activated zeolite—suitable carriers for microorganisms in anaerobic digestion processes?

BIOENERGYAND BIOFUELS

Activated zeolite—suitable carriers for microorganismsin anaerobic digestion processes?

S. Weiß & M. Lebuhn & D. Andrade & A. Zankel &M. Cardinale & R. Birner-Gruenberger & W. Somitsch &

B. J. Ueberbacher & G. M. Guebitz

Received: 12 July 2012 /Revised: 2 January 2013 /Accepted: 3 January 2013 /Published online: 23 February 2013# Springer-Verlag Berlin Heidelberg 2013

Abstract Plant cell wall structures represent a barrier in thebiodegradation process to produce biogas for combustionand energy production. Consequently, approachesconcerning a more efficient de-polymerisation of celluloseand hemicellulose to monomeric sugars are required. Here,we show that natural activated zeolites (i.e. trace metalactivated zeolites) represent eminently suitable mineralmicrohabitats and potential carriers for immobilisation ofmicroorganisms responsible for anaerobic hydrolysis of bio-polymers stabilising related bacterial and methanogeniccommunities. A strategy for comprehensive analysis ofimmobilised anaerobic populations was developed thatincludes the visualisation of biofilm formation via scan-ning electron microscopy and confocal laser scanning

microscopy, community and fingerprint analysis as wellas enzyme activity and identification analyses. UsingSDS polyacrylamide gel electrophoresis, hydrolytical ac-tive protein bands were traced by congo red staining.Liquid chromatography/mass spectroscopy revealed cellulo-lytical endo- and exoglucanase (exocellobiohydrolase) as wellas hemicellulolytical xylanase/mannase after proteolytic di-gestion. Relations to hydrolytic/fermentative zeolite colonis-ers were obtained by using single-strand conformationpolymorphism analysis (SSCP) based on amplification ofbacterial and archaeal 16S rRNA fragments. Thereby, domi-nant colonisers were affiliated to the genera Clostridium,Pseudomonas and Methanoculleus. The specific immobilisa-tion on natural zeolites with functional microbes already

Electronic supplementary material The online version of this article(doi:10.1007/s00253-013-4691-6) contains supplementary material,which is available to authorized users.

S. Weiß :M. Cardinale :G. M. GuebitzInstitute of Environmental Biotechnology,Graz University of Technology, Petersgasse 12,8010 Graz, Austria

M. Lebuhn :D. AndradeBavarian State Research Centre for Agriculture,Vöttinger Straße 38,85354 Freising, Germany

A. ZankelInstitute for Electron Microscopy, Graz University of Technology,Steyrergasse 17,8010 Graz, Austria

R. Birner-GruenbergerProteomics Core Facility, Center for Medical Researchand Institute of Pathology, Medical University of Graz,Stiftingtalstrasse 24,8010 Graz, Austria

W. SomitschEngineering Consultant, Wiedner Hauptstrasse 90/2/19,1050 Vienna, Austria

B. J. UeberbacherIPUS Mineral- & Umwelttechnologie GmbH, Werksgasse 281,8786 Rottenmann, Austria

S. Weiß :G. M. Guebitz (*)ACIB Austrian Centre of Industrial Biotechnology, Petersgasse 14,8010 Graz, Austriae-mail: [email protected]

Present Address:G. M. GuebitzDepartment of Agrobiotechnology Tulln, Institute ofEnvironmental Biotechnology, University of Natural Resourcesand Life Science, Konrad Lorenz Str. 20,3430 Tulln, Austria

Appl Microbiol Biotechnol (2013) 97:3225–3238DOI 10.1007/s00253-013-4691-6

http://dx.doi.org/10.1007/s00253-013-4691-6

colonising naturally during the fermentation offers a strategyto systematically supply the biogas formation process respon-sive to population dynamics and process requirements.

Keywords Biogas . Zeolites . Hemicellulases . Cellulases .

Microbial community . Grass silage

Introduction

In order to stabilise and optimise biogas formation fromenergy plants several approaches were investigated to avoidunfavourable acidification, ammonia accumulation etc. de-creasing the production of methane. Adequate availability ofessential trace elements for the bacterial community wasidentified to be essential (Takashima and Speece 1989;Goodwin et al. 1990). Pobeheim et al. (2010a) showed thatthe addition of mixed trace elements increased methaneyields by up to 30 % in batch digestion of maize silage.Furthermore, effective removal of volatile solids (around73 %) has been reported for maize silage operated complete-ly stirred tank reactors (CSTR), when higher concentratedsolutions of trace metals were added (Bougrier et al. 2011).Enhancement of methane production (up to 20 %), was alsoobserved for Co and Ni addition which was corroborated byrisen cofactor F420 activities and dominance of hydrogeno-trophic Methanoculleus sp. (Pobeheim et al. 2010b). Theimportance of Co, Se and other trace metals was also ob-served for stable food waste digestion. Here, the accumula-tion of volatile fatty acids (VFAs) can only be avoided,when digesters are supplemented with a certain concentra-tion of Se, Mo, Co and W (Banks et al. 2012).

Activated zeolites as used in this study combine tracemetal activation on the surface with an high ion-exchangecapacity for supplementation with cationic macro-elementssuch as Na+, K+, Mg2+ or Ca2+ and trace elements as well,that zeolites can be loaded with (Holper et al. 2005).Consequently, these materials potentially provide an excel-lent operational environment for microorganisms participat-ing either in the biomass hydrolysis or methane formationprocess. Recent studies demonstrated the capability of thesezeolites to be colonised by certain microbial populationsduring anaerobic batch-wise operation and laboratory-scaled mono-grass silage or vinasses operated bioreactors(Weiß et al. 2011; Fernández et al. 2007). Furthermore, itwas shown that microorganisms acting on biomass asdegraders, obtained by selective cultivation, can be immo-bilised on zeolite surfaces. These populations can provide aspecific enzymatic activity as addition to the natural consor-tium activities, e.g. hydrolytic activity to increase recalci-trant biomass degradation (plant cell walls) and thus resultin higher methane yields in batch-culture experiments asdemonstrated before (Weiß et al. 2010). This effect could

not be seen with enzymatic treatment alone as reported byBruni et al. (2010) applying commercially available laccasesor mixture of cellulases and hemicellulases to digested bio-fibres. Consequently, there is strong need to identify zeoliteimmobilised organisms and their enzymes in order to exploitthis unique system to improve biogas formation. Here, weshow that a combination of sophisticated microscopic [con-focal laser scanning microscopy (CLSM)-fluorescence insitu hybridization (FISH), scanning electron microscopy(SEM)], genetic [(single-strand conformation polymor-phism analysis (SSCP)] and biochemical [liquid chromatog-raphy/mass spectroscopy (LC-MS) protein identification]methods can lead to a comprehensive mechanistic under-standing of the function of zeolite immobilised anaerobicpopulations.

Materials and methods

Semi-continuous fermentation and analogue batch operation

The activated zeolites used (IPUS GmbH, Rottenmann,Austria) consisted of a natural zeolitic tuff containing >85 %clinoptilolite, which was crushed to a grain size below2.5 mm. The material was loaded with Fe, Ni, Co, Mo,Se, Cu and Zn as trace metal elements to enhancemicrobial activity (Holper et al. 2005). Powderous par-ticles of <100-μm size provided by IPUS were introduced tobatch experiments (5.0 gl−1) using in situ concentrate bags(6.75×12 cm; Ankom, Macedon, NY, USA) to apply an insacco incubation during batch operation. Activated zeoliteparticles of 1.0- to 2.5-mm size were entrapped in 0.5-mmmesh size polyamide bags in six fractions of 0.4 g in order toapply an in sacco incubation during single-stage flow-throughmono-digestion of grass silage in bioreactors of 28-l operatingvolume (Lebuhn et al. 2008). Organic loading rates rangedfrom 1 to 2 kg ODM m−3day−1 after slow adjustment of thetemperature (1 °C day −1) to 45 °C. Zeolite sample fractionswere collected after 7, 14, 21, 42 and 84 days.

To investigate biofilm formation in more detail, a batch-operation was carried out in 1,000-ml ground flasks with atotal volume of 500 ml of seeding sludge, which wasobtained from fermenters long-term mono-digesting grasssilage (360 days) at increasing organic loading rates of 0.5–3.0 kg ODM m−3day−1 and mesophilic conditions (Wichernet al. 2009). Before use, the sludge was flushed withoxygen-free nitrogen gas for 20 minl−1 to obtain anaerobicconditions. Two different substrates were used: (1) grasssilage (DM content of 56.13 %), grained to fibrouswool with a blender and (2) a soluble model substrateresembling natural grass silage (Weiß et al. 2011). Re-activation of collected zeolites after batch-wise operationfor 6 weeks and long-term storage at 4 °C was then

3226 Appl Microbiol Biotechnol (2013) 97:3225–3238

carried out in mineral salt medium, which had the followingcomposition (in milligramme per litre): MgSO4·7H2O, 9.4;CaSO4·2H2O, 4.7; Na2HPO4·2H2O, 752; KH2PO4, 63.92;NH4CL, 18.8. After mixing, 0.47-ml trace element solutionwas added and pH adjusted to 7.2 by using 1 M HCl. Traceelement solution was composed of (in milligramme per litre):Na2EDTA·2H2O, 2,740; ZnSO4·7H2O, 100; MnCl2·4H2O,25.6; H3BO3, 300; CoCl2·6H2O, 200; CuCl2·2H2O, 10;NiCl2·2H2O, 20; Na2MO4·2H2O, 900; Na2SeO3·5H2O, 30.4;FeSO4·7H2O, 1,000. In order to induce production of hydro-lytic enzymes, the following model substrate was added fed-batch-wise to the mineral salt medium in a concentration of1%DM (w/v): microcrystalline cellulose (Fluka), 30%; xylanfrom Birchwood (Roth, Karlsruhe, Germany), 50 %; ligninwith low sulfonate content (Sigma-Aldrich, St. Louis, MO,USA), 14.5 %; pectin C (Roth), 0.5 % and starch (Roth), 5 %.The re-activation experiment was carried out over a totalperiod of 55 days at temperatures according to the formerbatch-fermentation experiments, i.e. 35 and 45 °C respective-ly. Samples were taken as duplicates at regular intervals.

PCR-based SSCP and sequencing analysis

The total bacterial community DNAwas extracted by usinga DNA extraction kit for soil samples (FastDNA® Spin Kitfor Soil, MP Biomedicals, Solon, OH, USA) following themanufacturer’s recommendations. To collect zeolite-immobilised microorganisms, 0.1 g of zeolite was rinsedthree times with 500 μl of 1× PBS (pH8.0) before applyingthe DNA extraction procedure. Free microorganisms werecollected from sample supernatants (1.0 ml) representing thetotal sludge inoculation phase. Samples were centrifuged for15 min at 16,750×g to collect the cells. Amplification ofbacterial 16S rRNA gene fragments was carried out usingthe bacterial primer pair Com-1 (5′-CAG CAG CCG CGGTAA TAC-3′) (Schwieger and Tebbe 1998) and Unibac-II-927rP (5′-CCC GTC AAT TYM TTT GAG TT-3′) for anamplicon size of 412 bp according to Lieber et al. (2002).Amplification of archaeal 16S rRNA gene fragments wasperformed using the primer pair ARC-787f (5′-ATTAGATACC CSBGT AGTCC-3′) and ARC-1059r (5′-GCCATGCACC WCCTC T-3′) for an amplicon size of 273 bpfollowing Yu et al. (2005), using a Biometra T personal/gradient system (Biometra, Göttingen, Germany). After pu-rification of PCR-generated DNA products using the GeneClean Turbo Kit (Qbiogene, Heidelberg, Germany) follow-ing the manufacturer’s recommendations, DNA-content wasdetermined before SSCP analysis. Therefore, 1 μl of puri-fied PCR-products were analysed using a micro-volumespectrophotometer for nucleic acid and protein quantitation(NanoDrop 2000, Thermo Fisher Scientific, Wilmington,DE, USA). SSCP analysis of amplified bacterial and archae-al rrs (16S rRNA gene) fragments was carried out according

to Schwieger and Tebbe (1998). For single-strand forma-tion, 10 μl of purified PCR products containing 150 ng ofdsDNA were used to generate comparable bacterial andarchaeal fingerprints after λ-exonuclease digestion.Individual bands of fingerprint patterns were excised forDNA elution through several incubation steps in sterile elutionbuffer pH8.0 (5 M ammonium acetate, 10 mM magnesiumacetate, 1 mM Na2EDTA, 0.1 % SDS): (1) 15 min at −70 °C,(2) thawed for 5 min at 50 °C, (3) heated for 5 min at 90 °C,(4) incubated for 3 h at 37 °C and (5) stored at −20 °C for atleast 3 days before DNA precipitation using ethanol (70 %)was carried out. Similar PA-gel band positions were combinedin order to increase DNA amounts for re-amplification usingthe same primer pairs as for the initial PCR. PCR productswere then used for sequencing analysis after adjusting the totalDNA concentration to 2 ngμl−1. Sequencing analyses wereperformed by Eurofins MWG Operon (Ebersberg, Germany).To identify similar sequences that are available in the NCBIGenbank, sequences were used in BLASTn searches(Altschul et al. 1997; http://www.ncbi.nlm.nih.gov/blast/).Database hits with minimum 95 % sequence identity and acoverage cut off of 90 % were considered for discussion.Accession numbers for primary nucleotide sequence datadetermined during 16S rRNA gene fragment-based commu-nity analyses were submitted to the NCBI GenBank database(reference code: BankIt1555395) representing band numbers1 to 17 from SSCP-generated fingerprint patterns: (1)JX436456, (3) JX436455, (4) JX436459, (7) JX436458, (8)JX436457, (9) JX436454 and (17) JX436453. According tothe guidelines of the database, SSCP sequence lengths werenot less than 150 bp, and quality of the data was granted byclipping.

Enzyme activity assay

Enzyme activity, i.e. cellulase, pectinase and xylanase weremeasured using a 3-amino,5-nitrosalicylic acid (DNS) assaywhich is based on the quantification of reducing sugarsreleased with DNS according to Bailey et al. (1992).Samples were collected at two different sample sites: (1)liquid phase, for determination of extracellular enzymes asample volume of 1 ml was centrifuged (13,000×g for5 min) and supernatants used; (2) pellet, for determinationof cell-bound enzymes of zeolite-immobilised microorgan-isms a sample volume of 1 ml was centrifuged, the pelletcollected and two times rinsed with 1x PBS, before it wasresuspended in 500 μl of 1× PBS (pH8.0). The resuspensionwas achieved by gently using the pipet tip (a) or treating thepellet with glass beads (0.15 to 2.00 μm) for 40 s. using aFastPrep Instrument (Qbiogene, Heidelberg, Germany) toensure complete detachment of zeolite-immobilised bacteriacells (b). The different sample fractions were then incubatedat 50 °C for 10 min in the presence of a substrate solution

Appl Microbiol Biotechnol (2013) 97:3225–3238 3227

http://www.ncbi.nlm.nih.gov/blast/

containing 1 % (w/v) carboxymethylcellulose (Fluka, St.Gallen, Switzerland), polygalacturonic acid (Fluka) or xylanfrom birchwood (Roth) in an adjusted pH range of 6.0 to 6.5by using potassium phosphate buffer. The photometricabsorption was determined at 530/540 nm in 96-wellplates (Greiner, Frickenhausen, Germany) using a platereader (Tecan Infinite 200 M, Männedorf, Switzerland).Enzyme activity is given as units of enzyme activity perlitre (Ul−1). An enzyme unit is defined as conversion of1 μmol reduced sugar (D-glucose, or D-xylose) per minuteunder the given conditions above. One unit corresponds to16.67 nkat.

SDS-PA-Gel analysis

For protein separation SDS-PAGE was performed in 10 %polyacrylamide slab gels according to Laemmli (1970). Theseparating gel composition (5 ml) contained: 1.25 ml acryl-amide (40 %), 1.25 ml Tris (1.5 M, pH8.8), 50 μl sodiumdodecyl sulphate (SDS; 10 %), 25 μl ammonium persulfate(APS; 10 %), 6 μlN,N,N′,N′-tetramethylethylenediamine(TEMED), 2.4 ml dH2O. The stacking gel (5 ml) wascomposed of: 0.5 ml acrylamide (40 %), 1.25 ml Tris(0.5 M, pH6.8), 50 μl SDS (10 %), 25 μl APS, 6 μlTEMED, 3.2 ml dH2O. For activity staining, polymericsubstrate, i.e. carboxymethylcellulose (Fluka, St. Gallen,Switzerland), lichenan (Megazyme, Wicklow, Ireland) orxylan (Roth, Karlsruhe, Germany) was incorporated intothe separating gel in a concentration of 0.5 % (w/v) priorto the addition of APS according to Zverlov et al. (2010).Protein samples were mixed with loading dye (1:1) contain-ing 2.2 % SDS and applied directly to the polymerised gelafter determination of the protein concentration followingLowry et al. (1951) using the Roti®-Nanoquant proteinquantitation assay (Roth, Karlsruhe, Germany). The runningbuffer contained the following components: 1 g SDS, 3 gTris, 14 g glycine (Roth). Electrophoresis was performed onice at a conduction current of 30 mA (∼60 V) until theloading dye front reached the bottom of the slab gel (∼2–3 h). Gels were then put between cellulosic filter papers(Schleicher and Schuell, Dassel, Germany) soaked with0.1 M succinate buffer (pH4.5) and incubated for 120 minat 50 °C for renaturation. Protein bands were optionallydetected by staining with Coomassie blue R-250 followingKang et al. (2002). After destaining with 10 % ethanolsolution containing 2 % (v/v) o-phosphoric acid, Congored (Sigma-Aldrich, St. Louis, MO, USA) was appliedto detect catalytically active bands by staining the poly-meric gel matrix. Therefore gels were submersed inCongo red solution (0.02 %) for 15 min at room tem-perature (∼20 °C). After destaining with 1 M NaCl,halo zones in the red-stained background revealed thedegradation of the matrix incorporated biopolymer,

which were therefore chosen for further characterisationvia LC-MS/MS

LC-MS/MS analysis

Protein identification and internal sequence information wasreceived from LC-MS/MS. Protein bands were excised fromgels and reduced, alkylated and digested with Promega-modified trypsin according to the method of Shevchenkoet al. (1996). Digests were separated by nano-HPLC(Agilent 1200 system, Vienna, Austria) equipped with aZorbax 300SB-C18 enrichment column (5 μm, 5×0.3 mm)and a Zorbax 300SB-C18 nanocolumn (3.5 μm, 150×0.075 mm). Forty microlitres of sample were injected andconcentrated on the enrichment column for 6 min using0.1 % formic acid as isocratic solvent at a flow rate of 20 μlmin−1. The column was then switched in the nanoflow circuit,and the sample was loaded on the nanocolumn at a flow rate of300 nlmin−1. Separation was carried out using the followinggradient, where solvent A is 0.3 % formic acid in water andsolvent B is a mixture of acetonitrile and water (4:1) contain-ing 0.3 % formic acid: 0–6 min, 13 % B; 6–63 min, 13–28 %B; 63–88 min, 28–50 % B; 88–89 min, 50–100 % B; 89–100 min, 100 % B; 100–101 min, 100–13 % B; 101–120 min,re-equilibration at 13 % B. The sample was ionized in thenanospray source equipped with nanospray tips (PicoTipTMStock# FS360-75-15-D-20, Coating: 1P–4P, 15+/− 1 μmEmitter, New Objective, Woburn, MA, USA). It was analyzedin a Thermo LTQ-FT mass spectrometer (Thermo FisherScientific, Waltham, MA, USA) operated in positive ionmode, applying alternating full scan MS (m/z 400 to 2,000)in the ion cyclotron and MS/MS by collision induced disso-ciation of the five most intense peaks in the ion trap withdynamic exclusion enabled. The LC-MS/MS data were ana-lyzed by searching the NCBI public database with Mascot 2.2(MatrixScience, London, UK). A maximum false discoveryrate of 5% using decoy database search, an ion score cut off of20 and a minimum of two identified peptides was chosen asidentification criteria.

Scanning electron microscopy

For the investigation of morphological colonisation charac-teristics of activated zeolite particles, scanning electron mi-croscopy was used (Zeiss ULTRA 55, Carl Zeiss MicroImaging GmbH, Germany). Particles of 1.0 to 2.5 mm indiameter were analysed. The biological material was imagedwith the high-efficiency “in-lens SE detector” using second-ary electrons (SE), which deliver topographic contrast(Goldstein et al. 2003). Additionally, imaging with an angleselective backscattered electron (AsB) detector was per-formed. It detects backscattered electrons (BSE) and deliv-ers compositional contrast, which is determined by the


differences in the local chemical composition of an investi-gated specimen (Goldstein et al. 2003). In the case of SEimaging, 5-kV acceleration voltage of the primary electronswas applied with the exception of one micrograph whichwas imaged simultaneously with the AsB detector at anacceleration voltage of 15 kV. Before application, freshsamples were fixed in glutaraldehyde (2.5 % in 0.1 Mphosphate puffer, pH7.2), dehydrated in a graded ethanolseries with a final step in propylene oxide and ultimately driedby lyophilisation (Labconco Freeze Dry System FreeZone4.5, Kansas City, MO, USA) or critical point drying (CO2;Bal-Tec CPD, Balzers, Fürstentum Liechtenstein). For obser-vation, prepared samples were mounted on aluminum stubsusing double-sided carbon tape and moreover a carbon filmwas applied onto the specimens’ surfaces by high-vacuumevaporation.

Confocal laser-scanning microscopy and fluorescencein situ hybridization

To investigate taxonomic groups embedded in biofilms ongrass silage fibres and activated zeolite particles <100 μm indiameter, in situ hybridization using fluorescence-labellednucleic acid probes that are complementary to target signa-ture regions of rRNA was done. Before fixation, sampleswere washed three times with 1× phosphate-buffered saline(PBS), approximately 500 μl for 0.3 g fresh matter. Forfixation 800 μl of a 4 % paraformaldehyde/1× PBS solution(4:1) were mounted for an incubation time of 12 h at 4 °Cand then discarded. Then three washing steps in ice-cold 1×PBS (rinse/5 min/10 min) followed, before the storage at −20 °C in 500 μl of a 96 % ethanol/1× PBS solution (1:1). Afteranother washing step with 200 μl of 1×PBS, 10 μl of 1 mgml−1 lysozyme (Sigma-Aldrich, St. Louis, MO, USA) wereadded for permeabilisation over 10 min at room temperatureon polysine slides (Menzel-Gläser, Braunschweig, Germany).The samples were then rinsed twice in ice-cold 1× PBS beforean ethanolic dehydration series (50–80–96 %) was applied.After quickly rinsingwith PBS, another washing step for 3 minat room temperature followed before hybridizations were per-formed. The hybridization buffer contained 0.9 M NaCl,0.02 M Tris–HCl (pH8.0), 0.01 % (w/v) SDS and 2.5–5.0 ngμl−1 per probe. The concentration of formamide(Invitrogen, Karlsruhe, Germany) was 10 % (w/v) for everyprobe. Probes used: EUB338Mix (5′-GCT GCC TCC CGTAGGAGT-3′, 5′-GCAGCCACCCGTAGGTGT-3′, 5′-GCTGCC ACC CGT AGG TGT-3′) adapted from Amann et al.(1990) and Daims et al. (1999) for most bacteria; ARCH915(5′-GTG CTCCCCCGCCAATTCCT-3′) described by Stahland Amann (1996) for Archaea and Univ-1390 (5′-GAC GGGCGGTGTGTACAA-3′) for counter-staining of all organisms(Zheng et al. 1996). All hybridizations were realised by incu-bation at 36 °C in the dark for 90–180 min. To eliminate the

hybridization buffer afterwards, samples were rinsed with40 μl of pre-warmed (38 °C) washing buffer before a furtherincubation step for 10–15 min at 38 °C. The washing buffercontained: 0.45 M NaCl and 0.02 M Tris–HCl. Finally, sam-ples were rinsed with 500 μl ice-cold double-distilled H2O andmounted with Molecular Probes® ProLong Gold antifadentwith or without DAPI (Invitrogen) over 30 min to 24 h at roomtemperature as recommended by the manufacturer. Samplestained with multiple probes were simultaneously observedusing a Leica TCS SPE confocal laser-scanning microscope(LeicaMicrosystems, Heidelberg, Germany), taking advantageof the non-overlapping emission wavelengths of the fluoro-chromes used (maximum excitation/emission in (nanometre):FITC 490/525; Cy3 548/562; Cy5 650/670).

Results

Community profile and characterisation of populationson zeolites

In order to visualize anaerobic hydrolytic populationsimmobilized on zeolite surfaces, SEM pictures were takenafter critical point drying (Fig. 1a) or lyophilisation(Fig. 1b–d). Extensive colonisation during batch-wise oper-ation was observed area-wide after 5 days of incubation at45 °C and 21 days when incubated at 35 °C, respectively(Fig. 1a, b). Microbial cells observed ranged from 0.3 to0.5 μm in width and 1.2 to 5.4 μm in length. In contrast,cells observed after in sacco incubation in semi-continuously operated bioreactors were filamentous andpronounced rod shaped with lengths of 5.5 to 7.0 μm(Fig. 1c, d). Pilus-like appendages (Fig. 1b) were observedto reach out from microbial cells directly either to thezeolite’s surface or to neighbouring cells. Pili-dimensionsranged from 230 nm to 3 μm in length and 45 to 77 nm inwidth. The organic origin of bacillary and fibrous structureswas confirmed by using In-lens and AsB detectors simulta-neously (Fig. 1c, d) to expose topographic and composition-al contrasts that allow the discrimination between carbon-based biological material (i.e. microbial cells) and inorganicmatter.

Rapid degradation of grass silage fibres was observedafter 7 days of incubation at 35 °C as determined byCLSM. As shown in Fig. 2a–f, bacteria were present ongrass fibres, forming degradation sites on leaves andstems. Approximately 4.02±7.79×1011 cells cm−3 basedon hybridizing probes specific for bacteria (yellow fluores-cence signals) were counted. Epidermis layers were affectedfirst, at which decomposition was beginning from the primarycell wall proceeding to the outer cuticula. Apart from grasssilage fibres, bacteria were also detected on activated zeoliteparticles (<100 μm in diameter) present in batch-digestion


reactors of a model substrate for grass silage (Fig. 3c, d). Here,bacteria were seen both on the surface and deep inside the

clefts of zeolites. Bacterial colonisation was less pronouncedwhen compared to grass silage, however, still approximately

Fig. 1 SEM images ofactivated zeolite particles (<2.5 mm in diameter) colonisedby anaerobic microbialpopulations during batch-wise(a, b) and in sacco incubation insemi-continuously operatedbioreactors (c, d). a, b (SE, in-lens detector) Activated zeoliteheavily colonised by microor-ganisms (biofilm formation). c,d Rod-shaped cells, vanishingwhen inorganic matter is fo-cused (AsB detector). Barlength: 1 μm

Fig. 2 a, c, e confocal laser scanning micrographs of grass silagefibres with fluorescence labelled bacteria after 7 days of fermentationat 35 °C. EUB338-Cy3 probe mix was used for the specific staining ofbacteria. Negative controls with non-sense FISH probe (NON-EUB)

showed no positive signals. b, d, f 3D reconstructions (formed by spotsand iso-surfaces) of images a, c and e. Red spots represent bacteriadegrading outer layers of grass leave cell walls (yellow/beige). Barlength: 5 μm


7.22±3.92×1010 cells cm−3 were counted. This finding con-firms zeolites to be quite suitable for microbial colonisationduring anaerobic batch-wise operation.

Fingerprints of microbial populations colonising zeolitesurfaces in large-scale bioreactors and batch-operation wereperformed by PCR-based SSCP analyses. Figure 4 showsfingerprints of the archaeal community (A) and bacterialcommunity (B) associated with activated zeolites duringfermentation of grass silage and the model substrate insemi-continuously operated mode and batch mode, respec-tively. Regarding bacterial community analysis, total sludgeof semi-continuously operated bioreactors was analysed tocompare fingerprints of activated zeolites and liquid phaseassociated populations. Bacterial fingerprints of activatedzeolite surfaces and total sludge show comparable patterns;however, certain bands (B: 10, 11) appeared exclusively inzeolite surface samples from semi-continuously operatedbioreactor I3 and batch-operation on grass silage. Mostdifferences in band patterns between samples from semi-continuous fermentation and batch-operation were observedin the lower part of the gel and therefore most interesting foridentification analyses. Regarding archaeal populations, dif-ferences between bioreactor I2 and I3 were observed. Whilebands 1 to 3 (A) were dominant in bioreactor I2, they werealmost vanishing in reactor I3. In return, bands 7 and 8 (A)intensified from I2 to I3. Both bioreactors had pH values

ranging from 8.0 to 8.6 during sampling period, but theinitial volatile organic acids to total inorganic carbon ratiowas much higher for I3 (3.63) compared to I2 (0.94) indi-cating an accumulation of volatile fatty acids (VFA). Highcontents of VFA in I3 were affecting the sampling periodfrom day 7 to 21 and thus presumably changed the archaealcommunity composition.

Clustering the fingerprint patterns observed during SSCPanalysis revealed three distinct clusters regarding the bacte-rial community (Fig. 5, I–III). One cluster (I) was specificfor grass silage fed batch-operation with relatively lowsimilarity of <58 % when both incubation temperatures areincluded. Nevertheless, patterns derived from incubation at45 °C alone showed a high homogeneity (similarity of 96 %).Interestingly, cluster I was not only different from the onefound for model substrate fed batch-cultures, but also frompatterns derived from bioreactors semi-continuously operatedwith grass silage at 45 °C. Replicates of model substrategrown populations formed the second cluster (II) with asimilarity value of >73 %, unravelling two minor clusters foreach incubation temperature with similarities of >80 %.Highest band pattern similarity value (>88 %) was found forfingerprints from semi-continuous digestion of grass silageforming a third major cluster, which revealed no significantdifferences between zeolite surface and total sludge originatedpatterns. Populations in the liquid phase and on the surface

Fig. 3 a Bright field total viewof washed zeolite particles<100 μm in diameter; b confocallaser scanning micrograph of azeolite particle colonised byanaerobic bacteria. EUB338-Cy3 probe mix was used for thespecific staining of bacteria.Negative controls with non-sense FISH probe (NON-EUB)showed no positive signals. c to e3D reconstructions (formed byspots and iso-surfaces) of imageb. Green spots represent bacteriacolonising the particle’s surfaceand inside as well. Bar length,100 μm (a); 5 μm (b–e)


were probably foremost equally abundant, but within themajor cluster, two more distinct ones were found for activatedzeolite surfaces from bioreactor I2 (after in sacco incubationfrom days 7 to 21) and I3 (days 21 to 84) with simi-larities of 88 and 90 % respectively. Sequence analysesof zeolite surface samples collected from semi-continuously and batch-operated fermentation of grasssilage and model substrate indicated affiliations to two bacte-rial orders, i.e. Pseudomonadales and Clostridiales, and onearchaeal order, i.e. Methanomicrobiales. Within the familyPseudomonadaceae, affiliations to cellulases and pectinasesproducing Pseudomonas fluorescens (Pseudomonas cellu-losa) and Pseudomonas putida were indicated, when colon-ised zeolites from batch fermentation of model substrate at35 °C were analysed (Fig. 4, band spot 17). Correspondingbands were also observed when samples from semi-continuously operated fermentation approaches at 45 °C wereapplied to the SSCP analysis. In contrast, affiliations to thefamily Clostridiaceae were only present in samples frombatch and semi-continuously operated fermentations of grasssilage (Fig. 4, band spot 9). Within the archaeal domain,affiliations to the order Methanomicrobiales, i.e. thegenus Methanoculleus spp. was indicated during thetotal observation period of semi-continuously operated

fermentation of grass silage analysing colonised zeolites.Data on the community composition of zeolite-associatedmicroorganisms (archaea and bacteria) are summarised in-cluding indications to affiliations below the family level(Table 1).

Role of zeolite-associated populations in biomassdegradation

In order to analyse the role of microorganisms colonisingactivated zeolites on biomass degradation, hydrolytic en-zyme activities were determined. Therefore, washed colon-ised activated zeolites were re-incubated in buffer onsubstrate according to their pre-incubation, i.e. model sub-strate or grass silage. Then, cellulase and xylanase activitieswere measured after certain time points throughout the re-incubation (Fig. 6). Highest hydrolytic enzyme activities(2.8 kUl−1) were observed for colonized zeolite samplestaken from model substrate fermentations after 8 days ofre-incubation at 35 °C. When increased temperature (45 °C)was applied, an individual maximum was reached (1.9 kUl−1). Re-incubation in the presence of grass silage at 35 °Cled to 2.2 kUl−1 on average. As for model substrate, in-creased incubation temperature decreased hydrolytical

Fig. 4 Non-denaturating polyacrylamide gel from the SSCP analysisbased on (a) archaeal rrs (16S rDNA gene) fragments and (b) bacterialrrs (16S rDNA gene) fragments of (a) semi-continuously operatedmono-grass silage fermentation exposed zeolites (IS2/3),corresponding total sludge samples (I1–3) and of (b) batch-cultures

fed with grass silage (GS) or model substrate (MS). (M: 1 kb DNA-ladder marker lanes; sample-identity: IS2/3=bioreactors (in sacco sam-ple); MS/GS=batch-substrate, 35/45=fermentation temperature in (indegree Celsius), 7–353=fermentation time in (in day)


enzyme activities (1.8 kUl−1 at day27). The overall maxi-mum was found for re-incubation on model substrate (35 °C) at day27 with 3.2 kUl−1 on average. Assessing xylanaseand cellulase activities individually, highest xylanase activ-ities (around 1.5 to 1.6 kUl−1) were found for both sub-strates (at day8 for model substrate and day27 for grasssilage) at 45 °C incubation temperature. In contrast, maxi-mum cellulase activities occurred during incubation at 35 °Cranging from 1.3 kUl−1 upon grass silage to 1.9 kUl−1 uponmodel substrate at days8 and 27, respectively. As a result,cellulase activities pushed the maximum value of totalhydrolytical enzyme activity towards the lower incubationtemperature condition. During incubation, pH was adjustedto 6.0 to 6.5 initially and at each sampling point. This pHrange was not exceeded, except for batches re-incubated onmodel substrate at 45 °C. Here, pH values were around 7.5from days8 to 27, which suggests that decreased hydrolyticenzyme activities from day8 on where due to unstable pHconditions, possibly exceeding the optimal pH range forhydrolytic enzymes. Further feedings of model substrate or

grass silage after day27 could not provoke additionalincreases of enzyme activities (data not shown).

Overall, bacteria immobilised on activated zeolites,washed and stored for more than 20 months, started produc-tion of hydrolytic enzymes after only a few days upon re-incubation. These results clearly demonstrated that zeolite-habituating functional microorganisms endure or gain along-term revitalisation ability throughout immobilisationon zeolites. Thus, bacteria cells are able to immediately starthydrolytic enzyme production when re-incubated in synthet-ic medium with model substrate at 35 °C or grass silage at45 °C fermentation temperature.

For further characterisation of enzymes associated withzeolite surfaces involved in the degradation of grass silage,identification was realised through LC-MS peptide sequenc-ing analysis and protein database searches. Hydrolyticalactive proteins secreted by zeolite immobilised microorgan-isms after re-incubation were preselected by separation viagel electrophoresis on polyacrylamide gel matrix incorpo-rating compounds of plant cell wall biopolymers. All three

similarity [%]

100

95908580757065605550

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

surface

surface

surface

surface

surface

surface

surface

surface

surface

surface

surface

surface

surface

surface

surface

surface

surface

surface

surface

surface

surface

surface

GS1.35.17GS2.35.17GS3.35.17GS2.45.17GS3.45.17GS1.45.17MS1.45.17MS2.45.17MS3.45.17MS1.35.17MS3.35.17MS2.35.17I2.45.285I3.45.285IS2.45.84IS2.45.42I2.45.264IS3.45.21IS3.45.42IS3.45.84I3.45.264I1.45.285IS3.45.7I1.45.264I3.45.353IS3.45.14IS2.45.14IS2.45.21IS2.45.7

I II III

zeolite surfaceb

zeolite surfacea

I

III

II

Fig. 5 Clustering analysis generated dendrogram using band-basedcomparison (DICE; 1 % tolerance, 1 % optimisation, grouped byUPGMA) of anaerobic populations on activated zeolite surfaces(highlighted in grey) taken from batch-cultures (zeolite surface b)and semi-continuous operated bioreactors [zeolite surface a and

total sludge samples (highlighted in white)]. Sample-identity: I2/3=bioreactors (IS=in sacco sample); MS/GS=batch-substrate, 35/45=fermentation temperature in (in degree Celsius), 7–353=fermentationtime [in days])


substrates for cellulases/hemicellulases were useful fordetecting active enzymes, although CMC delivered thesharpest protein bands compared to lichenan and was there-fore preferred for enzyme separation. Samples were collect-ed from batch-cultivated activated zeolite inhabitinganaerobic populations during re-incubation in syntheticmineral medium on model substrate or grass silage(Table 2). Sampling points were coordinated with positiveenzyme activity findings. When lichenan was used, onlyexocellobiohydrolase II (3.2.1.91) and amylase (3.2.1.1)

were identified throughout the total re-incubation periodupon model substrate at 35 °C. At day1, several positivematches for grass silage fed batch-cultures re-incubated at35 °C appeared, i.e. cellulases (exocellobiohydrolases),hemicellulase (1,4-β-mannanohydrolase) and a sugar-hydrolysing enzyme with molecular weights ranging from47 to 149 kD exclusively related to Paenibacillus polymyxastrains from the division Firmicutes. At day8 of re-incubation, where highest enzyme activities in batch-cultures fed with model substrate (at 35/45 °C) and grass

Table 1 Identified clinoptilolite-associated microorganisms separatedby SSCP analysis and direct sequencing from bioreactors semi-continuously operated with grass silage; sample-identity: I2/3=bio-reactors (IS=in sacco sample); fermentation temperature during the

sampling period: 45 °C; 7–84=days of fermentation, when appearan-ces were observed; band numbers are in accordance with band selec-tion marked in Fig. 4 (see above)

Fermentation BLASTacc. no.

Indicated organism Family Querycoverage (%)

Maximumidentity (%)

Mode Time (day) Band

Continuous I3 7, 21 4 AY196674.1 Methanoculleus bourgensis Methanomicrobiaceae 90 98

Continuous I3 7, 21 EF118904.1 Methanoculleus thermophilus Methanomicrobiaceae 90 98

Continuous I3 7, 21 DQ787476.1 Methanoculleus receptaculi strain ZC-2,3 Methanomicrobiaceae 90 98

Continuous I3 7, 21–42 7 AF107104 Methanoculleus sp. MAB2 Methanomicrobiaceae 92 99

Continuous I3 7, 21–42 AY196674 Methanoculleus bourgensis Methanomicrobiaceae 92 99

Continuous I3 7, 21–42 DQ787476.1 Methanoculleus receptaculi strain ZC-2,3 Methanomicrobiaceae 92 98

Continuous I3 7, 21–42 NR_028253.1 Methanoculleus palmolei DSM 4273 Methanomicrobiaceae 92 97

Continuous I3 7, 21–42 AJ862839.1 Methanoculleus thermophilus Methanomicrobiaceae 92 97

Continuous I3 7, 21–42 AF095270.1 Methanoculleus olentangyi Methanomicrobiaceae 92 96

Continuous I3 7, 21–42 8 AF107103.1 Methanoculleus sp. MAB1 Methanomicrobiaceae 96 99

Continuous I3 7, 21–42 AY196674 Methanoculleus bourgensis Methanomicrobiaceae 96 99

Continuous I3 7, 21–42 DQ787476.1 Methanoculleus receptaculi strain ZC-2,3 Methanomicrobiaceae 96 98

Continuous I3 7, 21–42 NR_028253.1 Methanoculleus palmolei DSM 4273 Methanomicrobiaceae 96 97

Continuous I3 7, 21–42 NR_028163.1 Methanolinea tarda NOBI-1 Methanomicrobiaceae 96 96

Continuous I2 42–84 1 AY196674 Methanoculleus bourgensis Methanomicrobiaceae 90 99

Continuous I2 42–84 EF118904 Methanoculleus thermophilus strain JB-1 Methanomicrobiaceae 90 98

Continuous I2 42–84 DQ787476 Methanoculleus receptaculi Methanomicrobiaceae 90 98

Continuous I2 42–84 EU722338.1 Methanoculleus marisnigri Methanomicrobiaceae 90 98

Continuous I2 42–84 NR_028856 Methanoculleus submarinus strain Nankai-1 Methanomicrobiaceae 90 98

Continuous I2 42–84 NR_028152.1 Methanoculleus chikugoensis MG62 Methanomicrobiaceae 90 98

Continuous I2 42–84 NR_028253.1 Methanoculleus palmolei DSM 4273 Methanomicrobiaceae 90 98

Continuous I2 7–84 3 AY196674 Methanoculleus bourgensis Methanomicrobiaceae 90 98

Continuous I2 7–84 EF118904 Methanoculleus thermophilus strain JB-1 Methanomicrobiaceae 90 98

Continuous I2 7–84 DQ787476 Methanoculleus receptaculi Methanomicrobiaceae 90 98

Continuous I2 7–84 EU722338.1 Methanoculleus marisnigri Methanomicrobiaceae 90 98

Continuous I2 7–84 NR_028856 Methanoculleus submarinus strain Nankai-1 Methanomicrobiaceae 90 98

Continuous I2 7–84 NR_028152.1 Methanoculleus chikugoensis MG62 Methanomicrobiaceae 90 98

Continuous I2 7–84 NR_028253.1 Methanoculleus palmolei DSM 4273 Methanomicrobiaceae 90 98

Continuous I2,3 7–84 9 FJ808611 Clostridium sp. Clostridiaceae 96 92

Continuous I2,3 7–84 17 JF432053 Pseudomonas sp. Pseudomonadaceae 100 100

Continuous I2,3 7–84 JF756250.1 Pseudomonas lundensis strain MFPB42A12-09 Pseudomonadaceae 100 100

Continuous I2,3 7–84 HM854252.1 Pseudomonas pelagia strain KTT-14 Pseudomonadaceae 100 100

Continuous I2,3 7–84 HM031486.1 Pseudomonas pertucinogena strain mol25 Pseudomonadaceae 100 100

Continuous I2,3 7–84 FJ232608 Pseudomonas fluorescens Pseudomonadaceae 100 100

Continuous I2,3 7–84 AF173970 Pseudomonas putida Pseudomonadaceae 100 100


silage (at 35 °C) were measured, mainly cellulolytic endo-1,4-/1,3-β-D-glucanohydrolase was identified (foremost re-lated to Paenibacillus spp.). At day16, endoglucanase fromPaenibacillus sp. was present in batch-cultures re-incubatedat 45 °C. Furthermore, endo-1,4-β-glucanase affiliated toSymbiobacterium thermophilum and isomerases ofThermoanaerobacter pseudethanolicus and Fusobacteriumnucleatum were present. At day27, xylose isomerases andglycoside hydrolase related to three genera from within theorder Bacillales were identified, possibly indicating a shifttowards carbohydrate conversion in succession to the for-mer degradation of plant cell wall biopolymers. Besidespositive matches for recalcitrant biomass degradingenzymes, flagellin of P. fluorescens was found with highestscore and sequence coverage (23 %) from day28 on to theend of the observation period at day84 of the semi-continuously operated bioreactor experiment. These findingconfirmed the positive identification of Pseudomonadaceaein the course of the community analysis.

Discussion

Scanning electron and confocal laser scanning microscopyusing FISH gave optical evidence of microbes preferringactivated zeolite surfaces as operational environment.Thereby, colonisation was not only observed on the surfaceof zeolite particles, it also took place inside the porousframework as well. Observed cell morphologies and piliformations on surfaces did match with findings of Jarrellet al. (2011), who reported that Flagella and pili ofMethanococcus maripaludis have a second and first role insurface attachment besides motility and cell-connectingfunctions and that those cells lacking such surface

appendages were unable to attach efficiently to any surfaces.Bacterial biomass degraders in anaerobic processes areknown to produce multi-enzyme complexes, i.e. cellulo-somes for efficient cellulose and hemicellulose degradation(Schwarz 2001; Karita et al. 1997). Here, the attachment tosubstrate surfaces, e.g. fibres of microcrystalline cellulose isrealised by protuberances with dimensions up to 550 nm inlength and between 50 and 300 nm in width (Bayer et al.1998) that were also observed in our study. Furthermore,FISH-labeling confirmed dense biofilm formation on andinside zeolite particles by bacteria exclusively. Althougharchaeal colonisers were detected during community analy-sis and methane production during the colonisation processwas observed (Fig. S1), archaea-specific probing did notlead to corresponding detection signals, which is most likelydue to the lower threshold detection level of the in situhybridization technique compared to PCR-based methods,than real evidence for the absence of methanogens.

Cluster and sequencing analyses were useful tools for theverification of microbial populations associated with zeolitesurfaces. Amongst them, affiliations to the familyClostridiaceae were observed, representing a group of obli-gate anaerobes involving several species from the genusClostridium producing hemicellulolytic and cellulolytic en-zyme activities (Van Gylswyk and van der Toorn 1986).Furthermore, affiliations to the family Pseudomonadaceaewere found amongst zeolite colonisers. These gram-negativeaerobes are involving species that produce non-complexed,cell-free cellulolytic enzymes, e.g. P. cellulosa (Dees et al.1995). Recently, Potivichayanon et al. (2011) demonstratedbiogas production enhancement by the addition ofPseudomonas aeruginosa into the anaerobic biodegradationprocess. In general, Clostridiaceae and Pseudomonadaceaerender well-known degraders of cellulose and lipids

0

500

1000

1500

2000

2500

3000

3500

1 8 16 27 1 8 16 27 1 8 16 27 1 8 16 27

a b c d

hydr

olyt

ic e

nzym

e ac

tivity

[U l-1

]

sample and time [d]

cellulase

xylanase

Fig. 6 Hydrolytic enzymeactivities (cellulase [whitecolumn part]+xylanase [greycolumn part]; in units per litre)determined during batch-wiseoperated re-incubation ofwashed colonised zeolites inmineral salt medium on a mod-el substrate at 35 °C, b modelsubstrate at 45 °C, c grass silageat 35 °C, d grass silage at 45 °Cover a total time period of27 days


producing monomeric sugars, fatty acids and glycerol in an-aerobic processes (Bischofsberger et al. 2004). However,since clinoptilolite–zeolite possesses a nano- and micro-porous framework with an internal surface size of up to400 m2g−1, it possibly provides anaerobic and aerobic zonesfor both obligate anaerobes and aerobes.

Colonisers affiliated to the archaeal domain wereassigned solely to the order Methanomicrobiales. This

group uses the hydrogenotrophic pathway to generate meth-ane that allows faster growth rates than acetotrophic growth,which in turn probably led to its dominant appearance onzeolite surfaces. Sequence analyses indicated the presenceof Methanoculleus sp. amongst the colonisers, a genus thathas also been reported to appear in association with in-creased methane production from model substrate for maizeupon trace metal supplementation (Pobeheim et al. 2010b).

Table 2 Identification of hydrolases secreted by clinoptilolite immo-bilised microorganisms through SDS-PAGE with matrix incorporatedCMC, lichenan or xylan for activity staining. Active bands wereexcised and further analysed (LC-MS/MS, NCBI public databasesearch with Mascot). Only hits with highest score are listed (except

for samples collected at day8 where ID was still reliable, but highestscores not reached). Samples were collected from batch-cultivatedactivated clinoptilolite migulator inhabiting microorganisms duringre-incubation on model substrate (MS) or grass silage (GS) over atotal time period of 27 days

Cultivation gi no. EC no. Enzyme Mass(kD)

Related organism Score Peptides Sequence

Substrate Temp.(°C)

Time(day)

(unique) Coverage

GS 35 1 308067135 3.2.1.55 Arabinosidase ~57 Paenibacillus polymyxa E681 113 2 (2) 7

GS 35 1 308071054 3.2.1.78 1,4-β-Mannase B ~149 Paenibacillus polymyxa E681 160 2 (2) 2

GS 35 1 308067781 3.2.1.91 Exocellobiohydrolase 3 ~77 Paenibacillus polymyxa E681 775 24 (12) 18

GS 35 1 308069068 3.2.1.91 Exocellobiohydrolase II ~108 Paenibacillus polymyxa E681 271 4 (4) 6

GS 35 1 310640531 3.2.1.91 Exoglucanase A ~77 Paenibacillus polymyxa SC2 609 16 (10) 19

GS 35 1 308071411 4.2.2.2 Pectate lyase 2 ~47 Paenibacillus polymyxa E681 190 4 (3) 12

GS 35 1 308068567 3.2.1.3 Glucosidase yic I ~87 Paenibacillus polymyxa E681 103 2 (2) 3

GS 35 8 146350814 3.2.1.4 Endoglucanase ~62 Paenibacillus sp. KSM-N659 120 3 (3) 5

GS 35 8 4249556 3.2.1.4 Endoglucanase ~33 Humicola grisea var. Thermoidea 129 3 9

GS 35 8 232219 3.2.1.4 Endoglucanase A ~77 Paenibacillus lautus 168 4 (3) 5

MS 35 8 146350814 3.2.1.4 Endoglucanase ~62 Paenibacillus sp. KSM-N659 120 3 2




MS 35 8 232219 3.2.1.4 Endoglucanase A ~77 Paenibacillus lautus 168 3 5

MS 35 8 308069068 3.2.1.91 Exocellobiohydrolase II ~108 Paenibacillus polymyxa E681 130 2 2

MS 35 8 310641984 3.2.1.3 Glycoside hydrolase family 48 ~110 Paenibacillus polymyxa SC2 130 2 2

GS 35 16 4249556 3.2.1.4 Endoglucanase ~33 Humicola grisea var. Thermoidea 170 7 (4) 15

GS 35 16 16078876 3.2.1.4 Endoxylanase ~47 Bacillus subtilis 168 72 4 (2) 4

MS 35 16 1170135 3.2.1.4 Cellulase ~24 Humicola insolens 87 2 9


GS 45 16 51893813 3.2.1.4 Endo-1,4-β-glucanase ~40 Symbiobacterium thermophilum 570 8 35




MS 45 16 308068047 5.3.1.9 Glucose-6-phosphate isomerase ~50 Paenibacillus polymyxa E681 116 3 7

MS 45 16 167037114 5.3.1.9 Glucose-6-phosphate isomerase ~50 Thermoanaerobacter pseudethanolicusATCC 33223

79 2 6

MS 45 16 34763477 5.3.1.9 Glucose-6-phosphate isomerase ~42 Fusobacterium nucleatum ATCC 49256 139 3 7

MS 45 16 253574746 5.3.1.4 L-Arabinose isomerase ~55 Paenibacillus sp. D14 92 2 4

MS 35 27 232219 3.2.1.4 Endoglucanase A ~77 Paenibacillus lautus 619 12 (8) 19

MS 35 27 310641984 3.2.1.3 Glycoside hydrolase family 48 ~110 Paenibacillus polymyxa SC2 305 10 (5) 5

MS 35 27 253574413 5.3.1.14 Xylose isomerase ~49 Paenibacillus sp. D14 227 10 (4) 8

MS 35 27 1750124 5.3.1.14 Xylose isomerase ~50 Bacillus subtilis 164 2 6

MS 35 27 310641984 3.2.1.3 Glycoside hydrolase family 48 ~110 Paenibacillus polymyxa SC2 305 5 5

MS 35 27 27227837 3.2.1.8 Xylanase 5 ~143 Paenibacillus sp. W-61 128 2 (2) 1

MS 35 27 261409352 5.3.1.5 Xylose isomerase ~49 Geobacillus sp. Y412MC10 196 2 6


Amongst coccoid methanogens, which grow on H2–CO2,formate and some secondary alcohols, the presence ofMethanoculleus marisnigri was indicated, although its mor-phology could not be observed in microscopy approaches.Cells are usually 1 to 2 μm in diameter and have optimumgrowth conditions at mesophilic temperatures and slightlyalkaline pH values (Maestrojuán et al. 1990). All the meth-anogenic species we found were identified under mild ther-mophilic conditions.

Enzyme identification complemented the community anal-ysis revealing a broad spectrum of cellulolytical enzymes andrelated organisms that belonged to the genera Bacillus sp.,Clostridium sp. and Paenibacillus sp. Generated peptidesequences via LC-MS resulted in 36 positive hits coveringup several important enzymes for the degradation of plant cellwall polymers. Additionally, a set of isomerases was foundthat was affiliated to, e.g. Thermoanaerobacter pseudethano-licus, a rod-shaped, spore-forming, carbohydrate fermentingbacterium (Onyenwoke et al. 2007) and Fusobacteriumnucleatum, that is closely related to Bacteroides sp. andFlavobacterium sp. (Bolstad et al. 1996). The conver-sion of D-glucose-6-phosphate to D-fructose-6-phosphatemediated by isomerases represents an energy-deliveringstep within the bacterial metabolism and was probablyevolved subsequent to hydrolysis of complex polysac-charides. In addition, enzyme activity tests revealed along-term revitalisation ability regarding cellulases andxylanases associated with zeolite-habituating microor-ganisms. Former experiments showed that such popula-tions can be introduced to the fermentation processresulting in significantly increased methane productivity(Weiß et al. 2010). In conclusion, activated zeolites areeminently suited as natural carrier material for micro-organisms, potentially stabilising and enhancing the bio-gas production process from recalcitrant plant biomass.

Acknowledgments This study was performed within the AustrianCentre of Industrial Biotechnology ACIB and has been kindly sup-ported by IPUS GmbH (Austria), the Federal Ministry of Economy,Family and Youth (BMWFJ), the Federal Ministry of Traffic, Innova-tion and Technology (bmvit), the Styrian Business Promotion AgencySFG, the Standortagentur Tirol and ZIT–Technology Agency of theCity of Vienna through the COMET-Funding Program managed by theAustrian Research Promotion Agency FFG. Many thanks go to con-tributors of the collaborating organisations and institutes (FELMI-ZFE,ZMF and LfL).

References

Altschul SF, Madden TL, Schaffer AA, Zhang JH, Zhang Z, Miller W,Lipman DJ (1997) Gapped BLAST and PSI-BLAST: a newgeneration of protein database search programs. Nucleic AcidsRes 25(17):3389–3402

Amann RI, Binder BJ, Olson RJ, Chisholm SW, Devereux R, Stahl DA(1990) Combination of 16S rRNA-targeted oligonucleotideprobes with flow cytometry for analyzing mixed microbial pop-ulations. Appl Environ Microbiol 56(6):1919–1925

Bailey MJ, Biely P, Poutanen K (1992) Interlaboratory testing ofmethods for assay of xylanase activity. J Biotech 23:257–270

Banks CJ, Zhang Y, Jiang Y, Heaven S (2012) Trace element require-ments for stable food waste digestion at elevated ammonia con-centrations. Biores Technol 104:127–135

Bayer EA, Shimon LJW, Shoham Y, Lamed R (1998) Cellulosomes—structure and ultrastructure. J Struct Biol 124:221–234

Bischofsberger W, Böhnke B, Seyfried CF (2004) MikrobiologischeGrundlagen des anaeroben Abbaus – Überblick und Organismen.In: Dauber S (ed) Anaerobtechnik. Springer, Berlin, Germany, pp23–49

Bolstad AI, Jensen HB, Bakken V (1996) Taxonomy, biology andperiodontal aspects of Fusobacterium nucleatum. Clin MicrobiolRev 9:55–71

Bougrier C, Dognin D, Laroche C, Rivero JC (2011) Effects of addi-tives solutions of trace elements on anaerobic digestion of maizesilage. 1st International Conference on Biogas Microbiology.Leipzig, Germany, September 14–16, Proceedings P-10, p 75

Bruni E, Jensen AP, Angelidaki I (2010) Comparative study of me-chanical, hydrothermal, chemical and enzymatic treatments ofdigested biofibers to improve biogas production. Biores Technol101(22):8713–8717

Daims H, Bruhl A, Amann R, Schleifer KH, Wagner M (1999) Thedomainspecific probe EUB338 is insufficient for the detection ofall bacteria: development and evaluation of a more comprehensiveprobe set. System Appl Microbiol 22:434–444

Dees C, Ringelberg D, Scott TC, Phelps TJ (1995) Characterisation ofthe cellulose degrading bacterium NCIMB 10462. Appl BiochemBiotechnol 51:263–274

Fernández N, Montalvo S, Fernández-Polanco F, Guerrero L, Cortés I,Borja R, Sánchez E, Travieso L (2007) Real evidence aboutzeolite as microorganisms immobilizer in anaerobic fluidizedbed reactors. Process Biochem 42(4):721–728

Goldstein JI, Newbury DE, Echlin P, Joy DC, Lyman CE, Lifshin E,Sawyer L, Michael JR (2003) Scanning electron microscopy and X-ray microanalysis. Kluwer Academic/Plenum Publishers, NewYork

Goodwin JAS, Wase DAJ, Forster CF (1990) Effects of nutrientlimitation on the anaerobic up flow sludge blanket reactor.Enzyme Microbial Technol 12:877–884

Holper J, Lesjak M, Heinzel U, Boos B (2005) Zeolith in derBiogasgewinnung. Sonn. 05450052.5 [EP 1 577 269 A1].European Patent, Styria, Austria

Jarrell KF, Stark M, Nair DB, Chong JPJ (2011) Flagella and pili areboth necessary for efficient attachment of Methanococcus mar-ipaludis to surfaces. FEMS Microbiol Lett 319(1):44–50

Karita S, Sakka K, Ohmiya K (1997) Cellulosomes, cellulasecomplexes, of anaerobic microbes: their structure modelsand functions. In: Onodera R, Itabashi H, Ushida K, YanoH, Sasaki Y (eds) Rumen microbes and digestive physiologyin ruminants. Japan Science Society Tokyo/Karger, Basel, pp47–57

Kang D, Gho YS, Suh M, Kang C (2002) Highly sensitive and fastprotein detection with Coomassie Brilliant Blue in sodiumdodecyl sulfate-polyacrylamide gel electrophoresis. Bull KoreanChem Soc 23(11):1511–1512

Laemmli UK (1970) Cleavage of structural proteins during the assemblyof the head of bacteriophage T4. Nature 227:680–685

Lebuhn M, Liu F, Heuwinkel H, Gronauer A (2008) Biogas productionfrom mono-digestion of maize silage-long-term process stabilityand requirements. Wat Sci Tech 58(8):1645–1651

Lieber A, Kiesel B, Babel W (2002) Microbial diversity analysis of soilby SSCP fingerprinting technique using TGGE Maxi System. In:


Merbach W, Hütsch BW, August in J (eds) Teubner.Ökophysiologie des Wurzelraumes, Stuttgart, pp 61–65

Lowry OH, Rosebrough NJ, Farr AL, Randall RJ (1951) Proteinmeasurement with the Folin-Phenol reagents. J Biol Chem193:265–275

Maestrojuán GM, Boone DR, Xun L, Mah RA, Zhang L (1990)Transfer of Methanogenium bourgense, Methanogenium marisni-gri, Methanogenium olentangyi, and Methanogenium thermophi-licum to the genus Methanoculleus gen. nov., Emendation ofMethanoculleus marisnigri and Methanogenium, and descriptionof new strains of Methanoculleus bourgense and Methanoculleusmarisnigri. Int J Syst Bacteriol 40(2):117–122

Onyenwoke RU, Vadim VK, Anatolly ML, Wiegel J (2007)Thermoanaerobacter pseudethanolicus sp. nov., a thermophilicheterotrophic anaerobe from Yellowstone National Park. Int J SystEvol Microbiol 57:2191–2193

Pobeheim H, Munk B, Johansson J, Guebitz GM (2010a)Influence of trace elements on methane formation from asynthetic model substrate for maize silage. Biores Technol101(2):836–839

Pobeheim H, Munk B, Mueller H, Berg G, Guebitz GM (2010b)Characterization of an anaerobic population digesting a modelsubstrate for maize in the presence of trace metals.Chemosphere 80(8):829–836

Potivichayanon S, Sungmon T, Chaikongmao W, Kamvanin S (2011)Enhancement of biogas production from bakery waste byPseudomonas aeruginosa. World Acad Sci Eng Technol56:529–532

Schwarz WH (2001) The cellulosome and cellulose degradation byanaerobic bacteria. Appl Microbiol Biotechnol 56(5–6):634–649

Schwieger F, Tebbe C (1998) A new approach to utilize PCR-SSCP for16S rRNA gene-based microbial community analysis. ApplEnviron Microbiol 64(12):4870–4876

Shevchenko A, Wilm M, Vorm O, Mann M (1996) Mass spectrometricsequencing of proteins from silver-stained polyacrylamide gels.Anal Chem 68(5):850–858

Stahl DA, Amann R (1996) Development and application of nucleic acidprobes. In: Stackebrandt E, Goodfellow M (eds) Nucleic acid tech-niques in bacterial systematics. Wiley, Chichester, pp 205–248

Takashima M, Speece RE (1989) Mineral nutrient requirements forhigh-rate methane fermentation of acetate at low SRT. J Wat PollCont Fed 61:1645–1650

Van Gylswyk NO, van der Toorn JJTK (1986) Enumeration ofBacteroides succinogenes in the rumen of sheep fed maize-strawdiets. FEBS Lett 38(4):205–209

Weiß S, Tauber M, Somitsch W, Meincke R, Müller H, Berg G,Guebitz GM (2010) Enhancement of biogas production by addi-tion of hemicellulolytic bacteria immobilised on activated zeolite.Wat Res 44(6):1970–1980

Weiß S, Zankel A, Lebuhn M, Petrak S, Somitsch W, Guebitz GM(2011) Investigation of mircroorganisms colonising activated zeo-lites during anaerobic biogas production from grass silage. BioresTechnol 102(6):4353–4359

Wichern M, Gehring T, Fischer K, Andrade D, Luebken M, Koch K,Gronauer A, Horn H (2009) Monofermentation of grass silageunder mesophilic conditions: measurements and mathematicalmodeling with ADM 1. Biores Technol 100(4):1675–1681

Yu Y, Lee C, Kim J, Hwang S (2005) Group-specific primer and probesets to detect methanogenic communities using quantitative real-time polymerase chain reaction. Biotechnol Bioeng 89(6):670–679

Zheng D, Alm EW, Stahl DA, Raskin L (1996) Characterization ofuniversal small-unit rRNA hybridization probes for quantitativemolecular ecology studies. Appl Environ Microbiol 62:4504–4513

Zverlov VV, Hiegl W, Köck DE, Kellermann J, Köllmeier T, SchwarzWH (2010) Hydrolytic bacteria in mesophilic and thermophilicdegradation of plant biomass. Eng Life Sci 10(6):528–536


activated zeolite—suitable carriers for microorganisms in anaerobic digestion processes?

Documents