of the Clostridia were detected to be the dominant
part of the bacterial community in the LBR. Several genera belonging to the class
Clostridia were identified, indicating a high diversity within this class. Clostridium was
the most prevalent genus, followed by various genera, such as Alkaliphilus,
Caldicellulosiruptor, Defluviitalea, Desulfitobacterium, Moorella, Pelotomaculum,
Syntrophomonas and Thermoanaerobacter.
Numerous Clostridium species are known for their cellulolytic activity (Felix &
Ljungdahl, 1993; Bayer et al., 2004) and therefore for their potential key role in
anaerobic biomass degradation. Furthermore, the study of McDonald and coworkers
(2012) on anaerobic cellulose degradation showed that different members of the
Clostridium III cluster were associated with high cellulose degradation. In addition,
Thermoanaerobacter and Caldicellulosiruptor are also capable of degrading
carbohydrates (De Vos et al., 2009). For instance, Caldicellulosiruptor increases the
yield of methane after its addition to a thermophilic biogas fermenter as indicated by
114
Nielson and coworkers (2007). Further, species of the Moorella genus were also
predicted to ferment cellulosic material (Karita et al., 2003) and produce acetate as
fermentation end product (Drake & Daniel, 2004; De Vos et al., 2009).
Further degradation steps in the biogas formation process including oxidation of fatty
acids or degradation of proteins may be carried out by Syntrophomonas and
Alkaliphilus, respectively (McInerney et al., 1981; Takai et al., 2001), which were
detected in all metagenomes.
Hence, members of the Clostridia class play a key role in the degradation of plant-
derived biomass and the subsequent acido- and acetogenesis as mentioned in
previous studies (e.g., Klocke et al., 2007; Krause et al., 2008a; Schlüter et al., 2008;
Liu et al., 2009; Krakat et al., 2010b; Wirth et al., 2012).
Additionally, members of the Bacilli and Alpha-, Beta-, Gamma- and
Deltaproteobacteria classes were also detected in the biogas system at 55 to 65°C but
with lower abundances. Bacilli and Proteobacteria, in particular Deltaproteobacteria,
were recently found in a mesophilic biogas systems supplied with fodder beet silage or
maize silage (Klocke et al., 2007; Krause et al., 2008a). Members of the Thermotogae
were also identified in this study at temperatures of 55 to 65 °C, but also with minor
prevalence compared to Clostridia. Petrotoga, a genus of the Thermotogae class, was
found particularly at thermophilic temperatures of 55 °C. Some members of this genus,
such as Petrotoga mobilis are known to be fermentative and capable of degrading
xylan (Lien et al., 1998).
The high diversity of Clostridia and other fermentative bacteria at temperatures of 55 to
65 °C resulted in high degradation rates of ODM leading to high biogas yields and
efficient reactor performances.
After temperature increase to 70 °C, strong changes in the bacterial community and the
reactor performance occurred. Bacterial community changes were detected within all
three identically constructed biogas systems already after 7 days as determined by
TRFLP analysis.
Members of the Clostridia class were reduced and partially replaced by a member of
the Bacteroidia class, which was identified as being closely related to the
Acetomicrobium genus as resulted from TRFLP and rrs gene analyses. Further, the
metagenomic results also indicated an increase of the Bacillus (Bacilli), Bacteroides
(Bacteroidia) and Caldicellulosiruptor (Clostridia) genera.
115
Acetomicrobium species (Bacteroidia) grow at temperatures up to 75 °C (Krieg et al.,
2010). They degrade carbohydrates, such as glucose and cellobiose, which leads to
the fermentation end products acetate, lactate, ethanol, CO2 and H2 (Krieg et al.,
2010). In addition, species of the Bacteroides genus (Bacteroidia), identified by
metagenomic analysis, are saccharolytic bacteria producing mainly succinate and
acetate as fermentation end products (Krieg et al., 2010). In other studies, members of
the Bacteroidia were only detected in mesophilic and not in hyperthermophilic
fermentations. As example, Kongjan and coworkers (2011) did not identify Bacteroidia
when analyzing a two-stage UASB reactor system at 70 and 50 °C, supplied with
wheat hydrolysate. In contrast, some studies revealed Bacteroidetes as a prevalent
part of the bacterial community, but only in mesophilic biogas systems (Klocke et al.,
2007; Krause et al., 2008a; Schlüter et al., 2008; Liu et al., 2009; Kampmann et al.,
2012).
In addition, the Bacillus genus showed a high diversity, including species capable of
degrading carbohydrates (De Vos et al., 2009). Members of the Caldicellulosiruptor
genus have their growth optima between 65 and 75 °C. Various carbohydrates serve
as fermentable substrates leading to fermentation end products of acetate, ethanol,
CO2 and H2 (De Vos et al., 2009).
Furthermore at an LBR temperature of 75 °C, a member of the Thermotogae
(Fervidobacterium sp.) was detected in this two-phase biogas system by TRFLP
analysis. Members of the Fervidobacterium genus use various hexoses for energy
metabolism and produce lactate, CO2, H2, acetate as well as ethanol, but no butyrate
(Patel et al., 1985; Cai et al., 2007).
In accordance with the changes in the bacterial community, the intermediate production
changed in the biogas system at 70 °C. For instance, the concentration of the acetic
acid and butyric acid was reduced. The fermentation end products of the bacteria
prevalent at these temperatures may explain at least the strong reduction in n-butyric
acid. Accordingly, it has been reported that a certain number of extreme thermophilic
bacteria produce mainly acetate and H2 as end products and less butyrate and
propionate (Wiegel, 1992; Ahring, 2003).
Due to the fact that the intermediate butyrate can be converted into acetate (Wu et al.,
1993), the decrease in n-butyric acid also affects the acetic acid producing bacteria.
This could also explain the reduced acetic acid concentration at temperatures of
116
70 and 75 °C. A diminished consumption of the acetic acid by acid-consuming
microorganisms, such as syntrophic acetate-oxidizing bacteria or acetoclastic
methanogens, could also be assumed, due to a prolonged decomposition of acetic acid
in the leachate derived from the LBR at 70 and 75 °C. Ahring and coworkers (2001)
identified a disturbance in the proportion between acid-producing and acid-consuming
microorganisms in CSTRs already at 65 °C.
These changes in the bacterial community at 70°C finally lead to a strong reduction in
the performance of the two-phase biogas system.
4.4 Geneticpotentialforanaerobicdigestionofplantderived
biomass
The microbial degradation of biomass conducted by different bacteria is the time-
limiting step during the whole process of biomass conversion to methane. This study
focused on the degradation of carbohydrates, an important part of the whole biomass
degradation.
The lignin fraction of plant-derived biomass represents a source of energy, which is
currently not used by the degradation process due to the anaerobic conditions. Lignin
accounts for approximately 4% and 7% of the dry matter in the rye silage and straw
material used, respectively. However, to get more insights into the whole genetic
potential for carbohydrate degradation, the microbial potential for lignin degradation
was also investigated.
Only a minor number of EGTs was assigned to the protein family PF02578 encoding
multicopper polyphenol oxidoreductase laccases. Laccases seemed to play an
important role in lignin degradation not only in white-rot and brown-rot fungi, but also in
bacteria (Bugg et al., 2011). However, the assignment of EGTs, encoding enzymes
grouped in this protein family, revealed Alkaliphilus, Syntrophobacter and Clostridium,
which have not yet been associated with the degradation of lignin. Members of these
genera are strict anaerobic bacteria using proteinaceous substrates (Takai et al.,
117
2001), propionate (Boone & Bryant, 1980) or various substrates (e.g., Sleat et al.,
1984; Schnürer et al., 1996) for energy generation. In contrast, some studies have
shown that bacteria, involved in the degradation of lignin, belong to the Actinomycetes,
Alpha- and Gammaproteobacteria classes (e.g., Masai et al., 2007; Ahmad et al., 2010;
Bugg et al., 2011). Among those, members of the Streptomycetes, Sphingomonas and
Pseudomonas genera were identified. Although some metagenomic sequences could
be assigned to these genera (e.g., to Pseudomonas), the genetic potential of the
microbial community for lignin degradation in the biogas system analyzed was very
low.
This supports the results of the study of Jaenicke and coworkers (2011), who have
identified metagenomic sequences, which encode lignin degrading enzymes, only
rarely in a mesophilic biogas system. In comparison to that, sequences, encoding
cellulose degrading enzymes, were increased by a factor of 36 (Jaenicke et al., 2011).
Other metagenomic studies on biogas communities have also revealed a high genetic
potential particularly for carbohydrate degrading enzymes (Krause et al., 2008a;
Schlüter et al., 2008; Jaenicke et al., 2011; Wirth et al., 2012). A broad genetic
potential for the degradation of hemicelluloses and celluloses was also identified in the
microbial community of this study.
Interestingly, although the GH family profiles differed at least slightly in the
metagenomic samples analyzed in this study, GH families 3 and 43 were identified in
all metagenomes among the five most abundant GH families. These GH families were
also prevalently identified in metagenomes derived from a mesophilic and thermophilic
biogas system supplied with straw and hay and a full-scale, mesophilic biogas system
(Jaenicke et al., 2011; Rademacher et al., 2012; Hanreich, unpublished data). This
may indicate a potential key role of the GH families 3 and 43 in biomass degradation
independent of the biogas system analyzed. GH family 3 and 43 encompass enzymes,
such as β-glucosidase, xylan 1,4-β-xylosidase and β-xylosidase, arabinanase,
xylanase, respectively. These enzymes are involved in the breakdown of polymers,
such as xylan and cellulose.
However, to deepen the insights into the differences of the genetic potential for
carbohydrate degradation in the biogas system analyzed in this study, the abundance
of each GH family of interest derived from the four metagenomes was compared. The
GH families showed variations in abundance between the metagenomic samples. In
the biofilm sample from the AF almost all GH families of interest were decreased by a
118
factor of up to 20 and in extreme cases by a factor of 70 in comparison to the LBR
sample at 55 °C. Hence, the genetic potential for carbohydrate degradation seemed to
be strongly reduced in the AF.
Furthermore, the samples from the LBR at 55 and 65 °C revealed the most similar GH
family spectrum, indicating a comparable genetic potential for carbohydrate
degradation of the inherent microbial community in the biogas reactor at these
temperatures. In contrast, the metagenomic sample from the LBR at 70 °C showed
stronger alterations. The minor abundance of specific GH families (e.g., GH 8, 11) at
this temperature was mainly caused by a decrease of Firmicutes, particularly Clostridia,
as indicated by the taxonomic assignment of the corresponding EGTs. On the other
hand, the emergence of different GH families (e.g., GH 28, 39) was mainly caused by
members of the Bacteroidetes phylum, but also by Firmicutes (Clostridia),
Proteobacteria and Chloroflexi phyla. In particular the strong decrease of the number of
Firmicutes (Clostridia) and the emergence of Bacteroidetes are in agreement with the
alterations in the bacterial community after temperature increase to 70 °C as shown by
TRFLP and rrs gene analyses.
Although the total number of EGTs assigned to selected GH families was the highest in
the metagenomic sample at 70 °C, its composition and the assignment to taxonomic
groups changed. This might be the reason for the strong decrease in the degradation
rate of ODM at 70 °C. For instance, the GH families 9 and 5, encompassing the highest
number of up to now identified cellulase genes (Schülein, 2000), were less prevalent at
70 °C. Furthermore at this temperature, the presence of GH family 11, encompassing
the highest number of enzymes acting on xylan (Collins et al., 2005), was also
reduced. Hence, the genetic potential for carbohydrate degradation of the bacterial
community at hyperthermophilic temperatures was not as effective as the genetic
potential of the thermophilic community, although GH families with polysaccharide-
degrading activity were also detected.
The results clearly indicated that the prevalent GH families identified at 55 and 65 °C
might be important for efficient carbohydrate degradation due to the fact that the
highest biomass degradation was achieved at these temperatures. Hence these
GH families may provide a basis for the development of specific marker systems for the
monitoring of biogas reactors by qPCR analysis. Some studies have already focused
on the detection of genes encoding GH families 5 and 48 by qPCR (Pereyra et al.,
2010). Whereas the GH family 5 was identified with higher abundance in the
119
metagenomic samples at 55 and 65 °C, GH 48 was less frequently detected in this
study and in the metagenome of a mesophilic biogas community (Jaenicke et al., 2010;
Rademacher et al., 2012). However, the prevalently detected GH families 3 and 43
should also be considered for such a quantitative approach.
4.5 Putativepathogensinthetwophasebiogassystem
The presence of selected pathogens was analyzed to determine the potential risk for
infections of plants, animals and humans through field application of digestate.
The phylogenetic assignment of metagenomic sequences to plant pathogens, such as
Clavibacter michiganensis, Ps. syringae and Ralstonia solanacearum, was marginal in
the thermophilic two-phase biogas system. Furthermore, other widespread plant
pathogens, such as Rhizoctonia solani and Sclerotinia sclerotiorum, were not detected.
Other studies on this topic showed that plant pathogens did also not proliferate in
mesophilic biogas systems. The results of a project study (FNR, 2012d) showed that
different plant pathogens, such as Rhizoctonia solani and Sclerotinia sclerotiorum,
were inactivated after a few days of fermentation at mesophilic temperatures as
determined with culture-dependent methods. This was also confirmed by the study of
Kaemmerer and coworkers (2008), who analyzed different plant pathogens after
experimental and full-scale biogas fermentations. Furthermore, they showed that the
ensiling process also has the potential to inactivate plant pathogens.
Further potential risks might be derived from pathogens infecting animals or humans.
The number of metagenomic sequences, which could be assigned to potential animal
or human pathogens was very low. Only the Holin protein family was identified in
slightly higher abundances (PF05105). This family includes a protein, which is
supposed to be involved in toxin secretion in Cl. difficile (Tan et al., 2001). Other
protein families comprising pathogen-associated proteins were less frequently
identified in the two-phase biogas system. This might be achieved by the high reactor
temperatures and by the fact that the biogas system under study was supplied with
120
rye-silage and straw. Manure, a source of pathogens, such as Cl. botulinum, was only
added in the beginning of the reactor start-up.
However, other studies on this topic have also shown a low risk of proliferation of
potential pathogens even in mesophilic biogas systems supplied with manure. For
instance, Dohrmann and coworkers (2011), who have investigated mesophilic and
thermophilic biogas reactors supplied with different manures, did not detect pathogenic
Clostridium species after analyzing the rrs gene. In addition, no evidence for a
spreading of pathogens through digestate application was found by the study of
Eikmeyer and coworkers (accepted), who analyzed the four metagenomes of this study
and three further metagenomes from mesophilic biogas-producing communities. A
further work of Goberna and coworkers (2011) revealed a complete disappearance in
viable cells of E. coli and Salmonella sp. and a partial disappearance of Listeria sp.,
comparing fresh manure with manure fermented in a mesophilic biogas reactor.
Another study of Iwasaki and coworkers (2011) compared samples from a mesophilic
and thermophilic biogas plant, detecting viable numbers of the Coli-aerogenes
(Escherichia - Aerobacter) and Enterococcus group only in the mesophilic biogas plant.
However, a reduction in viable bacterial cells occurred during the subsequent
anaerobic digestion. Hence, a sanitation effect due to increased temperatures, but also
due to retention time might occur at least for specific bacterial groups.
In addition to that, Vinneras and coworkers (2006) indicated that the risk for an
unintended distribution of pathogens via the gas of biogas systems or through the
condensate water of gas pipes is also very low after applying culture-dependent
methods.
These results indicated that the risk of an unintended proliferation of potential
pathogens and therefore the risk for infections of plants, animal and humans after
applying digestate to agricultural land is rather low. Nevertheless, to deepen the
knowledge about the proliferation of pathogens in biogas systems and about
temperature-dependent sanitation effects, further studies need to be performed
applying culture-dependent methods or targeting specific genes of pathogens by
qPCR. The metagenomic results obtained in this study might serve as basis for these
studies.
121
4.6 Archaealcommunityinthetwophasebiogassystem
The production of methane is carried out by methanogenic archaea via three different
pathways: the hydrogenotrophic, acetoclastic and methylotrophic pathways, using CO2
and H2, acetate, methanol or carbon monoxide. To elucidate a potential predominance
of one of these pathways in the biogas reactor studied in this work, the composition of
the archaeal community was analyzed by metagenomic, TRFLP and rrs gene library
analyses.
4.6.1 CompositionofthearchaealcommunityintheLBRduring
temperatureincrease
The archaeal community in the LBR was analyzed for three different temperature
regimes (55, 65 and 70 °C) by metagenomic and TRFLP analysis. At an LBR
temperature of 55 °C, Methanosarcinales, i.e. Methanosarcina, were the most
prevalent methanogenic archaea in the LBR. At temperatures higher than 65 °C,
members of the Methanosarcinales order were less detected, indicating an inhibition of
growth. This is in accordance with the findings that thermophilic Methanosarcina
species, such as Msr. thermophila, grow at temperatures of up to 55 °C (Boone &
Castenholz, 2001).
On the other hand, members of the Methanobacteriales, i.e. Methanothermobacter,
were prevalently detected in the LBR at a temperature of 65 °C. At this temperature,
the hydrogenotrophic pathway was obviously the prevalent one for methane production
in the LBR. Methanothermobacter species have their growth optima between 55 and
65 °C (Boone & Castenholz, 2001), confirming the advantage of Methanothermobacter
over Methanosarcina in the LBR at temperatures of 65 °C.
At even higher temperatures, the TRFLP profiles also revealed Methanothermobacter
as prevalent. However, the metagenomic results indicated a strong growth inhibition of
the methanogenic archaea due to the fact that hardly any EGT could be assigned to
this taxonomic group. Although methanogens, such as members of the
122
Methanothermus and Methanocaldococcus genera, also grow at hyperthermophilic
temperatures (Boone & Castenholz, 2001), the members of the genera identified, such
as Methanosarcina and Methanothermobater, grew up to 55 and 65 °C, respectively,
supporting the assumed inhibition at 70 °C. This can be supported by the methane
yield obtained in the LBR. At 65 °C, the methane yield was reduced by 14%, whereas a
reduction of 59% occurred at 70 °C in comparison to the methane yield obtained at
55 °C. Furthermore, at an LBR temperature of 75 °C, only minor amounts of methane
were measured in the LBR.
4.6.2 CompositionofthearchaealcommunityintheAF
The archaeal community in the AF was tracked during the temperature increase in the
LBR. At an LBR temperature of 55 °C, Methanosaeta, a strict acetoclastic methanogen,
was detected with minor abundance in the AF. During the course of the experiment
Methanosaeta was strongly reduced or not found in the AF, although the temperature
of this reactor compartment, its pH and its concentration of free ammonia, remained
constant. For instance, a total ammonia nitrogen concentration of 0.85 to 1.00 g per L
and, due to the high temperature (55 °C) and the pH (7.80 - 7.95), a free ammonia
concentration of 0.21 to 0.27 g per L were determined in the AF. Although
Methanosaeta was found as predominant Archaea at total ammonia nitrogen
concentrations of up to 1.5 g per L (Karakashev et al., 2005; Nettmann et al., 2010),
they were not identified at free ammonia concentrations above 0.20 g per L in full scale
biogas plants (Nettmann et al., 2010). In this study, Methanosaeta was identified at
ammonia concentrations of 0.24 g per L in the AF at an LBR temperature of 55 °C.
However, this concentration might restrict the further proliferation of Methanosaeta in
the AF during the experimental runs. In addition, a VFA concentration of up to
5.03 g per L in the AF can also restrict Methanosaeta (Karakashev et al., 2005) and
instead favor the prevalence of Methanosarcina or strict hydrogenotrophic archaea.
Hence, members of the Methanosarcinales and Methanobacteriales were the most
prevalent orders in the AF independent of the LBR temperature and the analytical
approach. The major genera were determined to be Methanosarcina, a mixotrophic
123
methanogen, and Methanothermobacter (Methanobacterium as determined by RDP
classifier), a hydrogenotrophic methanogen. Furthermore, protein families, consisting of
enzymes for both pathways, were identified in the AF at least at an LBR temperature of
55 °C. During the following temperature increase in the LBR, both genera
(Methanosarcina and Methanothermobacter) showed an uneven progress in the AF.
Methanobacteriales, particularly Methanothermobacter (Methanobacterium as
determined by RDP classifier), which was prevalently detected in the AF, showed a
further increase at an LBR temperature of 70 °C. A study on a hyperthermophilic
two-stage UASB reactor, which was conducted at 70 and 55 °C and supplied with
wheat hydrolysate, also showed Methanothermobacter as prevalent (Kongjan et al.,
2011). An explanation for the increase in the biogas system studied in this work might
be that the concentration of VFA in the AF was clearly reduced at 70 °C and above. A
reduced acetic acid concentration combined with a thermophilic temperature of 55 °C
in the AF mean favorable conditions for acetate oxidizers acting with H2-scavenging
hydrogenotrophic archaea in syntrophy, having an advantage over the acetoclastic
methanogens (Zinder & Koch, 1984; Ahring 1995). This syntrophic interaction can be
conducted via interspecies H2 transfer, which has been described for
Methanothermobacter sp. and syntrophic oxidizing bacteria (Ishii et al., 2005).
Interestingly, Methanosarcina was also prevalent in the AF, particularly at higher LBR
temperatures. Some studies on archaeal communities in biogas reactors revealed a
high abundance of Methanosarcina (Methanosarcinaceae) in mesophilic and
thermophilic systems supplied with manure as main component (Karakashev et al.,
2005) or in thermophilic systems supplied with sludge (Lerm et al., 2012), but also in a
mesophilic anaerobic digester supplied with wine distillation waste (Godon et al., 1997).
Furthermore, an analysis of a thermophilic CSTR supplied with manure and maize
silage showed a potentially time-dependent increase of Methanosarcina during
14-weeks of fermentation (Bergmann, unpublished data). Hence, Methanosarcina
seemed to be widely distributed in various biogas systems. In contrast to the
acetoclastic Methanosaeta, Methanosarcina can tolerate free ammonia concentrations
of 0.45 g per L (Nettmann et al., 2010), which was underrun in the two-phase biogas
system studied in this work.
Furthermore, Methanosarcina can utilize acetate, but also CO2/H2, methanol and
carbon monoxide for methanogenesis. At hyperthermophilic temperatures in the LBR,
the acetic acid concentration, which altered during the 21-day fermentation period from
124
0.04 to 0.82 g per L (0.7 to 13.7 mM) in the AF, fell below the threshold for acetate
utilization of Methanosarcina in some cases. This threshold varies between 0.11 to
2.50 mM for Methanosarcina species from different anaerobic systems (Zinder, 1990;
Jetten et al., 1992; Mladenovska & Ahring, 2000). It is therefore conceivable that
Methanosarcina also utilized other methanogenic pathways than the acetoclastic one
depending on the apparent supply of acetic acid or other precursors. This is also
supported by the study of Karakashev and coworkers (2006), who have reported that
the acetoclastic pathway was only present in mesophilic and thermophilic reactors
when Methanosaetaceae were present. However, the use of CO2/H2 for the production
of methane has not been described for cultivable Methanosarcina strains growing at
thermophilic temperatures (Boone & Castenholz, 2001). Nevertheless, the potential
use of all methanogenic pathways could mean a key advantage over the strict
hydrogenotrophic methanogens in this reactor.
4.7 SyntrophicinteractionsofVFAoxidizingbacteriaand
methanogens
A further important point already mentioned is the syntrophic interaction of certain
bacteria with hydrogenotrophic methanogens, oxidizing acetate and other VFAs. The
bacterial community identified in the AF might interact in syntrophy with H2-scavenging
methanogens.
During the temperature increase in the LBR, the bacterial community in the AF showed
no strong alterations. Most of the prevalent bacterial TRFs that were identified in the
AF by TRFLP analysis showed the closest assignment to members of the Clostridia
class. The same is true for the metagenomic results, revealing a predominance of
Clostridia in the AF. Furthermore, the assignment at genus level revealed a prevalence
of Clostridium, Anaerobaculum, Thermacetogenium and Petrotoga, but also
Pelotomaculum and Syntrophomonas as determined either by RDP or CARMA.
The Clostridium genus, predominantly found in the biogas system under study, is
linked to the group of acetate-oxidizing bacteria (Schnürer et al., 1996). Certainly, the
125
acetate-oxidizing Cl. ultunense strain (NR_026531) grows at mesophilic and not at
thermophilic temperatures and showed only low rrs sequence identities of 81 to 88% to
TRF sequences assigned to the Clostridium genus (e.g., TRF 151, 161, 217).
However, currently unknown clostridial strains might be involved in syntrophic acetate
oxidation.
Members of the Anaerobaculum genus (Synergistia class) are not capable of
fermenting acetate, although the fermentation of various organic acids has been
reported (Rees et al., 1997; Menes & Muxi, 2002; Maune & Tanner, 2012). Therefore,
syntrophic acetate oxidation in cooperation with hydrogenotrophic methanogens by
members of the Anaerobaculum genus is rather unlikely. However, as members of this
genus are capable of producing acetate, they might be involved in the supply of acetate
for syntrophic oxidation. This might also be true for members of the Petrotoga genus,
which are fermentative bacteria producing, for instance, acetate as fermentation
product (Lien et al., 1998).
However, members of the Thermacetogenium genus were also prevalently identified in
the AF. The type species Thermacetogenium phaeum oxidizes acetate in syntrophic
association with Mtb. thermautotrophicus (Hattori et al., 2000). Due to the fact that
members of both genera were prevalently detected in the AF, a syntrophic conversion
of acetate to methane is likely.
Beside this potential syntrophic acetate oxidation, further syntrophic cooperations can
be assumed. Pelotomaculum, which was identified by CARMA, can oxidize lactate,
propionate and several alcohols in syntrophic cooperation with hydrogenotrophic
methanogens (Imachi et al., 2002). Moreover Syntrophomonas oxidizes fatty acids
under anaerobic conditions. This occurs in syntrophic association with H2-scavengers,
such as methanogens (McInerney et al., 1981).
4.8 DeterminationofcelldensitiesofBacteriaandArchaea
The quantitative detection of Bacteria and Archaea was performed by means of qPCR,
DAPI staining and DOPE-FISH analysis.
126
At LBR temperatures of 55 to 65 °C, the total cell count varied between 1.6 and
2.7 x 1010 per mL, whereas the number of Bacteria varied between 3.5 and 5.5 x 109
per mL. At these temperatures, the number of the bacterial rrs gene copies was higher
(1.3 x 1011 copies mL-1), presumably due to multiple copy numbers of the rrs gene in
cells. These results are in accordance with other quantitative studies in biogas
reactors. Blume and coworkers (2010) have identified approximately 1 x 109 to 1 x 1011
bacterial rrs gene copies per mL in their acidification experiment in a mesophilic,
laboratory-scale CSTR supplied with maize silage. Nettmann and coworkers (2010)
have detected lower total cell counts (1 - 9 x 108 mL-1), but also high bacterial rrs gene
copy numbers of 1 to 9 x 1011 per mL, analyzing six agricultural biogas plants operated
between 37 and 45 °C. The total cell counts detected in the study of Krakat and
coworkers (2010a) were also in line with this study. They detected 1 to 3 x 1010 total
cells per mL in a thermophilic biogas fermenter supplied with fodder beet silage and
juice.
Further increase in the operation temperature led to a reduction in cell numbers by
nearly one log in this study. Still, average values of 5.2 x 109 total cells per mL and
3.5 x 1010 bacterial rrs gene copies per mL were obtained at LBR temperatures of
70 and 75 °C. Although reduced, these values showed no strong differences to the
total cell and rrs gene copy numbers obtained by the studies of Nettmann and
coworkers (2010), analyzing full-scale biogas systems, and Blume and coworkers
(2010), analyzing a mesophilic lab-scale system. Hence, an evidence for the decrease
in the reactor performance due to reduced cell numbers is rather unlikely.
The number of archaeal cells represented only 0.3 to 0.6% of the total cell count. In the
full-scale biogas systems analyzed by Nettmann and coworkers (2010), the number of
archaeal cells represented even 3 to 7% of the total cells. However, although the
number of archaeal cells was low, some studies indicated that Archaea are highly
active in different biogas systems leading to high methane rates. For instance,
Zakrzewski and coworkers (2012) have suggested that Archaea are highly
transcriptionally active analyzing the metatranscriptome of a mesophilic biogas-
producing community. Furthermore, a metaproteome study on a thermophilic biogas
reactor revealed a considerable number of enzymes relevant for methanogenesis also
implying a high transcriptional activity of methanogenic archaea (Hanreich et al., 2012).
The quantification of specific taxonomic groups relevant for the anaerobic digestion and
the downstream methanogenesis is of major importance for gaining further insights into
127
the biogas-producing community and for monitoring changes within the microbial
community. Particularly, bacteria involved in the time-limiting step of carbohydrate
degradation, were of major interest. Therefore, a TaqMan primer-set for qPCR analysis
was developed for the monitoring of a potentially process-relevant bacterium, which
was dominantly detected by TRFLP analysis at LBR temperatures of 55 and 60 °C.
This specific primer set 304 targets a small group of unclassified Lachnospiracea,
representing bacteria whose rrs sequence resulted in a TRF 304 after TRFLP analysis.
At thermophilic temperatures, the number of rrs gene copies of these bacteria
accounted to 2 to 3% of the total bacterial rrs gene copies, which implies an
accumulation of these bacteria (unclassified Lachnospiraceae) at least at thermophilic
temperatures in the system studied here. Furthermore, these bacteria could also be
detected in another thermophilic biogas reactor system (UASSR) as shown in this
study, indicating a distribution and a putative process-relevance also in other
thermophilic biogas systems.
The development of primer sets, such as the primer set 304, is of importance for the
monitoring of biogas processes. Other studies have also developed primer sets for the
detection of process-relevant bacteria. For instance, McDonald and coworkers (2012)
have shown that qPCR analyses of different Clostridium clusters (III, IV, XIV) revealed
an emergence of these taxonomic groups in different leachate samples derived from a
landfill site. In most cases, the rrs gene copies of the Clostridium clusters showed a
relative abundance of 1 to 3% of the bacterial rrs gene copies. In two cases, the
relative abundance of the Clostridium III and XIV cluster increased up to approximately
7% and 17%, respectively. Furthermore, Bauer and coworkers (2012) developed
specific qPCR primer sets for different bacterial groups, such as for a cluster of the
Ruminococcaceae, Prevotellaceae and Syntrophomonadaceae, which are also
involved in the biogas formation process. They have demonstrated a temperature-
dependent proliferation of these groups. For instance, members of the
Syntrophomonadaceae evolved at thermophilic temperatures, whereas members of the
Prevotellaceae at mesophilic temperatures.
In addition, the archaeal community responsible for methanogenesis has also been in
the focus of interest. Yu and coworkers (2005) developed TaqMan qPCR primers for
the quantification of methanogenic archaea, which served as a basis for the monitoring
of the archaeal community in biogas plants by different authors (Yu et al., 2006; Blume
et al., 2010; Nettmann et al., 2010).
128
The development and application of such primer sets for the detection of specific
bacteria or archaea is of major importance for the monitoring of biogas systems and
may lead to a valuable tool for the detection of process instabilities in future.
129
5 Outlook
The results gained in this study provide a deeper understanding of biogas-producing
communities in phase-separated biogas systems operated at high temperatures
(55 - 75 °C). The impact of temperature increase on the microbial biogas community
and the overall reactor performance was shown as well as the dynamic changes in the
bacterial community as a consequence of the anaerobic digestion of the substrate
supplied.
A certain number of unclassifiable and hence currently unknown species were also
found in the biogas system. In future, culture-independent methods must be amended
by culture-dependent methods, such as isolation and characterization of microbial
strains. This will lead to an improved interpretability of results obtained by culture-
independent methods.
Nevertheless, this study identified bacteria and glycoside hydrolases, which are
potentially relevant for the degradation of carbohydrates. Specific markers for these
targets can be used for the monitoring of those potentially process-relevant, but
uncultured microorganisms. This is of major importance due to the fact that the majority
of microorganisms in biogas systems are still uncultivated. In a first approach, the
specific detection of one potentially process-relevant bacterium, belonging to the
unclassified Lachnospiraceae, was shown for the biogas system under study and also
for another thermophilic biogas system. In future, such marker-based approaches may
facilitate not only the monitoring of biogas communities, but also the detection of
emerging disturbances in biogas systems.
130
In addition, the results also indicated that an unintended proliferation of pathogens
infecting plants, animals or humans is very low, which may help to improve the public
opinion about the risk of biogas plants.
Furthermore, the results of this study provide insights into the operation and
optimization of such phase-separated systems. The best reactor performance of the
system was achieved between 55 and 60 °C. Although dynamic changes in the
bacterial community occurred between these temperatures, the community was mostly
dominated by the same members of the Clostridia class. Therefore, the bacterial
community of this biogas reactor seemed to be resilient to temperature fluctuations in
this temperature range (55 - 60 °C), which is important for reactor operation. In
addition, an option for an optimization of such biogas systems is bioaugmentation with
compost or microbial strains. The results indicated that a bioaugmentation could be
advantageous, although this effect was not persistent in this study. Future studies need
to evaluate the positive effect and profitability of continuous bioaugmentation.
Hence, the results gained in this study provide not only a promising basis for the
monitoring of biogas communities, but also for a prospective improvement of
thermophilic biogas systems with phase-separation.
131
6 REFERENCES
454 Life Science (2012) Performance of the GS FLX+ System. URL:
http://454.com/products/gs-flx-system/index.asp. Last access December 13, 2012.
Abdo Z, Schüette UME, Bent SJ, Williams CJ, Forney LJ & Joyce P (2006) Statistical methods
for characterizing diversity of microbial communities by analysis of terminal restriction
fragment length polymorphisms of 16S rRNA genes. Environ Microbiol 8, 929-938.
Ahmad M, Taylor CR, Pink D, Burton K, Eastwood D, Bending GD & Bugg TD (2010)
Development of novel assays for lignin degradation: comparative analysis of bacterial and
fungal lignin degraders. Mol Bio Syst 6, 815-821.
Ahring BK (1995) Methanogenesis in thermophilic biogas reactors. Anton Leeuw 67, 91-102.
Ahring BK, Ibrahim AA & Mladenovska Z (2001) Effect of temperature increase from 55 to 68 °C
on performance and microbial population dynamics of an anaerobic reactor treating cattle
manure. Water Res 35, 2446-2452.
Ahring BK (2003) Perspectives for anaerobic digestion. In Biomethanation I ed. Scheper, T, pp.
1-30. Berlin Heidelberg, Germany: Springer.
Altschul SF, Gish W, Miller W, Myers EW & Lipman DJ (1990) Basic local alignment search tool.
J Mol Biol 215, 403-410.
Amann RI, Binder BJ, Olson RJ, Chisholm SW, Devereux R & Stahl DA (1990) Combination of
16S ribosomal-RNA-targeted oligonucleotide probes with flow-cytometry for analyzing
mixed microbial-populations. Appl Environ Microbiol 56, 1919-1925.
Amann RI, Ludwig W & Schleifer KH (1995) Phylogenetic dentification and in-situ detection of
individual microbial cells without cultivation. Microbiol Rev 59, 143-169.
Anthonisen AC, Loehr RC, Prakasam TBS & Srinath EG (1976) Inhibition of nitrification by
ammonia and nitrous acid. J Water Pollut Control Fed 48, 835-852.
Applied Biosystems by Life Technologies (2011) 5500 SOLiD™ System. URL:
http://www3.appliedbiosystems.com/cms/groups/global_marketing_group/documents/gener
aldocuments/cms_088662.pdf. Last access November 23, 2012.
Ashelford KE, Chuzhanova NA, Fry JC, Jonas AJ & Weightman AJ (2006) New screening
software shows that most recent large 16S rRNA gene clone libraries contain chimeras.
Appl Environ Microbiol 72, 5734-5741.
132
Bagi Z, Acs N, Balint B, Horvath L, Dobo K, Perei KR, Rakhely G & Kovacs KL (2007)
Biotechnological intensification of biogas production. Appl Microbiol Biotechnol 76, 473-
482.
Bai Y, Shi Q, Wen D, Li Z, Jefferson WA, Feng C & Tang X (2012) Bacterial communities in the
sediments of Dianchi Lake, a partitioned eutrophic waterbody in China. Plos One 7,
e37796.
Balk M, Weijma J & Stams AJM (2002) Thermotoga lettingae sp. nov., a novel thermophilic,
methanol-degrading bacterium isolated from a thermophilic anaerobic reactor. Int J Syst
Evol Micr 52, 1361-1368.
Barnby FM, Morpeth FF & Pyle DL (1990) Endopolygalacturonase production from
Kluyveromyces marxianus. 1. resolution, purification, and partial characterization of the
enzyme. Enzyme Microb Technol 12, 891-897.
Bauer C, Korthals M, Gronauer A & Lebuhn M (2008) Methanogens in biogas production from
renewable resources - a novel molecular population analysis approach. Water Sci Technol
58, 1433-1439.
Bauer C, Marin-Perez C, Munk B & Lebuhn M (2012) Bakterielle Populationen in der
Hydrolysephase der Vergärung einer Stroh- und Heumischung - Bacterial populations in
the hydrolysis phase of an anaerobically digested straw/hay mixture. In Bornimer
Agrartechnische Berichte 79, pp. 87-98. Potsdam, Germany: Leibniz-Institut für
Agrartechnik Potsdam-Bornim e.V.
Bayer EA, Setter E & Lamed R (1985) Organization and distribution of the cellulosome in
Clostridium thermocellum. J Bacteriol 163, 552-559.
Bayer EA & Lamed R (1986) Ultrastructure of the cell-surface cellulosome of Clostridium
thermocellum and its interaction with cellulose. J Bacteriol 167, 828-836.
Bayer EA, Morag E & Lamed R (1994) The cellulosome - a treasure-trove for biotechnology.
Trends Biotechnol 12, 379-386.
Bayer EA, Belaich J-P, Shoham Y & Lamed R (2004) The cellulosomes: multienzyme machines
for degradation of plant cell wall polysaccharides. Annu Rev Microbiol 58, 521-554.
Belloc F, Dumain P, Boisseau MR, Jalloustre C, Reiffers J, Bernard P & Lacombe F (1994) A
flow cytometric method using Hoechst-33342 and propidium iodide for simultaneous cell-
cycle analysis and apoptosis determination in unfixed cells. Cytometry 17, 59-65.
Bench SR, Hanson TE, Williamson KE, Ghosh D, Radosovich M, Wang K & Wommack K
(2007) Metagenomic characterization of Chesapeake bay virioplankton. Appl Environ
Microbiol 73, 7629-7641.
Bergmann I, Nettmann E, Mundt K & Klocke M (2010) Determination of methanogenic Archaea
abundance in a mesophilic biogas plant based on 16S rRNA gene sequence analysis. Can
J Microbiol 56, 440-444.
Bergmann I (2012) Characterization of methanogenic Archaea communities in biogas reactors
by quantitative PCR. Dissertation. Berlin, Germany: Technische Universität Berlin.
BfR – Bundesinstitut für Risikobewertung (2011) EHEC O104:H4-Ausbruchsgeschehen in
Deutschland aufgeklärt: Auslöser waren Sprossen von aus Ägypten importierten
Bockshornkleesamen. URL: http://www.bfr.bund.de/de/presseinformation/2011/21/ehec_
o104_h4_ausbruchsgeschehen_in_deutschland_aufgeklaert__ausloeser_waren_sprossen
_von_aus_aegypten_importierten_bockshornkleesamen-82843.html. Last access
December 10, 2012.
133
Bhattarai S, Ross KA, Schmid M, Anselmetti FS & Buergmann H (2012) Local conditions
structure unique archaeal communities in the anoxic sediments of meromictic Lake Kivu.
Microbiol Ecol 64, 291-310.
Biswas K & Turner SJ (2012) Microbial community composition and dynamics of moving bed
biofilm reactor systems treating municipal sewage. Appl Environ Microbiol 78, 855-864.
Blackwood CB & Buyer JS (2007) Evaluating the physical capture method of terminal restriction
fragment length polymorphism for comparison of soil microbial communities. Soil Biol
Biochem 39, 590-599.
Blaut M (1994) Metabolism of methanogens. Anton Leeuw Int J G 66, 187-208.
Blume F, Bergmann I, Nettmann E, Schelle H, Rehde G, Mundt K & Klocke M (2010)
Methanogenic population dynamics during semi-continuous biogas fermentation and
acidification by overloading. J Appl Microbiol 109, 441-450.
BMU ed. - Bundesministerium für Umwelt, Naturschutz und Reaktorsicherheit (2012)
Erneuerbare Energien in Zahlen - Nationale und internationale Entwicklung. pp. 1-136.
Berlin, Germany: BMU.
Böckelmann U, Szewzyk U & Grohmann E (2003) A new enzymatic method for the detachment
of particle associated soil bacteria. J Microbiol Methods 55, 201-211.
Boone DR & Bryant MP (1980) Propionate-degrading bacterium, Syntrophobacter wolinii sp.
nov. gen. nov., from methanogenic ecosystems. Appl Environ Microbiol 40, 626-632.
Boone DR & Castenholz RW eds. (2001) Volume One: The Archaea and the deeply branching
and phototrophic Bacteria. In Bergey's Manual® of Systematic Bacteriology ed. Garrity,
GM, pp. 1-721, New York, USA: Springer.
Brosius J, Palmer ML, Kennedy PJ & Noller HF (1978) Complete nucleotide sequence of a 16S
ribosomal RNA gene from Escherichia coli. Proc Natl Acad Sci USA 75, 4801-4805.
Bugg TD, Ahmad M, Hardiman EM & Singh R (2011) The emerging role for bacteria in lignin
degradation and bio-product formation. Curr Opin Biotechnol 22, 394-400.
Bukovska P, Jelinkova M, Hrselova H, Sykorova Z & Gryndler M (2010) Terminal restriction
fragment length measurement errors are affected mainly by fragment length, G+C
nucleotide content and secondary structure melting point. J Microbiol Methods 82, 223-
228.
Burrell PC, O'Sullivan C, Song H, Clarke WP & Blackall LL (2004) Identification, detection, and
spatial resolution of Clostridium populations responsible for cellulose degradation in a
methanogenic landfill leachate bioreactor. Appl Environ Microbiol 70, 2414-2419.
Caffaro-Filho RA, Fantinatti-Garboggini F & Durrant LR (2007) Quantitative analysis of terminal
restriction fragment length polymorphism (T-RFLP) microbial community profiles: peak
height data showed to be more reproducible than peak area. Braz J Microbiol 38, 736-738.
Cai J, Wang Y, Liu D, Zeng Y, Xue Y, Ma Y & Feng Y (2007) Fervidobacterium changbaicum
sp. nov., a novel thermophilic anaerobic bacterium isolated from a hot spring of the
Changbai Mountains, China. Int J Syst Evol Microbiol 57, 2333-2336.
Cantarel BL, Coutinho PM, Rancurel C, Bernard T, Lombard V & Henrissat B (2009) The
Carbohydrate-Active EnZymes database (CAZy): an expert resource for glycogenomics.
Nucleic Acids Res 37, D233-D238.
Cao P, Zhang LM, Shen JP, Zheng YM, Di HJ & He JZ (2012) Distribution and diversity of
archaeal communities in selected Chinese soils. FEMS Microbiol Ecol 80, 146-158.
134
Carballa M, Smits M, Etchebehere C, Boon N & Verstraete W (2011) Correlations between
molecular and operational parameters in continuous lab-scale anaerobic reactors. Appl
Microbiol Biotechnol 89, 303-314.
Cardinali-Rezende J, Debarry RB, Colturato LF, Carneiro EV, Chartone-Souza E & Nascimento
AM (2009) Molecular identification and dynamics of microbial communities in reactor
treating organic household waste. Appl Microbiol Biotechnol 84, 777-789.
Cattell RB (1966) The scree test for the number of factors. Multivariate Behavioral Research 1,
245-276.
Chackhiani M, Dabert P, Abzianidze T, Partskhaladze G, Tsiklauri L, Dudauri T & Godon JJ
(2004) 16S rDNA characterization of bacterial and archaeal communities during start-up of
anaerobic thermophilic digestion of cattle manure. Bioresour Technol 93, 227-232.
Chan OC, Claus P, Casper P, Ulrich A, Lueders T & Conrad R (2005) Vertical distribution of
structure and function of the methanogenic archaeal community in Lake Dagow sediment.
Environ Microbiol 7, 1139-1149.
Chao A (1987) Estimating the population size for capture-recapture data with unequal
catchability. Biometrics 43, 783-791.
Chao A, Chazdon RL, Colwell RK & Shen TJ (2005) A new statistical approach for assessing
similarity of species composition with incidence and abundance data. Ecol Lett 8, 148-159.
Chen X & Stewart PS (2000) Biofilm removal caused by chemical treatments. Water Res 34,
4229-4233.
Chouari R, Le Paslier D, Daegelen P, Ginestet P, Weissenbach J & Sghir A (2005) Novel
predominant archaeal and bacterial groups revealed by molecular analysis of an anaerobic
sludge digester. Environ Microbiol 7, 1104-1115.
Clarke KR (1993) Nonparametric multivariate analyses of changes in community structure.
Austral Ecol 18, 117-143.
Claus H (2004) Laccases: structure, reactions, distribution. Micron 35, 93-96.
Cole JR, Wang Q, Cardenas E et al. (2009) The Ribosomal Database Project: improved
alignments and new tools for rRNA analysis. Nucleic Acids Res 37, D141-D145.
Collins T, Gerday C & Feller G (2005) Xylanases, xylanase families and extremophilic
xylanases. FEMS Microbiol Rev 29, 3-23.
Colwell RK (2009) EstimateS: Statistical estimation of species richness and shared species
from samples. URL: http://purl.oclc.org/estimates. Last access November 15, 2012.
Culman SW, Bukowski R, Gauch HG, Cadillo-Quiroz H & Buckley DH (2009) Software open
access T-REX: software for the processing and analysis of T-RFLP data. BMC
Bioinformatics 10, 171.
Dahlenborg M, Borch E & Radström P (2003) Prevalence of Clostridium botulinum types B, E
and F in faecal samples from Swedish cattle. Int J Food Microbiol 82, 105-110.
Daims H, Bruhl A, Amann R, Schleifer KH & Wagner M (1999) The domain-specific probe
EUB338 is insufficient for the detection of all Bacteria: Development and evaluation of a
more comprehensive probe set. Syst Appl Microbiol 22, 434-444.
Dang H, Li J, Chen M, Li T, Zeng Z & Yin X (2009) Fine-scale vertical distribution of bacteria in
the East Pacific deep-sea sediments determined via 16S rRNA gene T-RFLP and clone
library analyses. World J Microbiol Biotechnol 25, 179-188.
135
Danon M, Franke-Whittle IH, Insam H, Chen Y & Hadar Y (2008) Molecular analysis of bacterial
community succession during prolonged compost curing. FEMS Microbiol Ecol 65, 133-
144.
Dashtban M, Schraft H, Syed TA & Qin W (2010) Fungal biodegradation and enzymatic
modification of lignin. Int J Biochem Mol Biol 1, 36-50.
Davey HM & Hexley P (2011) Red but not dead? Membranes of stressed Saccharomyces
cerevisiae are permeable to propidium iodide. Environ Microbiol 13, 163-171.
DBFZ ed. - Deutsches BiomasseForschungsZentrum (2011) Monitoring zur Wirkung des
Erneuerbare-Energien-Gesetz (EEG) auf die Entwicklung der Stromerzeugung aus
Biomasse (interim report). pp. 1-124. Leipzig, Germany: DBFZ.
De Vos P, Garrity GM, Jones D, Krieg NR, Ludwig W, Rainey FA, Schleifer K-H & Whitman WB
eds. (2009) Volume three: The Firmicutes. In BERGEY'S MANUAL® OF Systematic
Bacteriology ed. Whitman, WB, pp. 1-1450. New York, USA: Springer.
DeLong EF (1992) Archaea in coastal marine environments. Proc Natl Acad Sci USA 89, 5685-
5689.
Demirer GN & Chen S (2005) Two-phase anaerobic digestion of unscreened dairy manure.
Process Biochem 40: 3542-3549.
Denaro R, D'Auria G, Di Marco G, Genovese M, Troussellier M, Yakimov MM & Giuliano L
(2005) Assessing terminal restriction fragment length polymorphism suitability for the
description of bacterial community structure and dynamics in hydrocarbon-polluted marine
environments. Environ Microbiol 7, 78-87.
Deppenmeier U, Müller V & Gottschalk G (1996) Pathways of energy conservation in
methanogenic archaea. Arch Microbiol 165, 149-163.
DERA ed. - Deutsche Rohstoffagentur in der Bundesanstalt für Geowissenschaften und
Rohstoffe (2011) DERA Rohstoffinformation - Reserven, Ressourcen und Verfügbarkeit
von Energierohstoffen 2011, pp. 1-92. Hannover, Germany: DERA.
Despres VR, Nowoisky JF, Klose M, Conrad R, Andreae MO & Pöschl U (2007)
Characterization of primary biogenic aerosol particles in urban, rural, and high-alpine air by
DNA sequence and restriction fragment analysis of ribosomal RNA genes. Biogeosciences
4, 1127-1141.
Dohrmann AB, Baumert S, Klingebiel L, Weiland P & Tebbe CC (2011) Bacterial community
structure in experimental methanogenic bioreactors and search for pathogenic clostridia as
community members. Appl Microbiol Biotechnol 89, 1991-2004.
Downs CG, Brady CJ & Gooley A (1992) Exopolygalacturonase protein accumulates late in
peach fruit ripening. Physiol Plant 85, 133-140.
Drake HL & Daniel SL (2004) Physiology of the thermophilic acetogen Moorella thermoacetica.
Res Microbiol 155, 422-436.
Droege M & Hill B (2008) The Genome Sequencer FLX™ System - longer reads, more
applications, straight forward bioinformatics and more complete data sets. J Biotechnol
136, 3-10.
Dugba PN & Zhang RH (1999) Treatment of dairy wastewater with two-stage anaerobic
sequencing batch reactor systems - thermophilic versus mesophilic operations. Bioresour
Technol 68, 225-233.
Dunbar J, Ticknor LO & Kuske CR (2001) Phylogenetic specificity and reproducibility and new
method for analysis of terminal restriction fragment profiles of 16S rRNA genes from
bacterial communities. Appl Environ Microbiol 67, 190-197.
136
Egert M & Friedrich MW (2003) Formation of pseudo-terminal restriction fragments, a PCR-
related bias affecting terminal restriction fragment length polymorphism analysis of
microbial community structure. Appl Environ Microbiol 69, 2555-2562.
Eikmeyer F, Rademacher A, Hanreich A et al. (2013) Detailed analysis of metagenome
datasets obtained from biogas-producing microbial communities residing in biogas reactors
does not indicate the presence of putative pathogenic microorganisms. Biotechnology for
Biofuels 6, 49.
Ellis JT, Tramp C, Sims RC & Miller CD (2012) Characterization of a methanogenic community
within an algal fed anaerobic digester. ISRN Microbiology 2012, 12pp.
Engebretson JJ & Moyer CL (2003) Fidelity of select restriction endonucleases in determining
microbial diversity by terminal-restriction fragment length polymorphism. Appl Environ
Microbiol 69, 4823-4829.
Enwall K & Hallin S (2009) Comparison of T-RFLP and DGGE techniques to assess denitrifier
community composition in soil. Lett Appl Microbiol 48, 145-148.
Farrelly V, Rainey FA & Stackebrandt E (1995) Effect of genome size and rrn gene copy
number on PCR amplification of 16S rRNA genes from a mixture of bacterial species. Appl
Environ Microbiol 61, 2798-2801.
Felix CR & Ljungdahl LC (1993) The cellulosome: the exocellular organelle of Clostridium. Annu
Rev Microbiol 47, 791-819.
Feng XM, Karlsson A, Svensson BH & Bertilsson S (2010) Impact of trace element addition on
biogas production from food industrial waste - linking process to microbial communities.
FEMS Microbiol Ecol 74, 226-240.
Ferguson RL, Buckley EN & Palumbo AV (1984) Response of marine bacterioplankton to
differential filtration and confinement. Appl Environ Microbiol 47, 49-55.
Fernandez A, Huang SY, Seston S, Xing J, Hickey R, Criddle C & Tiedje J (1999) How stable is
stable? Function versus community composition. Appl Environ Microbiol 65, 3697-3704.
Finn RD, Tate J, Mistry J et al. (2008) The Pfam protein families database. Nucleic Acids Res
36, D281-D288.
FNR ed. - Fachagentur Nachwachsende Rohstoffe e.V. (2010) Leitfaden Biogas - Von der
Gewinnung zur Nutzung. pp. 1-274. Gülzow-Prüzen, Germany: FNR.
FNR ed. - Fachagentur Nachwachsende Rohstoffe e.V. (2012a) Biogas. pp. 1-44. Gülzow-
Prüzen, Germany: FNR.
FNR ed. - Fachagentur Nachwachsende Rohstoffe e.V. (2012b) Biomethan. pp. 1-25. Gülzow-
Prüzen, Germany: FNR.
FNR ed. - Fachagentur Nachwachsende Rohstoffe e.V. (2012c) Bioenergie - die vielfältige
erneuerbare Energie. pp. 1-52. Gülzow-Prüzen, Germany: FNR.
FNR ed. - Fachagentur Nachwachsende Rohstoffe e.V. (2012d) Verbundvorhaben:
Untersuchungen zum phytosanitären Risiko durch die anaerobe Vergärung von
pflanzlichen Biomassen in Biogasanlagen; Teilvorhaben 1 (Projekt 22013207). URL:
http://www.fnr-server.de/ftp/pdf/berichte/22013207.pdf. Last access January 10, 2013
Fröhling A (2010) A flow cytometric approach to monitor the effects of gentle preservation
techniques in the postharvest chain. Dissertation. Berlin, Germany: Technische Universität
Berlin.
Gaillard BD (1965) Comparison of the hemicelluloses from plants belonging to two different
plant families. Phytochemistry 4, 631-634.
137
Garrity GM, Lilburn TG, Cole JR, Harrison SH, Euzeby J & Tindall BJ (2007) The taxonomic
outline of Bacteria and Archaea. TOBA Release 7.7, March 2007. Michigan State
University Board of Trustees. (http://www.taxonomicoutline.org/)
Gartemann KH, Kirchner O, Engemann J, Gräfen I, Eichenlaub R & Burger A (2003) Clavibacter
michiganensis subsp. michiganensis: first steps in the understanding of virulence of a
gram-positive phytopathogenic bacterium. J Biotechnol 106, 179-191.
Gilbert M, Brisson JR, Karwaski MF, Michniewicz J, Cunningham AM, Wu YY, Young NM &
Wakarchuk WW (2000) Biosynthesis of ganglioside mimics in Campylobacter jejuni
OH4384 - Identification of the glycosyltransferase genes, enzymatic synthesis of model
compounds, and characterization of nanomole amounts by 600-MHz H-1 and C-13 NMR
analysis. J Biol Chem 275, 3896-3906.
Gilles A, Meglecz E, Pech N, Ferreira S, Malausa T & Martin JF (2011) Accuracy and quality
assessment of 454 GS-FLX Titanium pyrosequencing. BMC Genomics 12, 245.
Giovannoni JJ, Dellapenna D, Bennett AB & Fischer RL (1989) Expression of a chimeric
polygalacturonase gene in transgenic rin (ripening inhibitor) tomato fruit results in
polyuronide degradation but not fruit softening. Plant Cell 1, 53-63.
Goberna M, Podmirseg SM, Waldhuber S, Knapp BA, Garcia C & Insam H (2011) Pathogenic
bacteria and mineral N in soils following the land spreading of biogas digestates and fresh
manure. Appl Soil Ecol 49, 18-25.
Godon JJ, Zumstein E, Dabert P, Habouzit F & Moletta R (1997) Molecular microbial diversity of
an anaerobic digestor as determined by small-subunit rDNA sequence analysis. Appl
Environ Microbiol 63, 2802-2813.
Good IJ (1953) The population frequencies of species and the estimation of population
parameters. Biometrika 40, 237-264.
Großkopf R, Janssen PH & Liesack W (1998) Diversity and structure of the methanogenic
community in anoxic rice paddy soil microcosms as examined by cultivation and direct 16S
rRNA gene sequence retrieval. Appl Environ Microbiol 64, 960-969.
Hammer Ø, Harper DAT & Ryan PD (2001) Past: Paleontological statistics software package for
education and data analysis. Palaeontol Electronica 4, 9pp.
Hanreich A, Heyer R, Benndorf D, Rapp E, Pioch M, Reichl U & Klocke M (2012) Metaproteome
analysis to determine the metabolically active part of a thermophilic microbial community
producing biogas from agricultural biomass. Can J Microbiol 58, 917-922.
Harmsen HJM, van Kuijk BLM, Plugge CM, Akkermans ADL, de Vos WM & Stams AJM (1998)
Syntrophobacter fumaroxidans sp. nov., a syntrophic propionate-degrading sulfate-
reducing bacterium. Int J Syst Evol Microbiol 48, 1383-1387.
Hattori S, Kamagata Y, Hanada S & Shoun H (2000) Thermacetogenium phaeum gen. nov., sp.
nov., a strictly anaerobic, thermophilic, syntrophic acetate-oxidizing bacterium. Int J Syst
Evol Microbiol 50, 1601-1609.
Henrissat B (1991) A classification of glycosyl hydrolases based on amino-acid sequence
similarities. Biochem J 280, 309-316.
Hill MO (1973) Diversity and Evenness - unifying notation and its consequences. Ecology 54,
427-432.
Hu ZH, Wang G & Yu HQ (2004) Anaerobic degradation of cellulose by rumen microorganisms
at various pH values. Biochem Eng J 21, 59-62.
Huse SM, Huber JA, Morrison HG, Sogin ML & Welch DM (2007) Accuracy and quality of
massively parallel DNA pyrosequencing. Genome Biol 8, R143.
138
Illumina Inc, (2012) HiSeq® 1500/2500 sequencing systems. URL: http://www.illumina.com/
Documents/products/datasheets/datasheet_hiseq2500.pdf. Last access November 23,
2012.
Imachi H, Sekiguchi Y, Kamagata Y, Hanada S, Ohashi A & Harada H (2002) Pelotomaculum
thermopropionicum gen. nov., sp. nov., an anaerobic, thermophilic, syntrophic propionate-
oxidizing bacterium. Int J Syst Evol Microbiol 52, 1729-1735.
Ishii S, Kosaka T, Hori K, Hotta Y & Watanabe K (2005) Coaggregation facilitates interspecies
hydrogen transfer between Pelotomaculum thermopropionicum and Methanothermobacter
thermautotrophicus. Appl Environ Microbiol 71, 7838-7845.
Iwasaki M, Yamashiro T, Beneragama N, Nishida T, Kida K, Ihara I, Takahashi J & Umetsu K
(2011) The effect of temperature on survival of pathogenic bacteria in biogas plants. Anim
Sci J 82, 707-712.
Jabari L, Gannoun H, Cayol JL, Hamdi M, Fauque G, Ollivier B & Fardeau ML (2012)
Characterization of Defluviitalea saccharophila gen. nov., sp. nov., a thermophilic
bacterium isolated from an upflow anaerobic filter treating abattoir wastewaters, and
proposal of Defluviitaleaceae fam. nov. Int J Syst Evol Microbiol 62, 550-555.
Jaenicke S, Ander C, Bekel T et al. (2011) Comparative and joint analysis of two metagenomic
datasets from a biogas fermenter obtained by 454-pyrosequencing. PLoS ONE 6, e14519.
Jayani RS, Saxena S & Gupta R (2005) Microbial pectinolytic enzymes: A review. Process
Biochem 40, 2931-2944.
Jetten MSM, Stams AJM & Zehnder AJB (1992) Methanogenesis from acetate - A comparison
of the acetate metabolism in Methanothrix soehngenii and Methanosarcina spp.. FEMS
Microbiol Rev 88, 181-198.
Johansen C, Falholt P & Gram L (1997) Enzymatic removal and disinfection of bacterial
biofilms. Appl Environ Microbiol 63, 3724-3728.
Kaemmerer D, Büttner P, Seigner L & Friedrich R (2008) Phytopathogene im Biogasprozess -
Phytopathogens in biogas plants. In Mitteilungen aus dem Julius Kühn-Institut (JKI) 417,
pp. 239-240. Quedlinburg, Germany: JKI.
Kampmann K, Ratering S, Kramer I, Schmidt M, Zerr W & Schnell S (2012) Unexpected stability
of Bacteroidetes and Firmicutes communities in laboratory biogas reactors fed with
different defined substrates. Appl Environ Microbiol 78, 2106-2119.
Kaplan CW & Kitts CL (2003) Variation between observed and true terminal restriction fragment
length is dependent on true TRF length and purine content. J Microbiol Methods 54,
121-125.
Karakashev D, Batstone DJ & Angelidaki I (2005) Influence of environmental conditions on
methanogenic compositions in anaerobic biogas reactors. Appl Environ Microbiol 71,
331-338.
Karakashev D, Batstone DJ, Trably E & Angelidaki I (2006) Acetate oxidation is the dominant
methanogenic pathway from acetate in the absence of Methanosaetaceae. Appl Environ
Microbiol 72, 5138-5141.
Karam NE & Belarbi A (1995) Detection of polygalacturonases and pectin esterases in lactic-
acid bacteria. World J Microbiol Biotechnol 11, 559-563.
Karita S, Nakayama K, Goto M, Sakka K, Kim WJ & Ogawa S (2003) A novel cellulolytic,
anaerobic, and thermophilic bacterium, Moorella sp. strain F21. Biosci Biotechnol Biochem
67, 183-185.
139
Kastner V, Somitsch W & Schnitzhofer W (2012) The anaerobic fermentation of food waste: a
comparison of two bioreactor systems. J Clean Prod 34, 82-90.
Khan MM, Pyle BH & Camper AK (2010) Specific and rapid enumeration of viable but
nonculturable and viable-culturable gram-negative bacteria by using flow cytometry. Appl
Environ Microbiol 76, 5088-5096.
Kim OS, Imhoff JF, Witzel KP & Junier P (2011) Distribution of denitrifying bacterial
communities in the stratified water column and sediment-water interface in two freshwater
lakes and the Baltic Sea. Aquat Ecol 45, 99-112.
Kitts CL (2001) Terminal restriction fragment patterns: a tool for comparing microbial
communities and assessing community dynamics. Curr Issues Intest Microbiol 2, 17-25.
Klocke M, Mähnert P, Mundt K, Souidi K & Linke B (2007) Microbial community analysis of a
biogas-producing completely stirred tank reactor fed continuously with fodder beet silage
as mono-substrate. Syst Appl Microbiol 30, 139-151.
Klocke M, Nettmann E, Bergmann I, Mundt K, Souidi K, Mumme J & Linke B (2008)
Characterization of the methanogenic Archaea within two-phase biogas reactor systems
operated with plant biomass. Syst Appl Microbiol 31, 190-205.
Kongjan P, Thong S & Angelidaki I (2011) Performance and microbial community analysis of
two-stage process with extreme thermophilic hydrogen and thermophilic methane
production from hydrolysate in UASB reactors. Bioresour Technol 102, 4028-4035.
Koster IW & Lettinga G (1984) The influence of ammonium-nitrogen on the specific activity of
pelletized methanogenic sludge. Agricultural Wastes 9, 205-216.
Kouguchi H, Watanabe T, Sagane Y, Sunagawa H & Ohyama T (2002) In vitro reconstitution of
the Clostridium botulinum type D progenitor toxin. J Biol Chem 277, 2650-2656.
Krakat N, Westphal A, Schmidt S & Scherer P (2010a) Anaerobic digestion of renewable
biomass: thermophilic temperature governs methanogen population dynamics. Appl
Environ Microbiol 76, 1842-1850.
Krakat N, Westphal A, Satke K, Schmidt S & Scherer P (2010b) The microcosm of a biogas
fermenter: Comparison of moderate hyperthermophilic (60 °C) with thermophilic (55 °C)
conditions. Eng Life Sci 10, 520-527.
Krause L, Diaz NN, Edwards RA et al. (2008a) Taxonomic composition and gene content of a
methane-producing microbial community isolated from a biogas reactor. J Biotechnol 136,
91-101.
Krause L, Diaz NN, Goesmann A, Kelley S, Nattkemper TW, Rohwer F, Edwards RA & Stoye J
(2008b) Phylogenetic classification of short environmental DNA fragments. Nucleic Acids
Res 36, 2230-2239.
Krieg NR, Staley JT, Brown DR, Hedlund BP, Paster BJ, Ward NL, Ludwig W & Whitman WB
eds. (2010) Volume Four: The Bacteroidetes, Spirochaetes, Tenericutes (Mollicutes),
Acidobacteria, Fibrobacteres, Fusobacteria, Dictyoglomi, Gemmatimonadetes,
Lentisphaerae, Verrucomicrobia, Chlamydiae, and Planctomycetes. In BERGEY'S
MANUAL® OF Systematic Bacteriology ed. Whitman, WB, pp. 1-948. New York, USA:
Springer.
Kröber M, Bekel T, Diaz NN et al. (2009) Phylogenetic characterization of a biogas plant
microbial community integrating clone library 16S-rDNA sequences and metagenome
sequence data obtained by 454-pyrosequencing. J Biotechnol 142, 38-49.
Krsek M & Wellington EMH (1999) Comparison of different methods for the isolation and
purification of total community DNA from soil. J Microbiol Methods 39, 1-16.
140
Krumrei C (2010) Mikrobiologie thermophiler Biogasfermentationen im zweistufigen Verfahren.
Diploma thesis. Berlin, Germany: Beuth Hochschule für Technik Berlin.
Kruskal JB (1964) Multidimensional scaling by optimizing goodness of fit to a nonmetric
hypothesis. Psychometrika 29, 1-27.
Kruskal JB & Wish M (1978) Multidimensional scaling. Beverly Hills, USA: Sage Publications.
Lamed R, Naimark J, Morgenstern E & Bayer EA (1987) Specialized cell-surface structures in
cellulolytic bacteria. J Bacteriol 169, 3792-3800.
Lane DJ (1991) 16S/23S rRNA sequencing. In Nucleic Acid Techniques in Bacterial
Systematics ed. Stackebrandt, E and Goodfellow, M, pp. 115-147. New York, USA: Wiley.
Lee MJ & Zinder SH (1988) Isolation and characterization of a thermophilic bacterium which
oxidizes acetate in syntrophic association with a methanogen and which grows
acetogenically on H2-CO2. Appl Environ Microbiol 54, 124-129.
Lehtomäki A, Huttunen S, Lehtinen T & Rintala JA (2008) Anaerobic digestion of grass silage in
batch leach bed processes for methane production. Bioresour Technol 99, 3267-3278.
Leininger S, Urich T, Schloter M, Schwark L, Qi J, Nicol GW, Prosser JI, Schuster SC &
Schleper C (2006) Archaea predominate among ammonia-oxidizing prokaryotes in soils.
Nature 442, 806-809.
Lerm S, Kleyboecker A, Miethling-Graff R, Alawi M, Kasina M, Liebrich M & Wuerdemann H
(2012) Archaeal community composition affects the function of anaerobic co-digesters in
response to organic overload. Waste Manag 32, 389-399.
Leyer I & Wesche K (2008) Multidimensionale Skalierung. In Multivariate Statistik in der
Ökologie, pp. 137-152. Berlin Heidelberg, Germany: Springer.
Lien T, Madsen M, Rainey FA & Birkeland NK (1998) Petrotoga mobilis sp. nov., from a North
Sea oil-production well. Int J Syst Bacteriol 48, 1007-1013.
Liu WT, Marsh TL, Cheng H & Forney LJ (1997) Characterization of microbial diversity by
determining terminal restriction fragment length polymorphisms of genes encoding 16S
rRNA. Appl Environ Microbiol 63, 4516-4522.
Liu FH, Wang SB, Zhang JS, Zhang J, Yan X, Zhou HK, Zhao GP & Zhou ZH (2009) The
structure of the bacterial and archaeal community in a biogas digester as revealed by
denaturing gradient gel electrophoresis and 16S rDNA sequencing analysis. J Appl
Microbiol 106, 952-966.
Lloyd M & Ghelardi RJ (1964) A table for calculating the "equitability" component of species
diversity. J Anim Ecology 33, 217-225.
Loy A, Maixner F, Wagner M & Horn M (2007) ProbeBase - an online resource for rRNA-
targeted oligonucleotide probes: new features 2007. Nucleic Acids Res 35, D800-D804.
Lueders T & Friedrich M (2000) Archaeal population dynamics during sequential reduction
processes in rice field soil. Appl Environ Microbiol 66, 2732-2742.
Lueders T & Friedrich MW (2003) Evaluation of PCR amplification bias by terminal restriction
fragment length polymorphism analysis of small-subunit rRNA and mcrA genes by using
defined template mixtures of methanogenic pure cultures and soil DNA extracts. Appl
Environ Microbiol 69, 320-326.
Lukow T, Dunfield PF & Liesack W (2000) Use of the T-RFLP technique to assess spatial and
temporal changes in the bacterial community structure within an agricultural soil planted
with transgenic and non-transgenic potato plants. FEMS Microbiol Ecol 32, 241-247.
141
Lynd LR, Weimer PJ, van Zyl WH & Pretorius IS (2002) Microbial cellulose utilization:
Fundamentals and biotechnology. Microbiol Mol Biol Rev 66, 506-577.
Magurran AE (1988) Ecological diversity and its measurement. Princeton, USA: Princeton
University Press.
Margulies M, Egholm M, Altman WE et al. (2005) Genome sequencing in open microfabricated
high density picoliter reactors. Nature 437, 376-380.
Marsh TL (1999) Terminal restriction fragment length polymorphism (T-RFLP): an emerging
method for characterizing diversity among homologous populations of amplification
products. Curr Opin Microbiol 2, 323-327.
Marsh TL (2005) Culture-independent microbial community analysis with terminal restriction
fragment length polymorphism. Environ Microbiol 397, 308-329.
Masai E, Katayama Y & Fukuda M (2007) Genetic and biochemical investigations on bacterial
catabolic pathways for lignin-derived aromatic compounds. Biosci Biotechnol Biochem 71,
1-15.
Maune MW & Tanner RS (2012) Description of Anaerobaculum hydrogeniformans sp. nov., an
anaerobe that produces hydrogen from glucose, and emended description of the genus
Anaerobaculum. Int J Syst Evol Microbiol 62, 832-838.
McDonald JE, Houghton JNI, Rooks DJ, Allison HE & McCarthy AJ (2012) The microbial
ecology of anaerobic cellulose degradation in municipal waste landfill sites: evidence of a
role for Fibrobacters. Environ Microbiol 14, 1077-1087.
McEniry J, O'Kiely P, Clipson NJW, Forristal PD & Doyle EM (2008) Bacterial community
dynamics during the ensilage of wilted grass. J Appl Microbiol 105, 359-371.
McInerney MJ, Bryant MP, Hespell RB & Costerton JW (1981) Syntrophomonas wolfei gen.
nov., sp. nov., an anaerobic, syntrophic, fatty-acid oxidizing bacterium. Appl Environ
Microbiol 41, 1029-1039.
McKay AM (1990) Degradation of polygalacturonic acid by Saccharomyces cerevisiae. Lett Appl
Microbiol 11, 41-44.
Mende DR, Waller AS, Sunagawa S, Jaervelin AI, Chan MM, Arumugam M, Raes J & Bork P
(2012) Assessment of metagenomic assembly using simulated next generation sequencing
data. Plos One 7, e31386.
Menes RJ & Muxi L (2002) Anaerobaculum mobile sp. nov., a novel anaerobic, moderately
thermophilic, peptide fermenting bacterium that uses crotonate as an electron acceptor,
and emended description of the genus Anaerobaculum. Int J Syst Evol Microbiol 52, 157-
164.
Merlino G, Rizzi A, Villa F, Sorlini C, Brambilla M, Navarotto P, Bertazzoni B, Zagni M, Araldi F
& Daffonchio D (2012) Shifts of microbial community structure during anaerobic digestion
of agro-industrial energetic crops and food industry byproducts. J Chem Technol
Biotechnol 87, 1302-1311.
Mladenovska Z & Ahring B (2000) Growth kinetics of thermophilic Methanosarcina spp. isolated
from full-scale biogas plants treating animal manures. FEMS Microbiol Ecol 31, 225-229.
Moeseneder MM, Arrieta JM, Muyzer G, Winter C & Herndl GJ (1999) Optimization of terminal-
restriction fragment length polymorphism analysis for complex marine bacterioplankton
communities and comparison with denaturing gradient gel electrophoresis. Appl Environ
Microbiol 65, 3518-3525.
Mott IEC, Stickler DJ, Coakley WT & Bott TR (1998) The removal of bacterial biofilm from water-
filled tubes using axially propagated ultrasound. J Appl Microbiol 84, 509-514.
142
Muha I, Grillo A, Heisig M, Schönberg M, Linke B & Wittum G (2011) Mathematical modeling of
process liquid flow and acetoclastic methanogenesis under mesophilic conditions in a two-
phase biogas reactor. Bioresour Technol 106, 1-9.
Muyzer G, Teske A, Wirsen CO & Jannasch HW (1995) Phylogenetic relationships of
Thiomicrospira species and their identification in deep-sea hydrothermal vent samples by
denaturing gradient gel-electrophoresis of 16S rDNA fragments. Arch Microbiol 164, 165-
172.
Nettmann E, Bergmann I, Mundt K, Linke B & Klocke M (2008) Archaea diversity within a
commercial biogas plant utilizing herbal biomass determined by 16S rDNA and mcrA
analysis. J Appl Microbiol 105, 1835-1850.
Nettmann E, Bergmann I, Pramschüfer S, Mundt K, Plogsties V, Herrmann C & Klocke M
(2010) Polyphasic analyses of methanogenic archaeal communities in agricultural biogas
plants. Appl Environ Microbiol 76, 2540-2548.
Neumann L & Scherer P (2011) Impact of bioaugmentation by compost on the performance and
ecology of an anaerobic digester fed with energy crops. Bioresour Technol 102, 2931-
2935.
Nielsen HB, Mladenovska Z & Ahring BK (2007) Bioaugmentation of a two-stage thermophilic
(68 °C/55 °C) anaerobic digestion concept for improvement of the methane yield from
cattle manure. Biotechnol Bioeng 97, 1638-1643.
Nizami A, Thamsiriroj T, Singh A & Murphy J (2010) Role of leaching and hydrolysis in a two-
phase grass digestion system. Energy Fuels 24, 4549-4559.
Nizami AS & Murphy JD (2011) Optimizing the operation of a two-phase anaerobic digestion
system digesting grass silage. Environ Sci Technol 45, 7561-7569.
Nizami AS, Singh A & Murphy JD (2011) Design, commissioning, and start-up of a sequentially
fed leach bed reactor complete with an upflow anaerobic sludge blanket digesting grass
silage. Energy Fuels 25, 823-834.
Nolte C (2011) Entwicklung der mikrobiellen Lebensgemeinschaften in thermophilen
Biogasfermentationen. Bachelor thesis. Berlin, Germany: Humboldt-Universität zu Berlin.
O’Neill MA, Albersheim P & Darvill A (1990) The pectic polysaccharides of primary cell walls. In
Methods in Plant Biochemistry ed. Dey, PM, pp. 415–441. London, UK: Academic Press.
Ogata H, Goto S, Sato K, Fujibuchi W, Bono H & Kanehisa M (1999) KEGG: Kyoto
Encyclopedia of Genes and Genomes. Nucleic Acids Res 27, 29-34.
Opatkiewicz AD, Butterfield DA & Baross JA (2009) Individual hydrothermal vents at Axial
Seamount harbor distinct subseafloor microbial communities. FEMS Microbiol Ecol 70,
413-424.
Osborn AM, Moore ERB & Timmis KN (2000) An evaluation of terminal-restriction fragment
length polymorphism (T-RFLP) analysis for the study of microbial community structure and
dynamics. Environ Microbiol 2, 39-50.
Osborne CA, Rees GN, Bernstein Y & Janssen PH (2006) New threshold and confidence
estimates for terminal restriction fragment length polymorphism analysis of complex
bacterial communities. Appl Environ Microbiol 72, 1270-1278.
Pacific Biosciences (2012) SMRT sequencing advantage. URL: http://www.pacific
biosciences.com/products/smrt-technology/smrt-sequencing-advantage/. Last access
November 23, 2012.
143
Pagani I, Liolios K, Jansson J, Chen IMA, Smirnova T, Nosrat B, Markowitz VM & Kyrpides NC
(2012) The Genomes OnLine Database (GOLD) v.4: status of genomic and metagenomic
projects and their associated metadata. Nucleic Acids Res 40, D571-D579.
Patel BKC, Morgan HW & Daniel RM (1985) Fervidobacterium nodosum gen. nov. and spec.
nov., a new chemoorganotrophic, caldoactive, anaerobic bacterium. Arch Microbiol 141,
63-69.
Pereira MG, Latchford JW & Mudge SM (2006) The use of terminal restriction fragment length
polymorphism (T-RFLP) for the characterisation of microbial communities in marine
sediments. Geomicrobiol J 23, 247-251.
Pereyra LP, Hiibel SR, Riquelme MVP, Reardon KF & Pruden A (2010) Detection and
quantification of functional genes of cellulose-degrading, fermentative, and sulfate-reducing
bacteria and methanogenic archaea. Appl Environ Microbiol 76, 2192-2202.
Pernthaler A, Pernthaler J & Amann R (2002) Fluorescence in situ hybridization and catalyzed
reporter deposition for the identification of marine bacteria. Appl Environ Microbiol 68,
3094-3101.
Ponpium P, Ratanakhanokchai K & Kyu KL (2000) Isolation and properties of a cellulosome-
type multienzyme complex of the thermophilic Bacteroides sp. strain P-1. Enzyme Microb
Technol 26, 459-465.
Punta M, Coggill PC, Eberhardt RY et al. (2012) The Pfam protein families database. Nucleic
Acids Res 40, D290-D301.
Pycke B, Etchebehere C, Van de Caveye P, Negroni A, Verstraete W & Boon N (2011) A time-
course analysis of four full-scale anaerobic digesters in relation to the dynamics of change
of their microbial communities. Water Sci Technol 63, 769-775.
Rademacher A, Zakrzewski M, Schlüter A, Schönberg M, Szczepanowski R, Goesmann A,
Pühler A & Klocke M (2012) Characterization of microbial biofilms in a thermophilic biogas
system by high-throughput metagenome sequencing. FEMS Microbiol Ecol 79, 785-799.
Raizada N (2004) Application of molecular-biological method for optimization of anaerobic
reactors. In Berichte aus Wassergüte- und Abfallwirtschaft. Garching, Germany:
Technische Universität München.
Reddy CA & D’Souza TM (1994) Physiology and molecular-biology of the lignin peroxidases of
Phanerochaete chrysosporium. FEMS Microbiol Rev 13, 137-152.
Rees GN, Patel BKC, Grassia GS & Sheehy AJ (1997) Anaerobaculum themoterrenum gen.
nov. sp. nov., a novel, thermophilic bacterium which ferments citrate. Int J Syst Bacteriol
47, 150-154.
Rheims H & Stackebrandt E (1999) Application of nested polymerase chain reaction for the
detection of as yet uncultured organisms of the class Actinobacteria in environmental
samples. Environ Microbiol 1, 137-143.
Rochex A, Masse A, Escudie R, Godon JJ & Bernet N (2009) Influence of abrasion on biofilm
detachment: evidence for stratification of the biofilm. J Ind Microbiol Biot 36, 467-470.
Rozen S & Skaletsky HJ (2000) Primer3 on the WWW for general users and for biologist
programmers. In Bioinfomatics: Methods and Protocols ed. Misener, S and Krawetz, SA,
pp. 365-386. Totowa, USA: Humana Press.
Sait L, Galic M, Strugnell RA & Janssen PH (2003) Secretory antibodies do not affect the
composition of the bacterial microbiota in the terminal ileum of 10-week-old mice. Appl
Environ Microbiol 69, 2100-2109.
144
Sakai S, Imachi H, Hanada S, Ohashi A, Harada H & Kamagata Y (2008) Methanocella
paludicola gen. nov., sp. nov., a methane-producing archaeon, the first isolate of the
lineage ‘Rice Cluster I’, and proposal of the new archaeal order Methanocellales ord. nov.
Int J Syst Evol Micr 58, 929-936.
Sanger F, Nicklen S & Coulson AR (1977) DNA sequencing with chain-terminating inhibitors.
Proc Natl Acad Sci USA 74, 5463-5467.
Schlüter A, Bekel T, Diaz NN et al. (2008) The metagenome of a biogas-producing microbial
community of a production-scale biogas plant fermenter analysed by the 454-
pyrosequencing technology. J Biotechnol 136, 77-90.
Schnürer A, Schink B & Svensson BH (1996) Clostridium ultunense sp. nov., a mesophilic
bacterium oxidizing acetate in syntrophic association with a hydrogenotrophic
methanogenic bacterium. Int J Syst Bacteriol 46, 1145-1152.
Scholten JCM & Conrad R (2000) Energetics of syntrophic propionate oxidation in defined batch
and chemostat cocultures. Appl Environ Microbiol 66, 2934-2942.
Schönberg M & Linke B (2012) The influence of the temperature regime on the formation of
methane in a two-phase anaerobic digestion process. Eng Life Sci 12, 279-286.
Schülein M (2000) Protein engineering of cellulases. Biochim Biophys Acta 1543, 239-252.
Schütte UME, Abdo Z, Bent SJ, Shyu C, Williams CJ, Pierson JD & Forney LJ (2008) Advances
in the use of terminal restriction fragment length polymorphism (T-RFLP) analysis of 16S
rRNA genes to characterize microbial communities. Appl Microbiol Biotechnol 80, 365-380.
Schwarz WH (2001) The cellulosome and cellulose degradation by anaerobic bacteria. Appl
Microbiol Biotechnol 56, 634-649.
Sekiguchi Y, Kamagata Y, Nakamura K, Ohashi A & Harada H (2000) Syntrophothermus
lipocalidus gen. nov., sp. nov., a novel thermophilic, syntrophic, fatty-acid-oxidizing
anaerobe which utilizes isobutyrate. Int J Syst Evol Microbiol 50, 771-779.
Shannon CE & Weaver W (1963) The mathematical theory of communication. Urbana, USA:
University of Illinous press.
Sharkey FH, Banat IM & Marchant R (2004) Detection and quantification of gene expression in
environmental bacteriology. Appl Environ Microbiol 70, 3795-3806.
Shoham Y, Lamed R & Bayer EA (1999) The cellulosome concept as an efficient microbial
strategy for the degradation of insoluble polysaccharides. Trends Microbiol 7, 275-281.
Simpson EH (1949) Measurement of diversity. Nature 163, 688-688.
Singh BK, Nazaries L, Munro S, Anderson IC & Campbell CD (2006) Use of multiplex terminal
restriction fragment length polymorphism for rapid and simultaneous analysis of different
components of the soil microbial community. Appl Environ Microbiol 72, 7278-7285.
Sipos R, Szekely AJ, Palatinszky M, Revesz S, Marialigeti K & Nikolausz M (2007) Effect of
primer mismatch, annealing temperature and PCR cycle number on 16S rRNA gene-
targetting bacterial community analysis. FEMS Microbiol Ecol 60, 341-350.
Siriwongrungson V, Zeng RJ & Angelidaki I (2007) Homoacetogenesis as the alternative
pathway for H2 sink during thermophilic anaerobic degradation of butyrate under
suppressed methanogenesis. Water Res 41, 4204-4210.
Sleat R, Mah RA & Robinson R (1984) Isolation and characterization of an anaerobic,
cellulolytic bacterium, Clostridium cellulovorans sp. nov. Appl Environ Microbiol 48, 88-93.
145
Smith CJ, Danilowicz BS, Clear AK, Costello FJ, Wilson B & Meijer WG (2005) T-Align, a web-
based tool for comparison of multiple terminal restriction fragment length polymorphism
profiles. FEMS Microbiol Ecol 54, 375-380.
Soo RM, Wood SA, Grzymski JJ, McDonald IR & Cary SC (2009) Microbial biodiversity of
thermophilic communities in hot mineral soils of Tramway Ridge, Mount Erebus, Antarctica.
Environ Microbiol 11, 715-728.
Speel EJM, Hopman AHN & Komminoth P (1999) Amplification methods to increase the
sensitivity of in situ hybridization: Play CARD(S). J Histochem Cytochem 47, 281-288.
Stahl DA & Amann R (1991) Development and application of nucleic acid probes. In Nucleic
acid techniques in bacterial systematics ed. Stackebrandt, E, pp. 205-248. Chichester,
England: John Wiley & Sons Ltd.
Staley JT & Konopka A (1985) Measurement of in situ activities of nonphotosynthetic
microorganisms in aquatic and terrestrial habitats. Ann Rev Microbiol 39, 321-346.
Statistisches Bundesamt (2012) Landwirtschaftlich genutzte Fläche nach Hauptnutzungsarten.
https://www.destatis.de/DE/ZahlenFakten/Wirtschaftsbereiche/LandForstwirtschaft/Bodenn
utzung/Tabellen/HauptnutzungsartenLF.html. Last access December 03, 2012.
Steinberg LM & Regan JM (2009) McrA-targeted real-time quantitative PCR method to examine
methanogen communities. Appl Environ Microbiol 75, 4435-4442.
Stoecker K, Dorninger C, Daims H & Wagner M (2010) Double labeling of oligonucleotide
probes for fluorescence in situ hybridization (DOPE-FISH) improves signal intensity and
increases rRNA accessibility. Appl Environ Microbiol 76, 922-926.
Sun L, Ouyang X, Tang Y, Yang Y & Luo Y (2012) Effects of different methods of DNA
extraction for activated sludge on the subsequent analysis of bacterial community profiles.
Water Environ Res 84, 108-114.
Takai K, Moser DP, Onstott TC, Spoelstra N, Pfiffner SM, Dohnalkova A & Fredrickson JK
(2001) Alkaliphilus transvaalensis gen. nov., sp. nov., an extremely alkaliphilic bacterium
isolated from a deep South African gold mine. Int J Syst Evol Microbiol 51, 1245-1256.
Talbot G, Roy CS, Topp E, Kalmokoff ML, Brooks SPJ, Beaulieu C, Palin MF & Masse DI
(2010) Spatial distribution of some microbial trophic groups in a plug-flow-type anaerobic
bioreactor treating swine manure. Water Sci Technol 61, 1147-1155.
Tale VP, Maki JS, Struble CA & Zitomer DH (2011) Methanogen community structure-activity
relationship and bioaugmentation of overloaded anaerobic digesters. Water Res 45, 5249-
5256.
Tamura K, Peterson D, Peterson N, Stecher G, Nei M & Kumar S (2011) MEGA5: Molecular
evolutionary genetics analysis using maximum likelihood, evolutionary distance, and
maximum parsimony methods. Mol Biol Evol 28, 2731-2739.
Tan KS, Wee BY & Song KP (2001) Evidence for holin function of tcdE gene in the
pathogenicity of Clostridium difficile. J Med Microbiol 50, 613-619.
Thauer RK, Hedderich R & Fischer R (1993) Reactions and enzymes involved in
methanogenesis from CO2 and H2. In Methanogenesis ed. Ferry, JG, pp. 209-252. New
York London: Chapman & Hall.
Tsuge H, Nagahama M, Nishimura H, Hisatsune J, Sakaguch Y, Itogawa Y, Katunuma N &
Sakurai J (2003) Crystal structure and site-directed mutagenesis of enzymatic components
from Clostridium perfringens iota-toxin. J Mol Biol 325, 471-483.
146
Turnbaugh PJ, Ley RE, Mahowald MA, Magrini V, Mardis ER & Gordon JI (2006) An obesity-
associated gut microbiome with increased capacity for energy harvest. Nature 444, 1027-
1031.
Ulrich A, Klimke G & Wirth S (2008) Diversity and activity of cellulose-decomposing bacteria,
isolated from a sandy and a loamy soil after long-term manure application. Microb Ecol 55,
512-522.
v. Wintzingerode F, Göbel UB & Stackebrandt E (1997) Determination of microbial diversity in
environmental samples: pitfalls of PCR-based rRNA analysis. FEMS Microbiol Rev 21,
213-229.
Van de Peer Y, Chapelle S & De Wachter R (1996) A quantitative map of nucleotide substitution
rates in bacterial rRNA. Nucleic Acids Res 24, 3381-3391.
Vanysacker L, Declerck SAJ, Hellemans B, De Meester L, Vankelecom I & Declerck P (2010)
Bacterial community analysis of activated sludge: an evaluation of four commonly used
DNA extraction methods. Appl Microbiol Biotechnol 88, 299-307.
Vetriani C, Jannasch HW, MacGregor BJ, Stahl DA & Reysenbach AL (1999) Population
structure and phylogenetic characterization of marine benthic Archaea in deep-sea
sediments. Appl Environ Microbiol 65, 4375-4384.
Vinneras B, Schönning C & Nordin A (2006) Identification of the microbiological community in
biogas systems and evaluation of microbial risks from gas usage. Sci Total Environ 367,
606-615.
Wagner M, Amann R, Lemmer H & Schleifer KH (1993) Probing activated sludge with
oligonucleotides specific for Proteobacteria - Inadequacy of culture-dependent methods for
describing microbial community structure. Appl Environ Microbiol 59, 1520-1525.
Wallner G, Amann R & Beisker W (1993) Optimizing fluorescent in situ hybridization with
ribosomal-RNA-targeted oligonucleotide probes for flow cytometric identification of
microorganisms. Cytometry 14, 136-143.
Wang Q, Garrity GM, Tiedje JM & Cole JR (2007) Naive Bayesian Classifier for rapid
assignment of rRNA sequences into the new bacterial taxonomy. Appl Environ Microbiol
73, 5261-5267.
Wang H, Vuorela M, Keränen AL, Lehtinen TM, Lensu A, Lehtomäki A & Rintala J (2010)
Development of microbial populations in the anaerobic hydrolysis of grass silage for
methane production. FEMS Microbiol Ecol 72, 496-506.
Wang TY, Chen HL, Lu MYJ et al. (2011) Functional characterization of cellulases identified
from the cow rumen fungus Neocallimastix patriciarum W5 by transcriptomic and
secretomic analyses. Biotechnol Biofuels 4, 24.
Warnecke F, Luginbuehl P, Ivanova N et al. (2007) Metagenomic and functional analysis of
hindgut microbiota of a wood-feeding higher termite. Nature 450, 560-565.
Weiland P (2010) Biogas production: current state and perspectives. Appl Microbiol Biotechnol
85, 849-860.
Weisburg WG, Barns SM, Pelletier DA & Lane DJ (1991) 16S ribosomal DNA amplification for
phylogenetic study. J Bacteriol 173, 697-703.
Westerholm M, Roos S & Schnürer A (2010) Syntrophaceticus schinkii gen. nov., sp. nov., an
anaerobic, syntrophic acetate-oxidizing bacterium isolated from a mesophilic anaerobic
filter. FEMS Microbiol Lett 309, 100-104.
147
Westerholm M, Roos S & Schnürer A (2011) Tepidanaerobacter acetatoxydans sp. nov., an
anaerobic, syntrophic acetate-oxidizing bacterium isolated from two ammonium-enriched
mesophilic methanogenic processes. Syst Appl Microbiol 34, 260-266.
Wheeler DL, Barrett T, Benson DA et al. (2007) Database resources of the National Center for
Biotechnology Information. Nucleic Acids Res 35, D5-D12.
Whittaker MM, Kersten PJ, Cullen D & Whittaker JW (1999) Identification of catalytic residues in
glyoxal oxidase by targeted mutagenesis. J Biol Chem 274, 36226-36232.
WHO - World health organization (2011) Ausbrüche von E. coli O104:H4-Infektionen:
Lagebericht 30. URL: http://www.euro.who.int/de/where-we-work/member-states/
germany/sections/news/2011/07/outbreaks-of-e.-coli-o104h4-infection-update-30. Last
access December 10, 2012.
Wiegel J (1992) The anaerobic thermophilic Bacteria. In Thermophilic Bacteria ed. Kristjansson,
JK, pp. 105-184. Boca Raton, USA: CRC Press.
Willms M, Hufnagel J & Wagner B (2008) Nutzen und Risiken des Energiepflanzenanbaus für
den Boden. In Kongressband 2008 Jena - Workshop "Bioenergie" des 120. VDLUFA-
Kongresses pp. 90-99. Darmstadt, Germany: VDLUFA-Verlag.
Wirth R, Kovacs E, Maroti G, Bagi Z, Rakhely G & Kovacs KL (2012) Characterization of a
biogas-producing microbial community by short-read next generation DNA sequencing.
Biotechnol Biofuels 5, 41.
Wittmann J, Eichenlaub R & Dreiseikelmann B (2010) The endolysins of bacteriophages CMP1
and CN77 are specific for the lysis of Clavibacter michiganensis strains. Microbiology SGM
156, 2366-2373.
Wu WM, Thiele JH, Jain MK & Zeikus JG (1993) Metabolic properties and kinetics of
methanogenic granules. Appl Microbiol Biotechnol 39, 804-811.
Xu Q, Gao WC, Ding SY, Kenig R, Shoham Y, Bayer EA & Lamed R (2003) The cellulosome
system of Acetivibrio cellulolyticus includes a novel type of adaptor protein and a cell
surface anchoring protein. J Bacteriol 185, 4548-4557.
Yu Y, Lee C, Kim J & Hwang S (2005) Group-specific primer and probe sets to detect
methanogenic communities using quantitative real-time polymerase chain reaction.
Biotechnol Bioeng 89, 670-679.
Yu Y, Kim J & Hwang S (2006) Use of real-time PCR for group-specific quantification of
aceticlastic methanogens in anaerobic processes: Population dynamics and community
structures. Biotechnol Bioeng 93, 424-433.
Zakrzewski M, Goesmann A, Jaenicke S, Jünemann S, Eikmeyer F, Szczepanowski R, Abu Al-
Soud W, Sorensen S, Pühler A & Schlüter A (2012) Profiling of the metabolically active
community from a production-scale biogas plant by means of high-throughput
metatranscriptome sequencing. J Biotechnol 158, 248-258.
Zakrzewski M, Bekel T, Ander C, Pühler A, Rupp O, Stoye J, Schlüter A & Goesmann A (2013)
MetaSAMS - A novel software platform for taxonomic classification, functional annotation
and comparative analysis of metagenome datasets. J Biotechnol 167, 156-165.
Zhang QJ, Meitzler JC, Huang SX & Morishita T (2000) Sequence polymorphism, predicted
secondary structures, and surface-exposed conformational epitopes of Campylobacter
major outer membrane protein. Infect Immun 68, 5679-5689.
Zhang T & Fang HHP (2006) Applications of real-time polymerase chain reaction for
quantification of microorganisms in environmental samples. Appl Microbiol Biotechnol 70,
281-289.
148
Zheng DD, Alm EW, Stahl DA & Raskin L (1996) Characterization of universal small-subunit
rRNA hybridization probes for quantitative molecular microbial ecology studies. Appl
Environ Microbiol 62, 4504-4513.
Zhu CG, Zhang JY, Tang YP, Xu ZK & Song RT (2011) Diversity of methanogenic archaea in a
biogas reactor fed with swine feces as the mono-substrate by mcrA analysis. Microbiol Res
166, 27-35.
Ziganshin AM, Schmidt T, Scholwin F, Il'inskaya ON, Harms H & Kleinsteuber S (2011) Bacteria
and Archaea involved in anaerobic digestion of distillers grains with solubles. Appl
Microbiol Biotechnol 89, 2039-2052.
Zinder SH & Koch M (1984) Non-aceticlastic methanogenesis from acetate: acetate oxidation
by a thermophilic syntrophic coculture. Arch Microbiol 138, 263-272.
Zinder SH (1990) Conversion of acetic-acid to methane by thermophiles. FEMS Microbiol Rev
75, 125-137.
149
7 APPENDIX
7.1 NMDSforTRFLPcommunityprofiles
Figure 7.1 NMDS of the bacterial
TRFLP profiles obtained fo
r
samples of the 21-day fermentation
at LBR temperatures of 55 °C (A),
60 °C (B), 65 °C (C), 70 °C (D) and
75 °C (E). Stress values: 0.017 (A),
0.053 (B), 0.100 (C), 0.110 (D) and
0.129 (E); L 0-21, leachate samples
at different time points of the
fermentation; D 21, digestate
sample at the end of the
fermentation
150
7.2 Proteinfamiliesrelevantforthemetagenomicanalyses
Table 7.1 Accession numbers of protein families relevant for carbohydrate
degradation (A) and methanogenesis (B) obtained from the Pfam
database
(A) Carbohydrate degrading enzymes (B) Methanogenic enzymes
Pfam
accession Protein family Pfam
accession Protein family
PF00150 Cellulase (glycosyl hydrolase
family 5)
PF00871 Acetokinase family
PF00295 Glycosyl hydrolases family 28 PF02552 CO dehydrogenase beta subunit/acetyl-CoA
synthase ε-subunit
PF00331 Glycosyl hydrolase family 10 PF03598 CO dehydrogenase/acetyl-CoA synthase complex
β-subunit
PF00457 Glycosyl hydrolases family 11 PF03599 CO dehydrogenase/acetyl-CoA synthase δ-subunit
PF00544 Pectate lyase PF02289 Cyclohydrolase (MCH)
PF00553 Cellulose binding domain PF01913 Formylmethanofuran tetrahyromethanopterin
formyltransferase, lobe
PF00703 Glycosyl hydrolases family 2,
domain
PF02741 Formylmethanofuran tetrahyromethanopterin
formyltransferase, proximal lobe
PF00759 Glycosyl hydrolase family 9 PF01493 GXGXG motif
PF00840 Glycosyl hydrolase family 7 PF03201 H2-forming N5,N10-methylene-
tetrahydromethanopterin dehydrogenase
PF00942 Cellulose binding domain PF04609 Methyl-coenzyme M reductase operon protein C
PF01055 Glycosyl hydrolases family 31 PF02505 Methyl-coenzyme M reductase operon protein D
PF01095 Pectinesterase PF02249 Methyl-coenzyme M reductase α-subunit,
C-terminal domain
PF01270 Glycosyl hydrolases family 8 PF02745 Methyl-coenzyme M reductase α-subunit,
N-terminal domain
PF01915 Glycosyl hydrolase family 3 C
terminal domain
PF02241 Methyl-coenzyme M reductase β-subunit,
C-terminal domain
PF02011 Glycosyl hydrolase family 48 PF02783 Methyl-coenzyme M reductase β-subunit,
N-terminal domain
PF02056 Family 4 glycosyl hydrolase PF02240 Methyl-coenzyme M reductase γ-subunit
PF02156 Glycosyl hydrolase family 26 PF01993 Methylene-5,6,7,8-tetrahydromethanopterin
dehydrogenase
PF02255 PTS system, Lactose/
cellobiose specific IIA subunit
PF02663 Molybdenum formylmethanofuran dehydrogenase
operon
PF02449 Beta-galactosidase PF01515 Phosphate acetyl/butaryl transferase
PF02836 Glycosyl hydrolases family 2,
TIM barrel domain
PF04208 Tetrahydromethanopterin S-methyltransferase,
subunit A
PF02837 Glycosyl hydrolases family 2,
sugar binding domain
PF05440 Tetrahydromethanopterin S-methyltransferase,
subunit B
PF02929 Beta galactosidase small
chain
PF04211 Tetrahydromethanopterin S-methyltransferase,
subunit C
PF03422 Carbohydrate binding module PF04207 Tetrahydromethanopterin S-methyltransferase,
subunit D
PF04616 Glycosyl hydrolases family 43 PF04206 Tetrahydromethanopterin S-methyltransferase,
subunit E
PF06165 Glycosyltransferase family 36 PF09472 Tetrahydromethanopterin S-methyltransferase,
subunit F
PF08532 Beta-galactosidase
trimerisation domain
PF04210 Tetrahydromethanopterin S-methyltransferase,
subunit G
PF08533 Beta-galactosidase C-terminal
domain
PF02007 Tetrahydromethanopterin S-methyltransferase,
subunit H
151
Table 7.2 Accession numbers of protein families relevant for pathogenicity of
selected pathogens obtained from the Pfam database
Pfam
accession Protein family
PF00161 Ribosome inactivating protein
PF01375 Heat-labile enterotoxin alpha chain
PF01376 Heat-labile enterotoxin beta chain
PF01742 Clostridial neurotoxin zinc protease
PF02048 Heat-stable enterotoxin
PF02258 Shiga-like toxin beta subunit family
PF03278 IpaB/EvcA family
PF03318 Clostridium epsilon toxin ETX/Bacillus mosquitocidal toxin MTX2
PF03495 Clostridial binary toxin B/anthrax toxin PA
PF03496 ADP-ribosyltransferase exoenzyme
PF03505 Clostridium enterotoxins
PF05058 ActA Protein
PF05105 Holin family
PF05538 Campylobacter major outer membrane protein
PF05588 Clostridium botulinum HA-17 protein
PF06002 Alpha-2,3-sialyltransferase (CST-I)
PF06109 Haemolysin E (HlyE)
PF06306 Beta-1,4-N-acetylgalactosaminyltransferase (CgtA)
PF07906 ShET2 enterotoxin, N-terminal region
PF07951 Clostridium neurotoxin, C-terminal receptor binding
PF07952 Clostridium neurotoxin, Translocation domain
PF07953 Clostridium neurotoxin, N-terminal receptor binding
PF08090 Heat stable E. coli enterotoxin 1
PF08191 Leucine-rich repeats (LLR) adjacent
PF08470 Nontoxic nonhaemagglutinin C-terminal
PF09052 Salmonella invasion protein A
PF09599 Salmonella-Shigella invasin protein C
PF12354 Bacterial adhesion/invasion protein N terminal
PF12918 TcdB toxin N-terminal helical domain
PF12919 TcdA/TcdB catalytic glycosyltransferase domain
PF12920 TcdA/TcdB pore forming domain
