The Pennsylvania State University
The Graduate School
Department of Crop and Soil Sciences
CHARACTERIZATION OF H2-PRODUCING BACTERIAL COMMUNITIES
FROM HEAT-TREATED SOIL AND ISOLATION OF DOMINANT
CLOSTRIDIUM SPP.
A Thesis in
Soil Science
by
Yonghua Luo
© 2007 Yonghua Luo
Submitted in Partial Fulfillment of the Requirements for the Degree of
Doctor of Philosophy
August 2007
The thesis of Yonghua Luo was reviewed and approved* by the following:
Mary Ann Bruns Associate Professor of Soil Science Thesis Advisor Chair of Committee
John M.Regan Assistant Professor of Civil and Environmental Engineering
David R.Huff Associate Professor of Turfgrass Breeding and Genetics
Hery Lin Assistant Professor of Hydropedology/Soil Hydrology
Curtis J.Dell Soil Scientist and Adjunct Professor of Soil Science
David M.Sylvia Professor of Soil Microbiology Head of the Department of Crop and Soil Sciences
*Signatures are on file in the Graduate School
iii ABSTRACT
Hydrogen is an attractive alternative to conventional fossil fuels. Currently most hydrogen is produced from nonrenewable natural gas, oil and coal. However, from the perspectives of economics and environmental quality, hydrogen production from renewable resources is gaining more and more attention. Biohydrogen production from the fermentation of wastewater containing carbohydrates is one approach for exploiting renewable resources. Much literature has been published on the effects of external factors such as substrate concentration, temperature, pH, etc. on hydrogen production by pure cultures or defined mixed cultures. However, only a few studies have been conducted on the effects of these external factors on H2-producing microbial communities from
uncharacterized, mixed inocula. The objectives of this research were threefold: to
determine the effect of glucose concentration on H2-producing bacterial community
composition in the fermentation of synthetic wastewater; to determine the effect of L-
cysteine and successive transfer on H2 producing bacterial community composition in
batch cultures; and to isolate and characterize a new strain of H2-producing Clostridium
spp. from mixed soil inocula using L-cysteine as reducing agent. In the first study, a
continuous flow bioreactor was continuously fed with synthetic wastewater for periods of
7-14 days. Heat-treated soil was introduced into the reactor as a mixed inoculum and L-
cysteine was added to a final concentration of 0.5g/L as reducing agent. PCR-based
ribosomal RNA intergenic spacer analysis (RISA) was used to characterize the bacterial
community composition in the bioreactor. It was revealed that RISA profiles of bacterial
communities grown in media with 5.0, 7.5 or 10g/L glucose yielded partial 16S rRNA iv sequences most related to Clostridium spp. In contrast, RISA profiles of cultures grown
in 2.5 g/L glucose yielded more diverse DNA sequences. The lowest glucose
concentration gave rise to sequences representing two bacterial families, Clostridiaceae
and Acidaminococcaeae, in the low G+C Gram-positive Firmicutes division, and one
family, Enterobacteraceae in the gamma Proteobacteria division. Using an oligonucleotide probe “LYHI”, complementary to rRNA of an isolate obtained from the reactor containing 10 g/L glucose, it was shown that about 90% percent of the bacterial cells in microscope fields from bioreactor samples at 10 g/L glucose hybridized with the
LYH1 probe. In samples from the bioreactor fed with 2.5 g/L glucose, only 26% of
bacterial cells hybridized to LYH1 probe. These results further confirmed the effect of
glucose on the composition of bacterial communities as judged by RISA. In the second
study, batch experiments were carried out with the same nutrient levels used in the continuous flow bioreactor. Regardless of the glucose concentration and number of serial
transfers, RISA profiles of bacterial communities grown without L-cysteine yielded partial 16S rRNA sequences closely related to Enterobacteriaceae. In contrast, RISA profiles of batch cultures with L-cysteine yielded results similar to those observed with the continuous flow bioreactor, which indicated that 5g/L appeared to be the lowest favorable glucose concentration for the successful competition of Clostridium spp. over other bacterial species. L-cysteine could have affected the composition of bacterial communities in batch experiments either due to its function as a quick oxygen scavenger or its role as a spore germinant. In the third study, a new Clostridium sp. strain LYH2
was isolated from a silty clay loam soil using L-cysteine as a reducing agent in the
growth medium. Closest cultured relatives of strain LYH2 were Clostridium acidisoli, v Clostridium akagii and Clostridium pasteurianum. This dissertation showed that heat- treated soil provided a useful source of Clostridium spp. for producing biohydrogen from renewable wastes, and that glucose concentrations greater than 5g/L and L-cysteine enhanced their growth and H2 production.
vi TABLE OF CONTENTS
ABSTRACT…………………………………………………………………………....iii
TABLE OF CONTENTS……………………………………………………………...... vi
LIST OF FIGURES…………………………………...... vviiii
LIST OF TABLES...... xiii
ACKNOWLEDGEMENTS...... xiv
Chapter 1 Introduction…………………………………………………………………...1
REFERENCES…………………………………………………………………………..9
Chapter 2 The effect of glucose concentration on composition of H2- producing
bacterial communities in a continuous flow bioreactor………………….………...…....14
ABSTRACT…………………………………………………………………………… 15
INTRODUCTION…………………………………………………………………….. ...17
MATERIALS AND METHODS……………………………………………………….19
RESULTS……………………………………………………………………………...... 26
DISCUSSION………………………………………………………………………...... 30
REFERENCES………………………………………………………………………...... 37
Chapter 3 The effect of L-cysteine on bacterial community composition and H2
production in batch cultures of heat-treated soil inocula……………………………….57
ABSTRACT………………………………………………………………………...... 58
INTRODUCTION…………………………………………………………….……. …...60
MATERIALS AND METHODS……………………………………………………...... 61
RESULTS…………………………………………………………………………….....66
DISCUSSION…………………………………………………………………………..69 vii REFERENCE………………………………………………………………………….74
Chapter 4 Strain LYH2, a N2-fixing, ferric-iron reducing Clostridium sp. isolated from a
silty clay loam soil………………………………………………………….………….85
ABSTRACT……………………………………………………………………………...86
INTRODUCTION……………………………………………………………………….87
MATERIALS AND METHODS……………………………………………………….88
RESULTS……………………………………………………………………………...93
DISCUSSION……………………………………………………………………….....96
REFERENCE…………………………………………………………………………..99
Chapter 5 GENERAL CONCLUSIONS……………………………………………..110
Conclusion
Appendix A…………………………………………………………………………..114
Appendix B…………………………………………………………………………... 116 viii LIST OF FIGURES
Figure 1-1: H2 production pathways…………………………………………………..3
Figure 1-2: Pathway to hydrogen production by fermentation………………………...4
Figure 1-3 Structural formula of L-cysteine. The thiol group can effectively bind
oxygen in liquid medium.………………………...... 7
Figure 2-1: RISA profiles showing the bacterial community fingerprints of bioreactor
samples with different glucose concentrations and hydraulic retention times. The
glucose concentrations and hydraulic retention times are indicated at the top. The
sizes of DNA marker bands are indicated by the arrows. The different DNA
bands in bioreactor samples are indicated with arrows in their respective lanes
..………………………………………………………………………………....44
Figure 2-2: Petri plate after 6 days of incubation on anaerobic medium containing
L-cysteine and 10g/L glucose, showing colonies of H2 producing bacteria...…...47
Figure 2-3: RISA fingerprint of genomic DNA of isolate from 10g/L glucose culture.
The sizes of DNA bands are indicated by the arrows………………………….. 48
Figure 2-4: Neighbor-joining dendrogram depicting phylogenetic relationships of
strain LYH1 with other Clostridium spp. based on nearly complete 16S rRNA
sequences, using Bacillus subtilis as outgroup. GenBank accession numbers are
in parentheses. The numbers at the nodes represent bootstrap values of 1,000
replicates. The scale bar represents the calculated number of changes per
nucleotide position………………………………………………………………49
Figure 2-5: Comparison of biogas production (A) and H2 percentage in biogas (B) in ix samples grown with 0.5g/L L-cysteine and without L-cysteine. Column heights
represent means and error bars indicate standard deviations. …………………....50
Figure 2-6: Accumulated gas production during growth of strain LYH1in batch cultures
with and without L-cysteine (squares) and without L-cysteine (diamonds)
…………………………………………………………………………………...51
Figure 2-7: Alignments of the LYH1 probe sequence, its target site, and sequences of
corresponding sites in reference organisms. 17 related clostridia,
Selenomonas strain SB90, Enterobacter cloacae strain CP1 and Bacillus
subtilis were selected as reference organisms. The target sequence is displayed in
the upper row; mismatches between target and sequences of corresponding sites
in reference organisms are shadowed……………………………………………...52
Figure 2-8: Confocal Scanning Laser Microscope images of sample culture at 10g/L
glucose concentration with 10-h HRT: (A) The hybridization of sample cells
with universal EUB probe labeled Cy-3 and LYH1 probe labeled by Cy-5. (B)
Morphology of sample cells examined by differential interference contrast
……………………………………………………………………………………53
Figure 2-9: Confocal Scanning Laser Microscope images of sample culture at 2.5g/L
glucose concentration with 10-h HRT: (A) The hybridization of sample cells with
universal EUB probe labeled Cy-3 and LYH1 probe labeled by Cy-5. (B)
Morphology of cells examined by differential interference contrast
…………………………………………………………………………………...55
Figure 3-1: RISA fingerprint of batch populations from the samples without L-cysteine
showing the bacterial community fingerprints of batch cultures without L-cysteine. x The glucose concentrations and the number of serial transfers indicated at the top.
The size of DNA bands are indicated by the arrows
…………………………………………………………………………………..80
Figure 3-2: RISA fingerprint of batch populations from the samples with L-cysteine
showing the bacterial community fingerprints of batch cultures without L-cysteine.
The glucose concentrations and the number of serial transfers indicated at the top.
The size of DNA bands are indicated by the arrows
……………………………………………………………………………...... 80
Figure 3-3: The cumulative biogas production by the batch cultures without L-
cysteine………………………………………………………………………...81
Figure 3-4: The cumulative biogas production by the batch cultures with L-
cysteine………………………………………………………………………....81
Figure 3-5: H2 percentage in the biogas produced from batch culture at 2.5 g/L glucose
during bacterial growth phase…………………………………………………..82
Figure 3-6: H2 yield from the batch cultures without L-cysteine…………………….83
Figure 3-7: H2 yield from the batch cultures with L-cysteine………………………..83
Figure 4-1: Neighbor-joining dendrogram illustrating phylogenetic relationship
between strain LYH2 and its Clostridium relatives based on nearly
complete 16S rRNA gene sequences. .The. tree .was constructed using neighbor-
joining algorithm with 100 bootstrappings. Numbers at branching points refer to
bootstrap values > 50%. Accession numbers of 16S rRNA gene sequences are
given in parentheses. The scale bar represents a 1% difference in nucleotide
sequence...... 104 xi Figure 4-2: Transmission electron micrographs of strain LYH2. (a) Negatively strain
preparation. (b) Thin section. Abbreviation: WL wall layer; CM, cytoplamic
membrane. (c) Thin section showing septum formation at two division points (S).
Bar: 11µm(a), 0.2µm(b), 5µm (c)……………………………………………..105
Figure 4-3: Effect of pH on the growth of LYH2 in TSB medium. Initial pH of medium:
§,3.7,4.2,8.3 and 8.7; ,4.6; , 5.4; ▲,6.0;
{,6.5;e,6.8;□,7.1, ,7.4;+,7.8….106 Figure 4-4: Optimal temperature for the growth of LYH2 in TSB medium (pH 6.2): o o o o o o o z, 25 C, 30 C, 50 C and 55 C ; , 35 C; ,40 C;{, 45 C……………………107
Figure 4-5: After 54-62 hour of incubation 30oC, non-magnetic, brown amorphous iron
(III) oxide in 5g/L glucose culture was converted by LYH2 to a black, solid
material of less volumn, which was strongly attracted to a magnet and contained a
significant amount of Fe(II)……………………………………………………108
Figure B-1: Fig A-1: Alignments of the LYH1 probe sequence, its target site, and
sequences of corresponding sites in reference organisms. 17 related Clostridia,
Bacillus subtilis were selected as reference organisms. The target sequence is
displayed in the upper row; mismatches between target and sequences of
corresponding sites in reference organisms are shadowed…………………..117
Figure B-2: Confocal Scanning Laser Microscope images of sample culture at 10g/L
glucose concentration with 10-h HRT: (A) The hybridization of sample cells with
universal EUB probe labeled by Alexa 488. (B) The hybridization of sample cells
with LYH1-specific probe labeled by Alexa 555. (C) The hybridization of sample
cells with NON EUB probe labeled by Alexa 488. (D) Morphology of sample cells xii examined by confocal scanning laser microscopy……………………………..118
xiii LIST OF TABLES
Table 2-1 Summary of hydrogen production from continuous flow bioreactor
…………………………………………………………………………..45
Table 2-2: Bacterial 16S rDNA sequence of RISA DNA bands derived from
bioreactor cultures…………………………………………...………….46
Table 3-1: Summary of peak gas production, final pH and H2 percent in
headspace when batch cultures with L-cysteine exhibited no further
accumulation of biogas …………………………………………………. 84
Table 4-1: Glucose-dependent product profile of LYH2…………………….109
Table A-1 Summary of continuous flow reactor operation………………….115 xiv ACKNOWLEDGEMENTS
My first acknowledgement goes to my advisor, Mary Ann Bruns, for her endless support of my research, constant guidance, encouragement, and patience during my studies at Penn State University. Her enthusiasm for research and her spirit of devotion to science positively influenced everyone in our lab. During my PhD research, I benefited from her keen insights, inspiration, and positive attitude towards work and life. What I learned from Mary Ann will definitely be an invaluable treasure in both my life and work.
I gratefully acknowledge my committee members: John M. Regan, David R. Huff,
Curtis J. Dell and Henry Lin for their time, suggestions, and valuable comments during the formulation and completion of this thesis. I especially would like to thank John M.
Regan for his useful guidance during my fluorescent in situ hybridization (FISH) experiment.
Many thanks to Bruce E. Logan and Carmen E. Martinez for their lab support,
Husen Zhang for trouble shooting on the problems in the design of my experiments with me and for his helpful advice, Steven Van Ginkel for operating bioreactor, Sang Eun Oh for his technical assistance in the measurement of hydrogen production, and David Jones for the analysis of fermentation end product. Kathleen M. Brown and Amy R. Annino deserve a special thank you for their assistance in helping me measure nitrogenase enzyme activities. I extend additional appreciation to Jennifer Loveland-Curtze, who made the spectrophotometer available for the determination of G+C content.
I am fortunate to have so many friends not only in the Crop and Soil Sciences
Department, but also in the Chinese Friendship Association. Thank you for your help, xv company and sharing pain and happiness with me in my best and worst of times. You made my journey at Penn State University one of the most memorable parts of my life.
Many thanks to my close friends: Sarah, Bill, Yuanhong Zhu, Xianzeng Niu, Keisha,
Michaele, Roren, just to name a few, for their encouragement.
Finally, I am deeply gratefully to my dear mother for her spiritual support, selfless love, and understanding, which offered me great motivation for my study and work.
Special thanks go to my brothers for their meticulous care for my mother, so I could concentrate on my study and work. I also thank my sister for always being in my corner and my wife, Yuchuang Huang, for her encouragement, care and steadfast support when I faced frustrating circumstances in my work. This work is dedicated to them with all my love.
The most important thing in life is to have a great aim and
determination to attain it.
Johann Wolfgang von Goethe (1749-1832)
1
Chapter 1
Introduction
2
Hydrogen production
Hydrogen is an environmentally friendly energy carrier. It is regarded as an important, non-polluting alternative to fossil fuel as only water is produced upon its combustion. Hydrogen can be produced by electrolytic means, but the cost of energy by this route is higher than for fossil fuels (U.S DOE 2003a). As a result, most hydrogen is currently provided from nonrenewable resources, such as gas, oil, and coal (U.S DOE 2003b) (Fig 1-
1). Thus, before fossil fuels are depleted, it will be beneficial to develop economical and sustainable technologies for hydrogen production. With this goal, biological H2 production by fermentative bacteria from renewable waste is being extensively investigated (Kalia et al.
1994, Kataoka et al. 1997, Kumar et al. 1995, Rachman et al. 1998, Taguchi et al. 1995,
Tanisho and Ishiwata, 1995, Lay 2001).
Biological hydrogen production is classified into two categories. One is production by photosynthetic bacteria such as Cyanobacteria, and the other is that by fermentative bacteria such as Clostridium spp. (Fig 1-1). Fermentative hydrogen production has advantages over photosynthetic production for industrial application because it permits hydrogen to be produced continuously from inexpensive waste material in reactors without light. Fermentation itself is a process in which an organic compound serves as the source of both the electron donor and the electron acceptor for ATP production by substrate-level
+ phosphorylation. H2 production results from electron flow to protons (H ) derived from water through the activity of hydrogenases, which serve to dispose of “excess” electrons in
3
the form of molecular hydrogen. Hydrogen production thus functions primarily to maintain
redox balance in the fermentation process.
To date, pure cultures known to produce hydrogen from carbohydrates include species of Enterobacter (Fabiano and Perego 2002, Hussy et al. 2003, Rachman et al. 1998),
Bacillus (Kalia et al. 1994), and Clostridum (Taguchi et al 1995, Yokoi et al. 1997). Among them, Clostridium spp. have been frequently observed in H2-producing mixed cultures from environmental samples and grown in non-sterile feedstocks or synthetic wastewater.
Therefore, Clostridum spp. were presumed to be one of key H2-producing populations in these systems.
Industrial: nonrenewable sources Biological: renewable sources
Natural gas, Organic wastes Solar energy Water Natural gas, oiloil andand coalcoal Clostridium spp Cyanobacteria
Electricity
Gasification and steam reforming Fermentation Photolysis Electrolysis
HydrogenHydrogen
Figure 1-1 Hydrogen production pathways
4
Clostridia
Clostridia are a diverse group of gram-positive, rod-shaped anaerobes. They are
ubiquitous in nature, especially in grassland and water-logged soils. The genus Clostridium
comprises over 150 validly described species (Bahl and Durre 2001) and 20 clusters (Collins
et al. 1994). Cluster I is the largest of the clostridial groups. However, the classification of
clostridia is still being revised. Many clostridial species can convert carbohydrate into
hydrogen as a fermentation by-product via hydrogenase. Carbohydrates could be hydrolyzed
into sugars by some clostridial species first. Sugars are then fermentated into liquid end
products such as acetate, butyrate, ethanol and lactate by clostridia. Hydrogen and CO2 are
released as fermentation by-products (Fig 1-2). Therefore, clostridia have great potential for
conversion of renewable biomass to commodity chemicals. However, hydrogen and CO2
produced during fermentation could also be converted into methane immediately by
methanogens. To increase hydrogen production by fermentation, the pathway leading to
hydrogen consumption by methanogens should be blocked. For example, methanogens could
be killed by heating mixed inocula such as soil and sludge. Or methanogen growth could be
prevented by controlling culture conditions such as low pH, short hydraulic retention time,
and low temperatures.
Hydrolysis Fermentation Methanogenesis Acetate Carbohydrates Sugars Butyrate Ethanol Lactate CH4
Hydrogen + CO2 Figure 1-2 Pathway to hydrogen production by fermentation
5
Research on hydrogen production by fermentation
Although most studies on conversion of organic compounds to hydrogen gas via fermentation have been conducted with pure cultures, such systems could easily become contaminated, particularly by organisms that consume hydrogen. Since significant yields of hydrogen can also be obtained from mixed cultures, mixed cultures of hydrogen-producing bacteria may be more feasible to maintain than pure cultures from an engineering point of view. The most commonly used sources of hydrogen-producing mixed microflora are sludges and soils (Dan et al. 2002, Gong et al. 2005, , Hussy et al. 2003, Iyer et al. 2004, Oh et al. 2004). The sludges and soils are generally heat-treated to destroy methanogens, which could consume H2 and reduce yield. Biohydrogen production from mixed cultures is influenced by carbon source concentration, dissolved H2, temperature, hydraulic retention time (HRT) and pH (Chang and Lin 2004, Kim et al. 2004, Mizuno et al. 2002, Ruzicka,
1996, Van Ginkel and Logan 2005), but only a few papers have reported the effect of these external factors on the composition of microbial communities derived from mixed inocula
(Iyer et al. 2004, Kim et al. 2006). Little information is available on the relationship between hydrogen production and the composition of microbial communities under different growth conditions.
DNA-based molecular methods have been used to characterize mixed microbial populations in biohydrogen-producing fermentations. These include RISA (Ribosomal RNA
Intergenic Spacer Analysis) (Iyer et al. 2004, Oh et al. 2004, Zhang et al. 2002) and DGGE
(denaturing gradient gel electrophoresis) (Borneman and Triplett 1997, Fang and Liu 2002,
Muyzer et al. 1993). Key H2-producing populations in these systems include clostridia,
6
which are frequently found in fermentative mixed microflora (Fang et al 2002, Iyer et al.
2004, Kim et al. 2006, Oh et al. 2004, Ueno et al. 2001). Since fewer than 1% of prokaryotes in most environments are culturable (Amann et al. 1990), molecular techniques are more rapid and reliable than culturing for characterizing community composition.
Research to date on treatment of wastewaters containing carbohydrates via fermentation has shown that clostridia readily convert organic substances into hydrogen, carbon dioxide, acids and solvent end products. Industrial hydrogen production by solventogenic clostridia like Clostridium acetobutylicum should therefore be feasible.
Because of its industrial importance, the metabolic pathways of C. acetobutylicum have been studied in considerable detail (Girbal et al. 1995) and its genome sequence has been determined (Nolling et al. 2001). This information could provide the basis for constructing recombinant Clostridium species with higher H2 production. The isolation of additional H2 producing Clostridium spp. with possibly better H2-production capabilities and different metabolic pathways would expand our capabilities for fermentative H2 production and increase our potential to utilize more diverse sources of wastewater.
In our studies, soil was chosen as the inoculum source of H2-producing mixed microflora because agricultural soils harbor high diversity of microorganisms (Borneman et al. 1996). Furthermore since aerobes and anaerobes are known to coexist in microenvironments of oxic soils, spore-forming anaerobes recovered from oxic soils are likely to be more tolerant to brief exposures to air. As for an O2 scavenger, we chose L- cysteine as the reducing agent. The thiol group of L-cysteine can effectively bind trace O2 in the medium (Fig 1-3). Therefore, the introduction of L-cysteine into the medium could
7
improve the competition of H2-producing anaerobes, such as clostridia, against other bacterial species in mixed culture. Thus, the addition of L-cysteine should be useful for the isolation of H2-producing clostridia from mixed cultures.
Thiol group
Fig 1-3 Structural formula of L-cysteine. The thiol group can
effectively bind oxygen in liquid medium.
The studies conducted during this dissertation research are described in the following chapters. Each chapter is prepared as a manuscript for submission to a peer reviewed journal.
The study objectives in each chapter are as follows:
(1) Determine the effect of glucose concentration on H2-producing bacterial community composition in a bioreactor (chemostat) inoculated with heat-treated soil.
(2) Determine the effect of L-cysteine and successive transfer on H2 producing bacterial community composition in batch cultures. Compare the composition of bacterial communities in chemostat and batch modes under the same nutrient conditions.
(3) Isolate and characterize new H2-producing Clostridium spp. from a mixed soil inoculum using L-cysteine as reducing agent.
8
The significance of this research to soil microbiology and energy science was its finding that agricultural soil provided a good source of clostridia for biohydrogen production by fermentation of sugar-containing wastes. Recommendations for industrial production of
H2 production from wastewater by soil clostridia are to keep concentrations of carbohydrate monomers above 28 mM and to add small amounts of the reducing agent L-cysteine to enhance growth and H2 yield.
9
REFERENCES
1. U.S DOE Website(2003a) http://www.eren.doe.gov/hydrogen/fags.html#cost.
2. U.S DOE Website (2003b) http://www.eren.doe.gov/hydrogenanfuelcells/.
3. Amann RI, Binder BJ, Olson RJ, Chisholm SW, Devereux R, Stahl DA.
(1990) Combination of 16S rRNA targeted oligonucleotide probes with flow-
cytometry for analyzing mixed microbial populations. Appl Environ
Microbiol 56:1919-1925.
4. Bahl H and Durre P (2001) Clostridia: Biotechnology and Medical Application,
Wiley-VCH Verlag GmbH, Weinheim, pp1
5. Borneman J, Skroch PW, O’Sullivan KM, Katherine M, Palus JA, Rumjanek,
Norma G, Jansen JL, Nienhuis J, Triplett EW. (1996) Molecular microbial diversity
of an agricultural soil in Wisconsin. Appli Envrion Microbial 62(6): 1935-1943
6. Borneman J and Triplett EW (1997) Molecular microbial diversity in soils
from eastern Amazonia: evidence for unusual microorganisms and
micropopulation shifts associated with deforestation. Appl Environ Microbiol.
63: 2647-2653
7. Chang F-Y and Lin C-Y (2004). Biohydrogen production using an up-flow
anaerobic sludge blanket reactor. Int J Hydrogen Energy. 29:33-9
8. Collins M.D, Lawson P.A, Willems A, Cordoba J.J, Fernandez-Garayzabal J,
Garcia P, Cai J, Hippe H & Farrow J.A.E (1994). The phylogeny of the genus
Clostridium: proposal of five new genera and eleven new species combination. Ins J
Syst Bacteriol 44,812-826
10
9. Fang HHP and Liu H (2002) Effect of pH on H2 production from glucose by a
mixed culture. Bioresour Technol 82:87-93
10. Fang HHP, Zhang T and Liu H (2002) Microbial diversity of mesophilic
hytrogen-producing sludge. Appl Microbiol Biotechnol 58:112-118
11. Girbal L, Croux C, Vasconcelos I and Soucaille P (1995) Regulation of
metabolic shifts in Clostridium acetobutylicum ATCC 824. FEMS
Microbiology Review. 17:287-297
12. Gong ML, Ren NQ and Xing DF (2005) Start-up of bio-hydrogen production
reactor seeded with sewage sludge and its microbial community analysis. War Sci
Tech 52(1-2): 115-121
13. Hussy, Hawkes FR, Dinsdale R and Hawkes DL (2003) Continuous
fermentative hydrogen production from a wheat starch co-product by mixed-
microflora. Biotechnol Bioeng 84(6): 619-26
14. Iyer P, Bruns MA, H Zhang, G.S Van, Logan BE (2004) H2-Producing
bacterial communities from a heat-treated soil inoculum. Appli Microbiol
Biotechnol 66:166-173
15. Kalia, VC, Jain, SR, and Joshi, AP. (1994) Fermentation of biowaste to
H2 by Bacillus licheniformis. World J Microbiol Biotechnol, 10, 224-227
16. Kataoka, N, Miya, A., and Kiriyama, K (1997) Studies on hydrogen production
by continuous cultures system of hydrogen-producing anaerobic bacteria, War Sci
Tech 36, 41-47.
17. Kim JO, Kim YH, Ryu JY, Song BK, Kim IH, Yeom SH (2004).
11
Immobilization methods for continuous hydrogen gas production biofilm
formation versus granulation. Proc Biochem. 40:1331-7
18. Kim SH, Han SK and Shin HS (2006) Effect of substrate concentration on
hydrogen production and 16S rDNA-based analysis of the microbial community
in a continuous fermenter. Process Biochemitry 41:199-207.
19. Kumar, A., Jain, S.R., Sharma, C.B, Joshi, AP, and Kalia, VC (1995)
Increased H2 production by immobilized microorganisms. World J Microbiol
Biotechnol. 11, 156-159
20. Lay JJ (2001) Biohydrogen generation by mesophilic anaerobic fermentation of
microcrystalline cellulose. Biotechnol Bioeng. 74(4),280-287.
21. Mizuno O, Dinsdale R, Hawkes FR, Hawkes DL. Noile T (2002).
Enhancement of hydrogen production from glucose by nitrogen gas sparaging.
Bioresource Technol
22. Muyer G, De Waal EC and Uitterliinden AG (1993) Profiling of complex
microbial population by denaturing gradient gel eletrophoresis analysis of
polymerase chain reaction-amplified genes coding for 16S rRNA. Appl
Environ Microbiol. 59:695-700
23. Nolling J, Breton G, Omelchenko MV, Makarova KS, Zeng Q, Gibson R,
Lee HM, Dubois J, Qiu D, Hitti J, Wolf Yuri, Tatusov RL, Savathe F,
Sabathe F, Stamm LD, Soucaille P, Daly MJ, Bennett GN, Koonin EV and
Smith DR.(2001) Genome sequence and comparative analysis of the solvent-
producing bacterium Clostridium acetobutylicum. Journal of Bacteriology. pp
12
4823-4838.
24. Oh SE, Lyer P, Bruns MA and Logan BE (2004) Biological hydrogen
production using a membrane bioreactor. Biotechnol.Bioeng 87(1): 119-27.
25. Raehman, MA, Nakashimada, Y., Kakizono, T, and Nishio, N (1998) Hydrogen
Production with high yield and high evolution rate by self-flocculated cells of
Enterobacter aerogenes in a packed-bed reactor. Appl. Microbiol. Biotechnol.,
49, 450-454
26. Ruzicka, M (1996) The effect hydrogen on acidogenic glucose cleavage. Water
Res. 30(10): 2447-2451.
27. Taguchi, F, Mizukami, N, Saito-Taki, T, and Hasegawa, K (1995) Hydrogen
production from continuous fermentation of xylose during growth of
Clostridium sp. strain no. 2. Can. J. Microbiol., 41, 536-540
28. Tanisho, S and Ishiwata, Y (1995) Continuous hydrogen production from molasses
by fermentation using urethane foam as a support of flocs. Int. J. Hydrogen
Energy, 20, 541- 545
29. Ueno Y, Haruta S, Ishii M and Igarashi Y (2001) Microbial community in
anaerobic hydrogen-producing microflora enriched from sludge compost. Appl
Microbiol Biotechnol. 57:555-562.
30. Van Ginkel S and Logan Bruce (2005) Increased biological hydrogen
production with reduced organic loading. Water Research. 39:3819-3826.
31. Yokoi H, Maeda Y, Hirose J, Hayashi S, and Takasaki Y (1997). H2 production
by immobilized cells of clostridium butyricum on porous glass beads. Biotechnol.
13
Lett. 11: 431-433
32. Zhang H, Bruns MA and Logan BE (2002) Perchlorate reduction by a novel
chemolithoautotrophic, H2-oxidizing bacterium. Environ Microbiol. 4:389-395
Strain no.2. Can. J . Mcrobiol.,41,536-540.
14
Chapter 2
The effect of glucose concentration on composition of H2- producing bacterial communities in a continuous flow bioreactor
Luo Y.H., Zhang H., Salerno M., Logan B.E., and Bruns M.A.
(To be submitted for publication to Applied Microbiology and
Biotechnology)
15
ABSTRACT
The effects of glucose concentration (2.5, 5.0, 7.5, 10g/L) and hydraulic retention time (HRT) (1.0, 2.5, 5, 10 hr) on bacterial community composition were assessed with mixed inocula from heat-treated soil in a bioreactor continuously fed with synthetic wastewater. L-cysteine (1 g) was added to the 2-L bioreactor to promote the germination of spores of H2 producers and to work as an oxygen scavenger. PCR-based ribosomal RNA intergenic spacer analysis (RISA) was used to characterize the bacterial community composition in reactors which had reached steady state H2 production. With one exception, all RISA profiles of bacterial communities grown in medium with 5.0, 7.5 or 10g/L glucose at all hydraulic retention time consisted of two DNA bands that yielded partial 16S rRNA sequences most closely related to Clostridium spp. In contrast, RISA profiles of cultures grown in 2.5 g/L glucose were more diverse than those at all other glucose concentrations, yielding DNA sequences representative of Clostridium, Enterobacter and Selenomonas spp.
The dominant culturable population from the reactor operated with 10g/L glucose and 10-h
HRT was designated as strain LYH1 after its isolation on agar plates containing 0.5 g/L L- cysteine and the same nutrient levels as the bioreactor feed. The ability of pure cultures of strain LYH1 to produce hydrogen at 2.5g/L glucose concentration was confirmed in batch experiments. An oligonucleotide probe complementary to the 16S rRNA of LYH1 was designed on the basis of an alignment with rRNA sequences of 17 related Clostridium species. Fluorescence in situ hybridization (FISH) was used to examine the abundance of strain LYH1 in samples from the bioreactor operated with 10 g/L glucose at 10-h HRT and with 2.5g/L glucose at 10h-HRT. In 10g/L bioreactor cultures, about 90 percent of the cells
16
viewed by confocal laser scanning microscopy hybridized with the LYH1 probe. In contrast, only 25 percent of bacterial cells from bioreactor samples fed with 2.5g/L glucose hybridized to the LYH1 probe. FISH data provided quantitative confirmation of RISA results showing that more diverse bacterial communities containing a larger proportion of non-sporeformers became established when the reactor was fed with the lowest glucose concentration.
17
INTRODUCTION
Hydrogen has been gaining the attention of energy research scientists primarily as an attractive alternative to conventional fossil fuels. As the combustion of hydrogen produces only water, it is regarded as a clean energy source that could mitigate the problem of rapidly diminishing, non-renewable fossil fuels (Billings RE, 1991). Aside from its use as an energy source, H2 has other environmental applications as an “universal electron donor.” For example, in wastewater treatment, H2 could take the place of external organic electron donors to stimulate the dechlorination of chlorinated organic pollutants (Ballapragada et al.,
1997; DeWeerd et al., 1991; Distefano et al., 1992; Warner et al., 2002). Hydrogen could also be applied in autotrophic denitrification processes as an electron donor to remove nitrate pollution with less sludge production and without the need for added organic carbon
(Rittmann and Snoeyink, 1984).
Currently most H2 is produced from non-renewable resources, such as natural gas, oil and coal (Chaudhary and Goodman, 1999), but more and more efforts are directed toward sustainable methods of harvesting hydrogen from renewable resources. One of these methods is fermentation of wastewater containing carbohydrates. The coupling of biohydrogen production to utilization of organic waste materials could provide important economic and environmental benefits.
Most studies of biohydrogen production so far have been conducted with pure cultures such as Bacillus licheniformis (Kalia et al. 1994), Enterobacter aerogenes
(Rachman et al, 1998), Clostridium butyricum (Karube et al. 1982) and Clostridium acetobutylicum (Chin et al. 2003). Significant yields of H2 can also be obtained from mixed
18
cultures. Clostridium spp. have been frequently observed to be present in these mixed cultures growing either in non-sterile feed stock or synthetic wastewater (Fang and Liu
2002a, Fang and Liu 2002b, Iyer et al. 2004, Lay et al. 1999). Since the need to use nonsterile wastewater would prohibit pure culture maintenance, the fermentation of wastewater through the activity of mixed microflora appears to be more feasible. Therefore, it is important to understand how mixed microflora can optimally produce hydrogen from organic wastewater. In our previous work, mixed bacterial communities from a heat-treated soil inoculum produced similar amounts of H2 in a continuous flow bioreactor under two different hydraulic retention times (HRTs of 30-h and 10-h) and temperatures (30 oC and 37 oC) at 9.5g/L glucose concentration (Iyer et al. 2004). Van Ginkel and Logan (2005) reported the effect of glucose loading on H2 production in chemostat reactors. In experiments with chemostat bioreactors, H2 yield increased as glucose concentration decreased from 10g/L to
2.5 g/L. The present study provides additional information on these latter bioreactor experiments by demonstrating compositional differences in bacterial communities from the same soil inoculum and grown in the bioreactor at four different glucose concentrations (2.5,
5.0, 7.5, 10g/L) and hydraulic retention times (1, 2.5, 5, 10 hr). An additional goal of our study was the isolation of a dominant culturable hydrogen producer and the confirmation of its capability to produce hydrogen in batch culture. FISH (fluorescent in situ hybridization) method was applied to confirm the result of RISA analysis and quantify the dominant bacterial populations in samples from the bioreactor.
19
MATERIALS AND METHODS
Inoculum
Soil was sampled from harvested tomato plots to 5 cm depth in early October 2004 at the Penn State University Park campus. This soil is classified as Hagerstown silty clay loam
(fine, mesic, typic hapludalfs). Soil was air-dried, ground with a sterile mortar-and-pestle, screened to obtain fine particles, and thoroughly mixed. Soil was then heated in a shallow layer at 105oC for 2 h to reduce numbers of vegetative microorganisms, particularly methanogens that would consume hydrogen (Lay 2001; Biebl 1999; Logan et al. 2002; Van
Ginkel et al. 2002). Soils were distributed into aliquots in sterile tubes and stored at -70oC.
Soil was subjected to only one freeze-thaw cycle prior to use as inoculum.
Culture Condition and Reactor Operation
The mineral medium used in the continuously-stirred tank reactor (CSTR) contained the following components (g/L): NH4Cl (0.88); K2HPO4 (0.44); KH2PO4 (0.44);
MgSO4.7H2O (0.36); FeCl3 (0.05); NiCl2.6H2O (0.0476); CaCl2.2H2O (0.0662); ZnCl2
(0.023); CoCl.6H2O (0.021); CuCl2 (0.01), MnCl2 (0.03), with four different glucose concentrations (10, 7.5, 5.0, 2.5 g/L). A 2-L fermentor (New Brunswick Scientific BioFlo
110) was used. Before each experiment at a given glucose concentration, the reactor was cleaned, filled to 2 L with distilled water and medium, and autoclaved for one hour. It was then allowed to cool and sparged with N2 gas for at least 30 minutes before inoculation. One
20
gram L-cysteine was added (final concentration of 0.5 g/L) to scavenge any remaining oxygen (Song and Logan 2004) and to promote spore germination (Gould, 1970; Montville et al, 1985). The inoculum consisted of 10 g heat-shocked agricultural soil as the microflora source for the bioreactor.
The reactor was initially operated in batch mode to establish a bacterial community without washing it out. After approximately fifty hours of operation in this mode, the reactor was then switched to continuously-fed mode. In continuously-fed mode, the bioreactor was fed with autoclaved medium contained in feed bottles and sparged with N2 to minimize oxygen content. The medium in the feed bottle was stirred continuously throughout. During continuous feeding, the pH was maintained at 5.5 by addition of 2 M KOH as monitored by pH electrode (405-DPAS-SC, Mettler Toledo). At each glucose concentration, the reactor was run first at the 10-hr hydraulic retention time (HRT) until steady state hydrogen production was achieved. Biogas production from the reactor was measured continuously with a respirometer (Challenge Envrionmental System AER-200, Fayetteville, AR), and steady-state hydrogen production at 10-hr HRT was determined when no further increases in biogas were measured. After that, the feeding rate of the same bioreactor culture was increased to 5-hr HRT, with continuous measurements being made until steady-state biogas production was again achieved. This sequence with the bioreactor operated at HRTs of 2.5-h and 1.0-h in turn. The temperature during the experiments was maintained at 30° C with a jacket heater. The bioreactor community samples at each HRT were all taken at steady-state hydrogen production, as determined from constant biogas production rates Prior to removing samples from the bioreactor, the air in the sampling outlet was removed with a syringe from
21
one side first, which caused culture in the bioreactor to flow to sterile tubes attached to the outlet from the other side. The samples were then prepared for molecular community analysis.
Bioreactor experiments were repeated for two contrasting treatment combinations that had been evaluated by van Ginkel and Logan (2005). These two combinations consisted of 10 g/L glucose with 10-h HRT and 2.5 g/L glucose with 10-h HRT. Although biogas and hydrogen in the reactor were measured in these trials, bioreactor samples were not analyzed for bacterial community composition. These latter trials were conducted to assess the reproducibility of results reported by van Ginkel and Logan (2005).
DNA Extraction
Bioreactor samples of 45 ml were centrifuged (AvantiTM J-25 Beckman Coulter) at
7000 rpm (4oC) for 20 minutes. Most of the supernatant was discarded except for 2 ml to resuspend the pellet. Resuspended pellets were homogenized by vortexing (Fisher
Scientific). Samples were frozen and stored at -20C prior to molecular analyses. The cell lysis and DNA extraction procedure followed the UltraClean Micorbial DNA Isolation Kit protocol (Mo Bio).
RISA and Cloning of PCR Product
22
Samples were obtained from the bioreactor operated with 10 g/L glucose at hydraulic retention times (HRTs) of 10-h and 5-h, with 7.5 and 5 g/L glucose concentration at HRTs of 10-h, 5-h, 2.5-h and 1-h, and with 2.5 g/L glucose concentration at HRTs of 10-h, 5-h and
2.5-h. Partial 16S rRNA genes of DNA extracted from bioreactor samples were amplified with the bacteria-specific 16S rDNA primer set 926f (16S rRNA) (5’-
AAACTYAAAKGAATTGACGG-3’) and 115r (23S rRNA) (5’-GGGTTBCCCCATTCGG-
3’) (Lane 1991). Each reaction tube contained PCR buffer (VWR), 2.5 U Taq DNA polymerase (VWR), 30pmol each primer, 250 uM each deoxyribonucleotide and 40 ng DNA template. PCR mixture was adjusted to a total volume of 50 ul in each tube. The reaction mixture was placed in a GeneAmp PCR system 9600 (Perkin Elmer). After an initial denaturation at 94 oC for 5 min, 30 temperature cycles was performed at 94 oC for 30 s, 54 OC for 30 s, and 72 oC for 1 min, followed by a final extension step of 72 oC for 7 min. PCR- amplified DNA products were separated by electrophoresis in 0.7% agarose gels. The gel was prepared and run in 1XTAE buffer (pH8.0). The products were separated by electrophoresis in a 0.7% agarose gel, stained with ethidium bromide, and imaged with an
Epi Chemi II Darkroom unit (UVP Laboratory Products, Upland, CA). DNA bands were then excised and eluted from the gel by the crush and soak procedure (Sambrook et al. 1989).
Eluted DNA bands were cloned using a TOPO TA cloning kit as specified by the manufacturer (Invitrogen). One sequence was obtained for each band. Sequencing was performed with an ABI Hitachi Genetic Analyzer. Sequence similarities between the 16S rRNA portions of PCR amplicons (ca. 600 bp) and GenBank accessions were determined using the Basic Local Alignment Search Tool (BLAST) (Altschul et al., 1990).
23
Isolation and Identification of Dominant H2-Producer
Fresh culture from the bioreactor at 10g/L glucose concentration with 10-h HRT was spread on 1.5% agar plates with the same nutrient levels as the bioreactor feed, including 0.5 g/L L-cysteine. Plates were incubated at 28oC under anaerobic conditions for up to 30 days with frequent visual inspection of colonies. Colonies with uniform morphology (white, opaque and convex) were observed after 6 days. A representative colony was streaked across a fresh plate and a well-isolated colony was picked for growth in liquid medium and preserved in 15% glycerol stocks. This isolate was designated LYH1 and anaerobically maintained in mineral medium previously used in the continuous-flow bioreactor with growth incubation at 30 oC. Genomic DNA of LYH1 was extracted and purified according to the UltraClean Microbial DNA Isolation Kit protocol (Mo Bio). DNA was PCR-amplified with two bacteria-specific 16S rDNA primer sets: (1) 27f (16S rRNA) and 907r (16S rRNA),
(2) 926f (16S rRNA) and 115r (23S rRNA). The nearly complete 16S rRNA gene sequence was obtained from the contiguous sequences derived from PCR amplification with these two primer sets. The 16S rRNA gene sequences most similar to LYH1 were used to identify its closest relatives by BLAST (Altschul et al., 1990). A phylogenetic tree was constructed based on an alignment of selected sequences from GenBank across approximately 1500 nucleotide positions (Escherichia coli positions 27–1,510) using Neighbor-joining phylogenetic analysis with CLUSTALX (1.81) software.
24
Batch Experiments
A fresh culture of strain LYH1 at late log phase was inoculated into two batch bottles in an anaerobic glove box. Each batch bottle contained 125 ml autoclaved mineral medium with 2.5g/L glucose. The total volume of the bottle was 160 ml so that the headspace volume was 35 ml. Some of the bottles contained 0.625g L-cysteine to test its effect on growth of
LYH1 and its capability to produce hydrogen. The initial medium pH was adjusted to 6.2.
The batch bottles were anaerobically incubated at 30oC on a magnetic stir base. The amount of accumulated biogas produced was measured with a respirometer (Challenge
Environmental System AER-200, Fayetteville, AR). Headspace hydrogen was measured with a gas chromatograph equipped with a thermal conductivity detector and a molecular sieve column (Molesieve 5A 80/100 6’X1/8X0.085, Alltech) with nitrogen as the carrier gas.
A pH/ISE meter with Normal Hydrogen Electrode (Thermo Orion model 250A) was used to measure the oxidation-reduction potential of autoclaved medium after equilibration in an anaerobic glove box (Coy).
Oligonucleotide Probes
A specific probe for strain LYH1 (5’-CTACTGCTTCATGCGAAGCTGTA--3’) was developed from the alignment of its sequence with 17 nearly complete 16S rRNA sequences of related Clostridium spp., as well as by the CHECK-PROBE program from the
RIBOSOMAL Database Project II (Cole et al., 2003). The newly designed LYH1 probe
25
corresponded to E. coli positions 174-196 and it was confirmed to have mismatches with closely related Clostridium spp (Fig7). EUB (5’-GCTGCCTCCCGTAGGAGT-3’) was used as general bacterial probe (positive control) (Amann et al., 1990). EUB and LYH1 probes were purchased from Thermo (Bremen, Germany) whose 5’ ends were labeled with Cy3 and
Cy5, respectively. To confirm probe specificity, a cell suspension from the pure culture of
LYH1 was used as positive control, and C. pasteurianum, C. acidisoli and C. akagii and C. acetobutylicum were used as controls to check for nonspecific hybridization with close relatives.
Fluorescent In Situ Hybridization and quantification of LYH1 in bioreactor cultures
Samples for FISH analysis were taken from the bioreactor after steady state H2 production at 10g/L and 2.5 g/L glucose concentration at 10 h HRT. Cells were first fixed with 3 volumes of 4% paraformaldehyde in 10 mM sodium phosphate buffer (pH 7.2) for 30 minutes. The fixed cells were then spun down, washed with 1× phosphate buffered saline
(PBS) (130 Mm NaCl, 10Mm sodium phosphate, pH 7.2), spun again and resuspended in 1×
PBS. This step was repeated 2-3 times. 3µl-volumes of cell suspension were then spotted in wells of a 10-well Teflon-coated slide (Erie Scientific), air dried and dehydrated in ethanol gradients (50, 80, 95%). A moisture chamber (50-ml polypropylene conical tube) equilibrated with 0.9 M sodium chloride was used for hybridization. To each well, 16 ul of hybridization buffer (40% formamide, 0.9 M Nacl, 20 mM Tris-HCl (pH 7.2), 0.01% SDS) containing a total amount of 100 ng of Cy5-labeled LYH1 probe and Cy3-labeled bacterial
26
probe was applied. The chamber was incubated at 46oC for at least 1.5 hours. The slide was washed with buffer (56 mM NaCl, 20mM Tris-HCl (pH: 8.0), 0.01% SDS) at 48oC for 20 minutes, then rinsed with pure H2O and air dried. An antifadent was used to mount the slide for microscopic observation.
Cells smears were examined with an Olympus Fluoview 1000 confocal laser scanning microscopy (Olympus American Inc). Two excitation laser lines, red (633nm) and
Blue (488nm) were applied to detect Cy3 and Cy5 labeled probes respectively. Images were viewed with a UAPO 40X water objective (numerical aperture=1.15). The abundance of
LYH1 in bioreactor cultures was evaluated by enumeration of bacteria hybridizing to LYH1 probe in viewed microscope fields. For each bioreactor culture, at least 300 cells were enumerated in a total of 3 to 4 microscope fields.
RESULTS
Hydrogen Production
Biogas and hydrogen results from the bioreactor experiments that generated samples for this study of bacterial community composition were reported in van Ginkel and Logan
(2005) and are summarized in Appendix A (Table A-1). In their study, van Ginkel and
Logan concluded that lower glucose concentrations favored higher H2 yields. That is, when they operated the bioreactor with a feed containing 2.5g/L glucose, the highest H2 yields of
2.4-2.8 mol-H2/mol-glucose were observed. In contrast, H2 yields obtained with 10 g/L glucose ranged from 1.7-2.2 mol-H2/mol-glucose. Van Ginkel and Logan proposed that the reason they observed lower H2 yields with 10g/L glucose was because of higher hydrogen
27
partial pressure in the liquid medium which could inhibit hydrogen production via hydrogenase.
In our study, hydrogen production results differed from those observed by van
Ginkel and Logan (2005) . Data from our follow-up bioreactor trials using 10-h HRT and glucose concentrations of 10 g/L and 2.5 g/L are presented in Table 2-1. In the follow-up trial using 10 g/L glucose and 10-h HRT, the H2 production rate and yield were lower than in the study by van Ginkel and Logan (2005). In the latter study, the H2 production rate was 8.4 mL/min and the yield was 2.2 mol-H2/mol-glucos, whereas in our study, the H2 production rate was 5.4 mL/min and the yield was 1.4 mol-H2/mol glucose. For the 2.5 g/L glucose and
10-h HRT combination, van Ginkel and Logan observed H2 production rate of 2.7mL/min and a yield of 2.8 mol-H2/mol-glucose. In contrast, a lower H2 production rate of 0.09 mL/min and a lower yield of 0.1 mol-H2/mol glucose were observed in our trials. The factors which led to these different results appear be due to experimental variability.
RISA Fingerprints
Overall the compositions of bacterial communities in the bioreactor during steady- state H2 production were relatively simple, with RISA fingerprints containing two to four distinct bands per lane (Fig 2-1). In samples from the reactor containing 5, 7.5 and10 g/L glucose, similar RISA fingerprints were consistently observed. All of these RISA fingerprints consisted of two DNA bands, regardless of hydraulic retention time. Most 16S
28
rRNA sequences from these bands showed closest affiliation to Clostridium sp. FRB1 (Table
2-2). The RISA fingerprint for the 7.5 g/L glucose at 10h-HRT contained two bands with lengths of 1066 and 960 bp, which yielded DNA sequences most similar to Clostridium acetobutylicum ATCC 824 and Clostridium acidisoii, respectively (Table 2-2). In contrast,
RISA fingerprints of bioreactor samples grown with 2.5g/L glucose contained three to four
DNA bands that yielded 16S rRNA sequences similar to Selenomonas strain SB 90,
Enterobacter sp. B509, and other clostridia such as Clostridium acetobutylicum ATCC 824 and C. magnum, as well as Clostridium sp. FRB1 (Table 2-2). Low-nutrient conditions appeared to permit other populations to compete better with clostridia in the reactor. The appearance of non-sporeforming Enterobacter and Selenomonas spp. in the samples at
2.5g/L glucose concentration indicated that 2 hours of heat-shock treatment of soil had not been sufficient to destroy all vegetative cells.
Isolation of Dominant Culturable Hydrogen Producers
During incubation at 28oC under anaerobic conditions, no colonies were observed on medium without L-cysteine for up to 30 days. On plates containing L-cysteine, on the other hand, colonies appeared after 6 days. These white, opaque and convex colonies were of similar morphology, generally 1-2 mm in diameter (Fig 2-2). Thus L-cysteine permitted successful culturing of a population, designated strain LYH1, which had a RISA fingerprint of two DNA bands with 16S rRNA sequences having 100% similarity and which were also identical to sequences from bioreactor samples (Figure 2-3). 16S rRNA gene sequences
29
obtained from LYH1 indicated that they had 99% similarity to that of Clostridium sp. FRB1.
Phylogenetic analysis indicated that strain LYH1 belonged to subcluster Ic of Clostridium spp. (Fig 2-4).
The Capability of Strain LYH1 to Produce Hydrogen
In batch experiments with medium containing 2.5g/L glucose, strain LYH1 grew much better with L-cysteine than without. From 125-ml cultures grown with L-cysteine, the biogas production peaked at 93 + 4.3 ml, and the percentage of H2 in the biogas was 63 +
1.9% (means + standard deviations, n = 12). Without L-cysteine, peak biogas production from the same volume of culture was 56 + 7.8 ml, and the percentage of H2 was 55 + 3.6%
(means + standard deviations, n = 6). One Way Analysis of Variance (ANOVA) showed there were significant differences in both biogas production and hydrogen percentage between the two treatments at p<0.05 (Fig 2-5).
The lag time for outgrowth of strain LYH1 in medium containing 0.5g/L L-cysteine was also much shorter than in medium without L-cysteine. After two successive transfers, the lag time in medium with L-cysteine was reduced to less than 3 hours, while the lag time in medium without L-cysteine remained at greater than 50 hours. (Fig 2-6).
Therefore, the addition of L-cysteine had a substantial impact on the outgrowth of strain LYH1 and its production of hydrogen in liquid medium.
Fluorescent In Situ Hybridization
30
A specific probe was designed from existing rRNA sequences (Fig. 2-7) to assess the abundance of strain LYH1 in bioreactor samples after steady state H2 production was attained. The LYH1 probe labeled with Cy5 and universal EUB probe labeled with Cy3 were applied in hybridization tests to a mixed culture of Escherichia coli and strain LYH1. No cross hybridization was observed between strain LYH1 probe and E. coli cells (data not shown). In controls for specific hybridization, positive hybridization results between the
LYH1 probe and cells of LYH1 was observed. No cells in pure cultures of C. pasteurianum,
C. akagii, C. acetobutylicum, and C. acidisoli were observed to hybridize to LYH1 probe
(data not shown). Therefore, the designed LYH1 probe could be applied to distinguish LYH1 from its close relatives in FISH experiments. In samples from the bioreactor operated with
10 g/L glucose concentration and 10-h HRT, about 90% cells in microscope fields were observed to hybridize with the LYH1 probe (Fig 2-8). In bioreactor cultures grown with 2.5 g/L glucose, the proportion of LYH1 in viewed fields decreased to 26% (Fig 2-9). These results were consistent with RISA fingerprint analysis results and indicated that the bacterial communities in the reactors were affected by glucose concentration.
DISCUSSION
Bacteria considered to belong to the same genus and species have 16S rRNA gene sequences with >98% nucleotide identity (Bennasar et al. 1996). In our study, slight differences were noted in the 16S rRNA portions of cloned sequences from individual RISA bands affiliated with Clostridium spp., which was apparently due to the coexistence of closely related strains in the bioreactor (Table 2-2). The relatively greater diversity of
31
sequences detected in RISA profiles at the longest HRT (10 h) may have resulted from the way the bioreactor was operated based on the experimental design. For each glucose concentration, the bioreactor was operated in batch mode for three days after inoculation with the heat-treated soil. Once a steady-state of biogas production was achieved, the bioreactor was switched to continuous mode at an initial HRT of 10 h. After steady state sampling at 10-h HRT, the bioreactor was switched to 5h HRT, then to 2.5h HRT, and then
1h HRT without changing inocula between different HRTs. The long initial HRT of 10 h could have allowed persistence of low-abundance populations in the early stages of community succession. Other bioreactor studies by Cohen et al. (1985) have shown that growth of non-sporeforming propionate producers, such as Selenomonas sp., was promoted by discontinuous or semicontinuous feeding modes. Detection of species other than Cluster I clostridia under these conditions could indicate that other fermentative pathways were occurring in the bioreactor, which would affect H2 production.
The maximum hydrogen yield from fermentation is about 4 mol H2/ mol glucose when glucose is converted to acetate (Jones and Woods, 1986):
C6H12O6 + 2H2O → 2CH3COOH + 2CO2 + 4H2 (2)
Fermentation by other species, however, could result in other products. For example, the production of propionate from hexose actually involves the consumption of H2 (Vavilin et al., 1995):
C6H12O6 + 2H2 → 2CH3CH2COOH + 2H2O (1)
At higher glucose concentrations (7.5 and 10 g/L), only two DNA bands were observed in the RISA profiles and both sequences were related to Clostridium spp. In the
32
bioreactor fed with 5.0g/L glucose, similar RISA profiles with two DNA bands were obtained, although a sequence related to Selenomonas spp. was cloned from one of the bands. The bacterial community in the reactor operated with 2.5g/L glucose, however, was more diverse than at any other glucose concentration. The results showed that glucose concentration was a factor influencing the bacterial community composition. In this study,
5g/L appeared to be the lowest favorable glucose concentration for the competition of
Clostridium spp. with other bacterial species.
Kim et al. reported the effect of sucrose concentrations from 10-60 g/L on hydrogen production and microbial communities in a continuous fermenter inoculated with sludge as mixed microflora. They suggested that microbial community composition was the main factor contributing to different hydrogen yields under different sucrose concentrations (Kim et al., 2006). In their study on the effect of glucose concentration on hydrogen yield in a continuously-fed bioreactor inoculated with heat-treated soil, van Ginkel and Logan (2005) reported that hydrogen yield increased as glucose concentration decreased from 10g/L to 2.5 g/L. In our study, however, we hypothesized that lower hydrogen yields would be obtained from the bioreactor operated with 2.5 g/L glucose than with higher glucose concentrations because of the presence of non-sporeforming propionate producers (Selenomonas sp.) at the lower glucose concentration. This implied that other factors besides community composition could have contributed to the contrasting hydrogen yield observation made by van Ginkel and Logan. They hypothesized that the increased yield of hydrogen at 2.5g/L glucose was likely due to the inhibition of hydrogenase caused by the relatively higher hydrogen partial pressure in the liquid phase at high glucose loading rates. As hydrogen production from
33
fermentation is a thermodynamically unfavorable process, hydrogen yield will decrease if hydrogen partial pressure increases in the liquid phase due to the shift of metabolic pathways towards production of more reduced substrates such lactate, ethanol, acetone, butanol or alanine (Levin et al. 2004).
Interestingly, when the HRT of 2.5 g/L glucose bioreactor was shortened from 10 h to 2.5 h, a shift in dominant populations from Selenomonas spp. and Clostridium acetobutylicum to Enterobacter spp. and strain LYH1 was observed. As Selenomonas spp. are non-spore-forming propionate-producers and C. acetobutylicum is a solventogenic clostridial species, this result may partially explain why hydrogen production was higher at
2.5-h HRT than at 10-h HRT at the 2.5 g/L glucose concentration. The detection of non- spore-forming bacteria in our experiments indicated that some non-sporeforming bacteria in soil are heat resistant. Two-hours of dry-heat soil treatment at 105 oC was apparently not long enough to kill all vegetative cells.
Clostridia are gram-positive, spore-forming, rod-shaped bacteria with low or no tolerance to O2. They are frequently observed during dark fermentation processes for hydrogen production. For spore-forming bacteria, spore germination into fully active vegetative cells may require specific nutrients such as certain amino acids. L-Alanine is the most studied supplement for enhancing germination of spores of many species (Gould,
1970). L-cysteine is one of a limited number of related amino acids which can replace L- alanine as a germinant. Combined with the function of L-cysteine as an oxygen scavenger
(Song and Logan, 2004), it was hypothesized that the addition of L-cysteine would promote the growth of spore-forming and obligately anaerobic H2-producing clostridia. In this
34
experiment, we found that the addition of L-cysteine was an indispensable condition for obtaining colonies on solid media of the dominant H2-producing strain LYH1 in the bioreactor with 10 g/l glucose and 10-h HRT. At the lowest glucose concentration tested
(2.5g/L), strain LYH1 could grow much better in batch culture with 0.5g/L L-cysteine than in medium without L-cysteine, and it yielded a satisfactory hydrogen production, reaching approximately 2.0 mol H2/ mol glucose. This hydrogen yield is close to that obtained from the chemostat bioreactor (2.2 mol H2/ mol) operated with 10g/L glucose and 10 hr HRT (van
Ginkel and Logan, 2005). The H2 production in batch experiments provided further evidence that LYH1 was a dominant bacterial population in the bioreactor community grown with
10g/L glucose. Addition of L-cysteine to the enrichment medium appears to be useful for isolating these clostridia.
Although RISA is a useful molecular tool for the diagnosis of microbial community composition, it is a qualitative, rather than quantitative, community analysis approach. To quantify targeted populations in microbial communities, further analysis methods, such as real-time PCR and fluorescent in-situ hybridization must be applied (Gaval et al., 2002; Hall et al., 2003; Liao et al., 2004; Limpiyakorn et al., 2005; Kaetzke A, 2005; Zilouei et al.,
2006). In FISH experiments, 16S rRNA probes can be used to quantify different populations of interest in microbial communities. In secondary structure models for prokaryotic small subunit rRNAs described by Neefs et al. (1990), the 16S rRNAs of prokaryotes contain 8 variable regions and 48 helices. In the design of 16S rRNA probes that are relatively more specific and hybridizable to complementary sequences within 16S rRNAs, target sites located in variable and non-helix regions are usually preferred. Based on the recovery of
35
strain LYH1 on solid medium and its 16S rRNA sequence information, we developed a probe specific for strain LYH1 that was complementary to a non-helix-forming sequence in variable area 2 (V2) of its 16S rRNA. After the application of LYH1-specific probes labeled with Cy5, the images viewed with confocal laser scanning microscopy revealed that bacteria hybridizing to LYH1 probes comprised 90% of all cells observed in 10g/L bioreactor samples. When the bioreactor was fed with 2.5 g/L glucose, the bacteria hybridizing to the
LYH1 probe constituted only 26% of all cells examined. Our conclusions on the effect of glucose on bacterial community composition as judged by RISA were validated and extended by our experiments using FISH. Since visible DNA bands in RISA fingerprints arise only from populations that constitute at least 1% of total cell numbers in the community, RISA does not reveal the presence of populations with low abundance
(Borneman and Triplett, 1997). FISH, on the other hand, can detect populations of lower abundance (0.1-1%) depending on the numbers of microscope fields that are scanned.
Although cells hybridizing to the LYH1 probe were observed in 2.5 g/L bioreactor cultures operated at 10-h HRT, no excised RISA bands derived from the bioreactor samples yielded
DNA sequences that were identical to that of strain LYH1. Because of the need to extract
DNA for RISA analyses, differential cell lysis susceptibility of gram-positive and gram- negative bacteria may also contribute to differences observed between RISA and FISH.
Furthermore, differences in the number of rRNA operons in individual bacterial genomes will also affect the likelihood of PCR amplification in RISA analysis (Farrelly et al., 1995).
Clostridium acetobutylicum ATCC 824, for example, has 11 rRNA operons with identical
16S rRNA and spacer lengths of 178bp or 179 bp (Nolling et al., 2001), and rRNA
36
sequences related to C. acetobutylicum, but not LYH1, were detected in RISA fingerprints from bioreactor samples at 2.5 g/L glucose. This study demonstrates that RISA is a useful qualitative tool that can be used to develop probes based on specific sequences for quantitative analyses such as FISH.
ACNOWLEDGEMENTS
We thank Steven Van Ginkel for bioreactor operation, Sang Eun Oh for technical advice, and Dr John M. Regan for his guidance in fluorescent in situ hybridization (FISH) experiment. Microscope images were obtained in the Centre for Quantitative Cell Analysis in Penn State’s Huck Institute for the Life Sciences. This research was funded by NSF grant
BES 01-24674.
37
REFERENCES
1. Altschul SF, Gish W, Miller W, Myers EW and Lipman DJ (1990) Basic
local alignment search tool. J.Mol.Biol. 215:403–410
2. Amann RI, Binder BJ, Olson RJ, Chisholm SW. Devereux R and Stahl DA
(1990) Combination of 16 S ribosomal-RNA-targeted oligonucleotide probes
with flow-cytometry for analyzing mixed microbial population.
Appl.Environ.Microbiol 56: 1919-1925
3. Bennasar, A., Rossello-Mora, R, Lalucat, J. and Moore, ERB (1996)
16S rRNA gene sequences analysis relative to genomovars of Pseudomonas stutzeri
and proposal of Pseudomonas balearica sp nov.. Int.J.Syst.Bacteriol. 46:200-205.
4. Ballapragada, BS, Stensel, H.D, Puhakka, JA, Ferguson, JF (1997)
Effect of hydrogen on reductive dechlorination of chlorinated ethenes. Environ.
Sci. Technol. 31, 1728–1734.
5. Biebl, H (1999) Clostridium acetobutylicum. In Encyclopedia of food
microbiology; Robison,R., Batt, C., Patel, P., Eds.; Academic press: New York.
6. Benemann J (1996) H2 biotechnology: progress and prospects. Nat Biotechnol
14:1101–1103
7. Borneman J and Triplett EW (1997) Molecular microbial diversity in soils
from eastern Amazonia: evidence for unusual microorganisms and
micropopulation shifts associated with deforestation.
Appl.Environ.Microbiol. 63: 2647-2653
8. Billings RE (1991) The hydrogen world view, 1st ed. American Academy of
38
Science.
9. Chaudhary TV, Goodman DW (1999) Stepwise methane steam reforming: a
route to CO-free H2. Catal Lett 59:93–94
10. Chin HL, Chen ZS, Chou CP (2003) Fedbatch operation using Clostridium
acetobutylicum suspension culture as biocatalyst for enhancing H2 production.
Biotechnol Prog 19:383–388
11. Cohen A. Distel B, van Deursen A, van Andel JG (1985) Role of anaerobic
spore- forming bacteria in the acidogenesis of glucose-changes induced by
discontinuous or low-rate feed supply. A van Leeuw J Micrbiol. 51(2): 179-
192
12. Cole JR, Chai B, Marsh TL, Farris RJ, Wang Q, Kulam SA, Chandra S,
McGarrell DM, Schmidt TM, Garrity GM, Tiedje JM (2003) The
Ribosomal Database Project (RDP-II): previewing a new autoaligner that
allows regular updatas and the new prokaryotic taxonomy. Nucl.Acid.Res
31:442-443
13. DeWeerd, KA, Concannon, F, Suflita, JM ( 1991). Relationship between
hydrogen consumption, dehalogenation, and the reduction of sulfur oxyanions
by Desulfomonile tiedjei. Appl. Environ. Microbiol. 57, 1929–1934.
14. Distefano, TD, Gossett, JM, Zinder, SH (1992) Hydrogen as an electron
donor for dechlorination of tetrachloroethene by an anaerobic mixed culture.
Appl. Environ. Microbiol. 58, 3622–3629.
15. Fabiano B and Perego P (2002). Thermodynamic study and optimization of
39
hydrogen production by Enternobacter aerogenes. Int.J.Hyfrogen.Energy. 27:
149-59
16. Fang HHP, Zhang T, Liu H (2002a) Microbial diversity of a mesophilic
Hydrogen-producing sludge. Appl.Microbiol.Biotechnol 58:112-118
17. Fang HHP, Zhang T, Liu H (2002b) Characterisation of a hydrogen-
producing granular sludge. Biotechnol.Bioeng 78:44-52.
18. 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.Eviron.Microbiol. 61:2798-2801.
19. Gaval G, Duchene P, Pernelle JJ ( 2002) Filamentous bacterial population
dominance in activated sludges subject to stresses.Water Sci Technol. 46(1-
2):49-53.
20. Gould GW (1970) Germination and the problem of dormancy. J.Appl.Bact
33:34-49.
21. Hall SJ, Keller J, Blackall LL (2003) Microbial quantification in activated
sludge: the hits and misses. Water Sci Technol. 48(3):121-6.
22 Iyer P, Bruns MA, H Zhang, GS Van, Logan BE (2004) H2-Producing
Bacterial communities from a heat-treated soil inoculum.
Appl.Microbiol.Biotechnol 66:166-173
23. Jones DT and Woods DR (1986) Acetone-butanol fermentation revisited.
Micro. Rev. 50(4): 484-524
24. Kaetzke A, Jentzsch D, Eschrich K (2005) Quantification of Microthrix
40
parvicella in activated sludge bacterial communities by real-time PCR. Lett
Appl .Microbio.40(3):207-11
25. Kalia VC, Jain SR, Sharma CB, Joshi AP (1994) Fermentation of bio-waste
to H2 by Bacillus licheniformis. World J Microbiol Biotechnol 10:224–227
26. Karube I, Urano N, Matsunaga T, Suzuki S (1982) H2 production from
glucose by immobilized growing cells of Clostridium butyricum. Eur J Appl
Microbiol 16:5–9.
27. Kim SH, Han SK and Shin HS (2006) Effect of substrate concentration on
hydrogen production and 16S rDNA-based analysis of the microbial community
in a continous fermenter. Process Biochemitry 41:199-207.
28. Lane DJ (1991) 16S/23S rRNA sequencing. In: Stackebrandt E, Goodfellow
M (eds) Nucleic acid techniques in bacterial systematics. Wiley, New York,
pp 115– 175
29. Levin DB, Pitt L, Love M (2004) Biohydrogen production; prospects and limitations
to practical application. Int J Hydrogen Energy 29:173-185.
30. Lay, JJ (2001) Biohydrogen generation by mesophilic anaerobic fermentation
of microcrystalline cellulose. Biotechnol Bioeng. 74(4), 280-287
31. Lay JJ, Lee YJ and Noike T (1999). Feasibility of biological hydrogen
Production from organic fraction of municipal solid waste. Water Res:
33(11):2579-86.
32. Liao J, Lou I, de los Reyes FL 3rd (2004) Relationship of species-specific
filament levels to filamentous bulking in activated sludge. Appl Environ
41
Microbiol 70(4):2420- 8.
33. Limpiyakorn T, Shinohara Y, Kurisu F, Yagi O (2005) Communities of
ammonia- oxidizing bacteria in activated sludge of various sewage treatment
plants in Tokyo. FEMS Microbiol Ecol. 54(2):205-17.
34. Logan B E, Oh, S.-E, Kim I S and Van Ginkel S (2002) Biological
Hydrogen Production Measured in Batch Anaerobic Respirometers. Environ.
Sci.Technol. 36: 2530- 2535.
35. Montville T.J, Jones SB, Conway L.K and Sapers GM (1985) Germination
of spores from Clostridium botulinum B-aphis and Ba410.
Appl.Environ.Mcirobiol. 50(4): 795-800.
36. Neefs JM, Van de Peer Y, Hendriks L. and De Wachter R (1990)
Compilation of small ribosomal subunit RNA sequences. Nucleic Acids Res.
18 Suppl:2237-231.
37. Nolling J, Breton G, Omelchenko MV, Makarova KS, Zeng Q, Gibson R,
Lee HM, Dubois J, Qiu D, Hitti J, Wolf YI, Tatusov RL, Sabathe F,
Doucette-Stamm L, Soucaille P, Daly MJ, Bennett GN, Koonin EV, Smith
DR (2001) Genome sequence and comparative analysis of the solvent-
producing bacterium Clostridium acetobutylicum, J Bacteriol 183:4823-4838.
38. Rachman MA, Nakashimada Y, Kakizono T, Nishio N (1998)
Hydrogen production with high yield and high evolution rate by self-
Flocculated cells of Enterobacter aerogenes in a packed-bed reactor. Appl
Microbiol Biotechnol 49:450–454
42
39. Rittmann BE, Snoeyink V (1984) Achieving biological stable drinking water.
J.Am.Water Works Assoc 76: 106-114
40. Sambrook J, Frutsch EF, Maniatis T (1989) Moleculae cloning: A
laboratory manual, 2nd edn. Cold Spring Harbor Laboratory Press, Plinview,
N.Y.
41. Song YG, Logan BE (2004) Effect of O2 exposure on perchlorate reduction
by Dechlorosoma sp. KJ. Water.Res. 38(6):1626-1632.
42. Warner, KA, Gilmour, CC, Capone, DG (2002) Reductive dechlorination
of 2,4-dichlorophenol and related microbial processed under limiting and non-
limiting sulfate concentration in anaerobic mid-Chesapeake Bay sediments.
FEMS Microbiol. Ecol. 40, 165–195.
43. Vallin VA, Rytow SV, Lokshina LY (1995) Modelling hydrogen partial
pressure change as a result of competition between the butyric and propionic
groups of acidogenic bacteria. Bioresour Technol. 54:171-7.
44. Van Ginkel S, Sung S and Lay JJ (2002) Biohydrogen Production as a
Function of pH and Substrate Concentrate Concentration. Environ. Sci.
Technol. 35: 4726-4730.
45. Van Ginkel S and Logan B (2005) Increased biological hydrogen production
with reduced organic loading. Water Research. 39:3819-3826
46. Zilouei H, Soares A, Murto M, Guierysse B, Mattiasson B (2006) Influence
of temperature on process efficiency and microbial community response during
the biological removal of chlorophenols in a packed-bed bioreactor. Appl
43
Microbiol Biotechnol. 10:1-9
44
10g/L 7.5g/L 5g/L 2.5g/L (glucose) Marker 10h 5h 10h 5h 2.5h 1h 10h 5h 2.5h 1h 10h 5h 2.5h (HRT)
1 1 1 2 1353bp 1 2 1078bp 872bp 3 2 603bp 2 4 3 3
Figure 2-1: RISA profiles showing the bacterial community fingerprints of bioreactor
samples with different glucose concentrations and hydraulic retention times. The
glucose concentration and hydraulic retention time are indicated at the top. The sizes of
DNA marker bands in outer lanes are indicated by the arrows. The different DNA bands
in bioreactor samples are indicated with arrows in their respective lanes.
45
Table 2-1 Summary of hydrogen production from continuous flow bioreactor
Glucose HRT Total gas production H2 production H2 yield
(g/L) (h) (mL/min) (mL/min) (mol H2/mol glucose)
10 10 9.36 5.39 1.4
2.5 10 1.7 0.09 0.1
46
Table 2-2 Bacterial 16S rDNA sequence of RISA DNA bands derived from bioreactor cultures
a Sample Band Number Closest relative in GeneBank Percent simlarity with LYH1 16S (Glucose-HRT) (907-1540) 10g/L-10h 1 Clostridium sp. FRB1 (AY925092) 98.89 10g/L-10h 2 Clostridium sp. FRB1 (AY925092) 99.20
10g/L-5.0h 1 Clostridium sp. FRB1 (AY925092) 99.36 10g/L-5.0h 2 Clostridium sp. FRB1 (AY925092) 99.36
7.5g/L-10h 1 Clostridium acetobutylicum ATCC824 (AT007815) 93.17 7.5g/L-10h 2 Clostridium acisisoli strain CK74 (AJ237756) 96.34 7.5g/L-5.0h 1 Clostridium sp. FRB1 (AY925092) 99.36 7.5g/L-5.0h 2 Clostridium acetobutylicum ATCC824 (AT007815) 95.54
7.5g/L-2.5h 1 Clostridium acetobutylicum ATCC824 (AT007815) 96.17 7.5g/L-2.5h 2 Clostridium sp. FRB1 (AY925092) 99.20
7.5g/L-1.0h 1 Clostridium acetobutylicum ATCC824 (AT007815) 94.91 7.5g/L-1.0h 2 Clostridium sp. FRB1 (AY925092) 99.36 5.0g/L-10h 1 *Selenomonadaceae strain SB90 (AJ229242) 81.73 5.0g/L-10h 2 Clostridium sp. FRB1 (AY925092)
5.0g/L-5.0h 1 Clostridium sp. FRB1 (AY925092) 98.25 5.0g/L-5.0h 2 Clostridium sp. FRB1 (AY925092)
5.0g/L-2.5h 1 Clostridium sp. FRB1 (AY925092) 99.52 5.0g/L-2.5h 2 Clostridium sp. FRB1 (AY925092) 5.0g/L-1.0h 1 Clostridium sp. FRB1 (AY925092) 99.01 5.0g/L-1.0h 2 Clostridium sp. FRB1 (AY925092)
2.5g/L-10h 1 *Selenomonadaceae strain SB90 (AJ229242) 81.89 2.5g/L-10h 2 *Selenomonadaceae strain SB90 (AJ229242) 82.36 2.5g/L-10h 3 Clostridium acetobutylicum ATCC824 (AT007815) 93.64 2.5g/L-10h 4 *Selenomonadaceae strain SB90 (AJ229242) 91.91 2.5g/L-5.0h 1 *Selenomonadaceae strain SB90 (AJ229242) 82.01 2.5g/L-5.0h 2 Clostridium magnum strain DSM (X77835) 94.11 2.5g/L-5.0h 3 *Selenomonadaceae strain SB90 (AJ229242) 81.89 2.5g/L-2.5h 1 *Enterobacter sp B509 (AB049108) 81.48 2.5g/L-2.5h 2 Clostridium sp. FRB1 (AY925092) 97.45 2.5g/L-2.5h 3 Clostridium sp. FRB1 (AY925092) 99.68
a Accession number of the closest relative indicated in perenthesis Asterisk ( ) mark non-Clostridium bacterial species Dash (-) indicated the sequences not available
47
Fig 2-2: Petri plate after 6 days of incubation on anaerobic medium containing
L-cysteine and 10g/L glucose, showing colonies of H2 producing bacteria.
48
Marker Isolate
1161bp 1353bp 1078bp 872bp 980bp 603bp
Fig 2-3: RISA fingerprint of genomic DNA of isolate from 10g/L glucose culture.
The sizes of DNA bands are indicated by the arrows
49
C.FRB1ClostridiumC.FRB1 AY925092 (AY925092)sp. FRB1 (AY925092) 64 C.pasteurianum M23930( ) 100 C.pasteruianumClostridium pasteurianum (M23930)(M23930)
LYH1 59 Strain LYH1 Cluster IC ( ) 100 C.acidisoliClostridiumC.acidisoli CK74acidisoli (AJ237756) AJ237756(AJ237756)
C.akagiiClostridiumC.akagii( AJ237755 (AJ9237755) akagii )(AJ9237755)
C.acetobutylicumClostridiumC.acetobutylicum acetobutylicum ATCC 824 ATCC AE0ATCC 824 824(AE007543) (AE007543) 99 Cluster Ib 61 C.collagenovoransClostridiumC.collagenovorans collagenovorans DSM 3089 X7 DSMDSM 3089 3084 (X7343) (X7343)
C.subterminaleClostridiumC.subterminale subterminale X68451 (X68451)(X71849) Cluster I 100 57 Cluster If C.argentinenseC.argentinenseClostridium X68316 argentinense (X68316)(X68316)
C.cellulovorans X71849 C.cellulovoranClostridium cellulovoran (X71849)(X71849) Cluster Ia 58 100 C.beijerinckiiClostridiumC.beijerinckii DSM beijerinckii 791 X68179 DSM DSM791 (X68179) 791 (X68179)
C.sporogenesClostridiumC.sporogenes sporogenes X68189 (X68189)(X68189) Cluster Id 100 78 ClostridiumC.tetani X74770 (X74770) tetani (X74770) Cluster Ie
C.proteolyticumClostridiumC.proteolyticum proteolyticum DSM 3090 DSM X744 DSM3090 3090 (X73448) (X73448) Cluster II C.cellulosiClostridiumC.cellulosi L09177 cellulosi (L09177)(L09177) Cluster IV 100 C.thermocellumClostridiumC.thermocellum thermocellum DSM 1237 L0917 DSM DSM1237 1237(L09173) (L09173) Cluster III
BacillusBacillus subtilis subtilissubtilis AF549498(AF549498) (AF549498) 0.01
Fig 2-4: Neighbor-joining dendrogram depicting phylogenetic relationships of strain LYH1 with
other Clostridium spp. based on nearly complete 16S rRNA sequences, using Bacillus subtilis as out group. GenBank accession numbers are in parentheses. The numbers at the nodes represent percentage bootstrap values of 1,000 replicates. The scale bar represents the calculated number of changes per nucleotide position
50
A Mean of accumulated gas production (ml) B Mean of H2 percent in headspace of batch bottles
120 70
100 60 50 80 Sample with 0.5g/L L- Sample with0.5 g/L L- cysteine 40 cysteine 60 Sample without L- 30 Sample without L- 40 cysteine cysteine
H2 percent (%) 20 Gas production (ml) production Gas 20 10
0 0
Fig 2-5. Comparison of biogas production (A) and H2 percentage in biogas (B) in samples
grown with 0.5g/L L-cysteine and without L-cysteine. Column heights represent means and
error bars indicate standard deviations.
120 51
100 80 LYH1 culture with 0.5 g/L L-cysteine 60 LYH1 culture without L-cysteine 40
productionGas (ml) 20
0 0 1020304050607080 Time (hour)
120
100
80 LYH1 culture with 0.5 g/L L-cysteine 60 LYH1 culture without L-cysteine
40
productionGas (ml) 20 0 0 102030405060708090 Time (hour)
100
80 LYH1 culture with 0.5 g/L L-cysteine 60 LYH1 culture without L-cysteine
40
Gas production (ml) 20
0 01020304050607080
Time (hour)
Figure 2-6. Accumulated biogas production during growth of strain LYH1 in batch cultures with L-cysteine (squares) and without L-cysteine (diamonds)
52
16 S rRNA sequences Target 16S rRNA positions ( E.coli numbering) Strain LYH1 probe 3’-ATGTCGAAGCGTACTTCGTCATC-5’ Strain LYH1 target 5’-TACAGCTTCGCATGAAGCAGTAG-3’ 174-196 Clostridium sp. FRB1 5’-TAAAGCTTCGCATGAAGCATTAA-3’ 173-195
C.pasteurianum strain CK74 5’-TAAAGTTTCACATGGAGCTTTAA -3’ 174-196
C.akagii 5’-TATAGCTTCGCATGGAGCCATAA -3’ 173-195
C.acidisoli 5’-TGTACGTTCGCATGAAGCAATAA-3’ 173-195
C.novyi 5’-TATTTTACGGCATCGTAGAATAA-3’ 181-203 C.sporogenes 5’-AAGAGAATCGCATGATTTTCTTA-3’ 181-203 C.acetobutylicum ATCC 5’-TCGAGAATCGCATGATTCTTGAG-3’ 171-193 C.collagenovorans 5’-ATTTTGATCGCATGGTCGAAATA-3’ 179-201
C.cellovorans 5’-GCAATTCTCGCATGAGAGATGTA-3’ 175-197
C.tyrobutylicum 5’-CA_AGTTTCACATGGAATTTGGA -3’ 178-200
C.subterminale 5’-ATGAAGGTCGCATGACTTTTATA-3’ 184-206
C.argentinense 5’-ATGAAGGTCGCATGACTTTTATA-3’ 181-204
C.tetani 5’-TAAGGTTTCGCATGAAACTTTAA-3’ 182-204
C.cellulosi 5’-ACGTTGGAGGCATCTCCGATGTA-3’ 179-201
C.thermocellum_DSM1237 5’-AACGGGGCGGCATCGTCCTGTTA-3’ 184-206
C.proteolyticum_DSM309 5’-CTAATTGCCGCATGGCAGATAGA-3’ 177-200
C.beijirinckii_DSM791 5’-TGTTAGTGCCGCATGGCATAGCA-3’ 179-201 Selenomonadaceae strain SB90 5’-TTGAGAAGCGCATGNTTTTCAA_-3’ ------
Enterobater cloacae stain CP1 5’-CGCAAGACCAAAGAGGGGGACCT-3’ 183-205
B.subtilis 5’-GTTTGAACCGCATGGTTCAAACA-3’ 182-204
Dash (-) indicated the numbering of target 16S rRNA positions is not available.
Fig 2-7: Alignments of the LYH1 probe sequence, its target site, and sequences of
corresponding sites in reference organisms. Sequences from 17 related Clostridium spp.,
Selenomonas sp.strain SB90, Enterobacter cloacae strain CP1 and Bacillus subtilis were
selected as reference organisms. The target sequence is displayed in the upper row;
mismatches between target and sequences of corresponding sites in reference organisms are
shadowed.
53
Fig 2- 8: Confocal Scanning Laser Microscope images of sample culture at
10g/L glucose concentration with 10-h HRT: (A) The hybridization of sample cells with universal EUB probe labeled by Cy3 and LYH1 probe labeled by
Cy5. (B) Morphology of sample cells as shown by differential interference contrast.
A B
54
A B
A B
55
Fig 2-9: Confocal Scanning Laser Microscope images of sample culture at
2.5g/L glucose concentration with 10-h HRT: (A) The hybridization of sample cells with universal EUB probe labeled by Cy3 and LYH1 probe labeled by Cy5.
(B) Morphology of sample cells examined by confocal scanning laser microscopy
A B
56
A B
A B
57
Chapter 3
The effect of L-cysteine on bacterial community composition and H2 production in batch cultures of heat-treated soil inocula
Luo Y.H., Zhang H., Logan B.E. and Bruns M.A.
(To be submitted for publication in Journal of Bioscience and Bioengineering)
58
ABSTRACT
The effect of L-cysteine on H2 production and bacterial community composition in batch systems containing artificial wastewater with varied glucose concentrations.was tested with inocula from heat-treated soil. Addition of L-cysteine (0.5g/L) resulted in a reduction of
Eh of the medium from +93 (±11) to -84.75(±4.57) mV. PCR-based ribosomal RNA intergenic spacer analysis (RISA) was used to characterize the bacterial community composition in each treatment combination through 5 consecutive transfers to determine effect on lag time and community composition. Regardless of the glucose concentration and number of serial transfers, all RISA profiles of bacterial communities grown without L- cysteine were identical, consisting of three DNA bands that yielded partial 16S rRNA sequences most closely related to Enterobacteriaceae. In contrast, RISA profiles of batch cultures with L-cysteine varied with glucose concentration. At the lowest glucose concentration (2.5g/L), RISA profiles were similar to those of batch cultures without L- cysteine, and the second, fourth, and sixth serial batch cultures had the same profiles. At the
5g/L glucose concentration, RISA band patterns shifted between the second and fourth cultures but remained the same for the fourth and sixth serial cultures, yielding two bands with sequences most related to clostridia. At glucose concentrations of 7.5 and 10g/L, all serial cultures had RISA profiles with these same two bands. In all cultures, the addition of
L-cysteine was associated with higher amounts of accumulated biogas and hydrogen.
Compared to 0.76-1.37 mol-H2/mol-glucose hydrogen yields in batch systems dominated by
59
Enterobacter spp or diverse bacterial populations, higher hydrogen production was obtained in clostridia-dominated cultures, which ranged from 1.18 mol-H2/mol-glucose to 1.89 mol-
H2/mol-glucose.
60
INTRODUCTION
Bioproduction of hydrogen can be attained by the dark fermentation of carbohydrate-rich wastewater by H2-producing bacteria. Sewage sludges and heat-treated soils have been used as sources of microflora for hydrogen production (Chen et al. 2001,
Ginkel et al., 2001, Hallembeck and Benemann, 2002, Nielsen et al., 2001, Oh et al., 2004,
Park et al., 2005;). To date, pure cultures known to produce hydrogen from carbohydrates include species of Enterobacter (Fabiano and Perego 2002, Hussy et al. 2003, Rachman et al. 1997), Bacillus (Kalia et al. 1994), and Clostridum (Taguchi et al 1995, Yokoi et al.
1997). Among them, Clostridium spp. have been frequently observed in H2-producing mixed cultures from environmental samples and grown in non-sterile feedstocks (Fang and Liu
2002a, Fang and Liu 2002b, Lay et al. 1999) or synthetic wastewater( Iyer et al. 2004).
Therefore Clostridum spp. were presumed to be one of key H2-producing populations in these systems. In general the yield of H2 from Clostridum spp. is higher than that from the facultatively anaerobic Enterobacter spp. (Yokoi et al. 1998).
In our previous work, success in culturing Clostridium spp. from a bioreactor on solid medium required the addition of L-cysteine. The addition of reducing agents such as L- cysteine has long been known to enhance the growth of anaerobic organisms by acting as a scavenger of O2. This has been demonstrated with H2-producing clostridia (Yokoi et al.
1998). For example, low oxidation-reduction potential with Eh of -250 mV or less is required for adequate growth of solvent-producing clostridia (Bahl and Durre 2001). The addition of
L-cysteine has been reported to enhance spore germination of clostridia (Gould, 1970;
61
Montville et al., 1985). On the other hand, sufficient substrate is another factor required for activation of spore-forming bacteria (Tortura et al., 1998; Mitchell, 2001). The concentration of substrates like glucose may affect biohydrogen production by influencing outgrowth of specific populations and the resultant composition of microbial communities.
Although biohydrogen production from mixed cultures is influenced by carbon source concentration, dissolved H2, temperature, hydraulic retention time (HRT) and pH
(Chang and Lin 2004, Mizuno et al. 2002, Kim et al. 2004, Ruzicka, 1996, Van Ginkel and
Logan 2005), only a few papers have reported the effect of these external factors on the composition of microbial communities derived from mixed inocula (Iyer et al. 2004, Kim et al. 2006). Little information is available on the relationship between hydrogen production and the composition of microbial communities under different growth conditions. The objective of this study was therefore to determine the effects of serial transfers, addition of
L-cysteine, and glucose concentration on hydrogen production and composition of bacterial communities in batch cultures. If the clostridia represent an important group of efficient hydrogen producers, it is important to understand how to optimize their growth from mixed inocula.
MATERIALS AND METHODS
Inoculum
Soil was sampled to 5 cm depth from harvested tomato plots in early October 2004 at the Penn State University Park campus. This soil is a Hagerstown silty clay loam (fine,
62
mesic, typic hapludalfs). Soil was air-dried, ground with a sterile mortar-and-pestle, screened to obtain fine particles, and thoroughly mixed. Soil was then heated in a shallow layer at
105oC for 2 h to reduce numbers of vegetative microorganisms, particularly methanogens that would consume hydrogen (Biebl 1999, Lay 2001, Logan et al. 2002, Van Ginkel et al.
2002;). Aliquots of treated soils were distributed into sterile tubes for storage at -70oC. When a tube was removed from storage, any remaining soil after inoculation was discarded to ensure that soils were subjected to only one freeze-thaw cycle.
Culture Conditions and Batch Operation
Batch bottles of mineral media were prepared with and without L-cysteine at four different glucose concentrations. The medium used in batch experiments contained the following (g/L): NH4Cl (0.88); K2HPO4 (0.44); KH2PO4 (0.44); MgSO4.7H2O (0.36); FeCl3
(0.05); NiCl2.6H2O (0.0476); CaCl2.2H2O (0.0662); ZnCl2 (0.023); CoCl.6H2O (0.021);
CuCl2 (0.01); MnCl2 (0.03). The initial pH of the medium was adjusted to 6.2 with NaOH, and MOPS pH buffer (C6H13NO4S.H2O(2-(N-morpholino)ethanesulfonic.acid, monohydrate) at a final concentration of 12g/L. The medium (with or without 0.5 g/L L-cysteine) and the stock glucose solutions (200 g/L) were autoclaved separately. The total volume of a batch bottle was 160 ml and each bottle was fitted with a screw cap containing a butyl rubber insert. The combined volumes of medium plus subculture transfer was 125ml medium, so that the headspace volume above liquid was 35 ml. All components were placed in an anaerobic glove box to maintain anaerobiosis during media preparation, initial inoculation,
63
and subsequent serial transfers. Autoclaved glucose solution was pipetted aseptically into the sterile medium to achieve glucose concentrations of 10, 7.5, 5.0, and 2.5 g/L. For both series of batch bottles (with or without L-cysteine), 0.5g heat-treated soil was inoculated into the first batch bottle, which was incubated at 30°C outside the anaerobic chamber on a magnetic stir base. Accumulated biogas in the headspace above the culture was measured continuously after inserting a sterile needle through the rubber insert of the cap and connecting the needle to a Challenge AER-200 respirometric system (Challenge Environmental Systems,
Fayetteville, AR). When no further biogas accumulation was observed, the first culture was taken into the anaerobic chamber for transfer to a fresh bottle of medium. Five serial transfers (1:10) were made using 12.5 ml culture transferred to 112.5 ml fresh medium when the maximum accumulated biogas at each glucose concentration was observed.
DNA Extraction
After a transfer was made to fresh medium, the culture was dispensed into 50-ml polypropylene tubes for frozen storage at -80oC. Thawed cultures were centrifuged at 7900x g for 20 min at 4oC (AvartiTM J-25 Beckman Coulter) to obtain a cell pellet. Most of the supernatant was discarded except for 2 ml left to resuspend the pellet. Resuspended cells were homogenized by vortexing. The second, fourth and sixth batch cultures were selected for DNA extraction and RISA analysis. The cell lysis and DNA extraction were carried out with the UltraClean Microbial DNA Isolation Kit protocol (Mo Bio), and genomic DNA samples were stored at -20°C until analysis.
64
RISA and Cloning of PCR Product
The bacteria-specific 16S rRNA primer set 926f (16S rRNA) (5’-
AAACTYAAAKGAATTGACGG-3’) and 115r (23S rRNA) (5’-GGGTTBCCCCATTCGG-
3’) was used to amplify purified DNA (Lane 1991). The reaction volume of each PCR mixture was 50 ul and contained PCR buffer (VWR), 2.5 U Taq DNA polymerase (VWR),
30pmol each primer, 250 uM each deoxyribonucleoside and 40 ng DNA template. The PCR was performed in a GeneAmp PCR system 9600 (Perkin Elmer) according to the following cycles: 1 cycle at 94оC for 5 min; 30 cycle at 94 оC for 30 s, 54 оC for 30 s, and 72 оC for 1 min; and 1 cycle at 72 оC for 7 min. PCR-amplified DNA products were separated by electrophoresis in 0.7% agarose gels. The gel was prepared and run in 1x TAE buffer (pH
8.0). After 5 hours of electrophoresis at 45 V, the gel was stained with ethidium bromide and imaged with an Epi Chemi II Darkroom unit (UVP Laboratory Products, Upland, CA). DNA fragment lengths in RISA gel bands were calculated from migration distances of DNA fragments in marker lanes. Based on RISA profile results for samples without L-cysteine, the sixth serial batch culture at 2.5g/L glucose was selected for DNA band excisions and sequence analysis. For samples with L-cysteine, the sixth batch culture at 2.5g/L glucose concentration, the second, fourth and sixth batch cultures at 5g/L glucose concentration, the sixth batch culture at 7.5g/L glucose concentration, the fourth and sixth batch cultures at
10g/L glucose concentration were selected for RISA band sequence analysis. DNA bands were excised and eluted from the gel by the crush and soak procedure (Sambrook et al.
65
1989). Cloning of eluted DNA followed the TOPO TA Cloning kit protocol (Invitrogen).
The plasmids containing RISA DNA fragments were purified by using QIA prep Spin Mini prep kit (QIAGEN) as described by the manufacturer. Sequencing was performed with an
ABI Hitachi Genetic Analyzer. At least two clones for each DNA bands were sequenced.
Closest affiliations for 16S rRNA gene sequences were obtained with the Basic Local
Alignment Search Tool (BLAST) (Altschul et al., 1990).
Chemical Assays
When respirometry readings indicated that cultures exhibited no further increases in accumulated biogas, headspace gas samples were collected using a Gas-tightTM syringe
(Hamilton, Reno, NV) and injected into a gas chromatograph (GC; Model 310, SRI
Instruments, Torrence, CA) equipped with a thermal conductivity detector and a molecular sieve column (Molesieve 5A 80/100 6’X1/8X0.085, Alltech). Nitrogen was used as the carrier gas when determining composition of injected gas samples. The pH of the culture was measured with a pH meter (Accumet model 10, Fisher Scientific). Oxidation-reduction potential was measured with a pH/ISE meter with Normal Hydrogen Electrode (Thermo
Orion model 250A). The initial oxidation-reduction potential of the medium was measured in an anaerobic glove box after the medium was autoclaved and cooled to ambient temperature.
66
RESULTS
RISA Fingerprints
Upon addition of L-cysteine, the oxidation-reduction potential of the culture medium decreased from Eh of +93(+11) mV to Eh of -84.75(+4.57) mV (mean and standard deviation of n = 4). This initial reduction in Eh appeared to influence the outgrowth of specific populations from the mixed inoculum. In cultures grown without L-cysteine, identical RISA patterns with 3 bands were obtained (ca 1200, 1100, and 990 bp), regardless of the glucose concentration and the number of serial transfers (Fig 3-1). Cloned DNA from all three RISA bands yielded sequences with 99% similarity to Enterobacter sp. B509.
The RISA profiles of cultures grown with L-cysteine, however, did vary with glucose concentration and number of serial transfers (Fig 3-2). In serial batch cultures grown with 2.5 g/L glucose, RISA profiles were similar to those of cultures grown without L- cysteine (three bands at ca 1200, 1100, and 990 bp). Cloning of all three DNA bands yielded sequences related to Enterobacter sp. B901-2 (98-99% similarity). No differences were observed in RISA profiles in the second, fourth, or sixth batch cultures in 2.5 g/L glucose.
For cultures growth with 5 g/L glucose, RISA profiles varied with the number of serial transfers. While the second serial culture yielded a RISA profile with three bands similar to those of cultures grown without L-cysteine, the RISA profiles of the fourth and sixth cultures each contained only two DNA bands (ca. 1120 and 960 bp). Both bands from each of these
RISA profiles yielded cloned DNA sequences that were most closely related to C. acetobutylicum. For cultures grown with 7.5g/L and 10g/L glucose, identical RISA profiles
67
were obtained and consisted of two bands yielding sequences related to Clostridium acetobutylicum and Clostridium sp. FRB1. These results indicated that outgrowth of
Clostridium spp. was favored by higher glucose concentrations in the presence of L-cysteine.
Biogas Production
In batch cultures without L-cysteine, as glucose concentration increased from 2.5 g/L to 7.5 g/L, the total accumulated gas produced in batch cultures increased. However, no further rise was observed when glucose concentration was 10 g/L (Fig 3-3), which was likely due to the final pH dropping to 4.9. In batch cultures with L-cysteine, biogas production also increased with the increase in glucose concentration. The highest gas production was observed with 10g/L glucose in batch cultures with L-cysteine though their final pH dropped to below 4.9 (Table 3-1). Whereas in batch cultures without L-cysteine, the highest biogas production occurred in 7.5 g/L glucose. One Way Analysis of Variance (ANOVA) showed biogas production was increased significantly in batch cultures dominated by Clostridium spp. compared to those dominated by Enterobacter spp. (p<0.05) (Fig 3-4), although the percentages of H2 in the biogas were similar in both types of cultures.
To check whether methanogens were consuming H2 in the 2.5 g/L glucose medium without L-cysteine, the percentage of H2 in biogas was measured periodically. The percentage of H2 in biogas stayed constant (62-64%), up to 9 hours after gas production ceased (Fig 3-5), which indicated that H2 was not being consumed.
68
Hydrogen yields from batch cultures
Theoretical optimal hydrogen yield from the fermentation of glucose is 4 mol-
H2/mol-glucose (Jones and Woods 1986). In our study, hydrogen yield was calculated for each batch culture and graphed in Fig 3-6 and Fig 3-7. After five transfers in media without
L-cysteine, the hydrogen yield was higher in 2.5 g/L glucose (1.12 ± 0.16 mol-H2/mol- glucose) than that in 10 g/L glucose (0.82 ± 0.06 mol-H2/mol-glucose) (Fig 2-6). The converse was true for cultures grown in medium with L-cysteine, where hydrogen yields for
2.5 g/L and 10 g/L glucose were 1.04 ± 0.39 mol-H2/mol-glucose and 1.81 ± 0.13 mol-
H2/mol-glucose, respectively (Fig 2-7). Lag times were reduced significantly upon serial transfer in either type of medium. Lag times for the first batch cultures were 50 h and 100 h for media with and without L-cysteine, respectively. After two transfers, lag times were reduced significantly to less than three hours at all four glucose concentrations.
Whether or not L-cysteine was added to the medium, cultures grown in 2.5 g/L glucose had similar hydrogen yields, which was about 1.1 mol-H2/mol-glucose. It is reasonable to propose that such similarity in hydrogen yields could be attributed to the similarity in community compositions, because both these communities were dominated by
Enterobacteraceae. On the other hand, at glucose concentrations higher than 2.5 g/L, hydrogen yields differed in media with and without L-cysteine, as did community compositions. Since the former communities were dominated by Clostridium spp., it is reasonable to propose that these populations gave rise to the higher hydrogen production.
69
DISCUSSION
Kim et al. investigated the relationship between hydrogen production and bacterial communities in a continuous-flow stirred-tank reactor (CSTR) with sucrose concentrations of 10-60 g COD/L (Kim et al., 2006). In their work, the highest hydrogen yield was obtained from the 30 g/L sucrose bioreactor. The decrease of hydrogen yield at other sucrose concentrations was attributed either to acetogens, which were detected in the bioreactor cultures at sucrose concentrations below 30 g/L, or to lactic acid bacteria, which were detected at sucrose concentrations above 30 g/L. In our study, hydrogen production was also found to be closely associated with the composition of bacterial communities in the cultures.
In the batch experiments, it was found that all RISA profiles of bacterial communities grown without L-cysteine were identical and dominated by Enterobacteriaceae. In contrast, RISA profiles of batch cultures grown with L-cysteine varied with glucose concentration. The
RISA profile of the batch culture with L-cysteine in 2.5 g/L glucose was similar to that of cultures grown without L-cysteine. Therefore, in 2.5 g/L glucose cultures, it was not surprising that the average hydrogen yield (1.04 ± 0.39 mol-H2/mol-glucose) was similar to that in medium without L-cysteine (1.12 ± 0.16 mol-H2/mol-glucose). In the second batch culture at 5 g/L glucose, one of the 16S rRNA sequences obtained from RISA DNA bands has 98% similarity to Selenomonas strain SB 90 and others were affiliated with Enterobacter spp. Selenomonas strain SB 90 is a non-spore-forming propionate-producer and was previously classified as a member of Clostridium cluster IX (Chin et al. 1999). Fermentation of glucose to propionate produces no hydrogen. On the other hand, Clostridium spp. are generally known to be capable of producing more hydrogen than the facultative anaerobes,
70
Enterobacter spp. Hence, in the cultures grown with L-cysteine, the hydrogen yield in the second batch culture at 5.0 g/L glucose was relatively low compared to those obtained from the batch cultures at 7.5 and 10 g/L glucose, reaching 1.43 mol H2/mol glucose. The average hydrogen yield (1.81± 0.13 mol H2/mol) in the 10g/L batch cultures was higher than that
(1.60 ± 0.30 mol H2/mol) in the 7.5 g/L batch cultures. One of the 16S rRNA sequences derived from RISA DNA bands in 10 g/L batch cultures was related to Klebsiella sp. KG1, and no sequences of this type were detected in 7.5 g/L batch cultures. Klebsiella spp. are known to include fermentative H2-producers (Brosseau and Zajic 1982; Kumar and Vatsala,
1989; Kumar and Das, 2001; Solomon et., 1995). Because RISA is a qualitative molecular community approach, it cannot be used to indicate relative population abundance. Thus, there may have been slight differences between the compositions of bacterial communities in the 7.5 and 10 g/L batch cultures that were not detected by RISA and which consequently gave rise to different hydrogen yields. In our experiments, no consumption of hydrogen caused by acetogens or methanogens was observed, which implied that methanogens were effectively eliminated from soil inocula by this heat treatment method. However, the presence of other non-spore-forming bacteria like Selenomonas spp. and Enterobacter spp. in our study indicated that additional heat treatment could be tried to kill all vegetative cells in inocula in future work.
Since Enterobacter spp. are facultative anaerobes while Clostridium spp. and
Selenomonas spp. are obligate anaerobes, the different community compositions observed in cultures with and without L-cysteine might reflect different redox requirements or sensitivities to O2. In a study of the effect of oxygen and oxidation-reduction potential on
71
growth of C. perfringens, it was found that aerobic conditions inhibited C. perfringens in both negative and positive ranges of oxidation-reduction potential (Walden and Hentges,
1975). If the oxygen was replaced by another electron acceptor, such as potassium ferricyanide, growth occurred in the absence of oxygen at either negative or positive oxidation-reduction potential. Therefore Walden and Hentges (1975) demonstrated that oxygen, rather than oxidation-reduction potential, was responsible for growth inhibition. In another experiment, cocultures of Enterobacter aerogenes were effective in eliminating oxygen and facilitating the growth of Clostridium butyricum and its H2 production from starch (Yokoi et al., 1998). Moreover, elimination of O2 in co-cultures with Enterobacter sp
B901-2 induced the N2-fixing activities of Clostridium sp. strain B901-1b, which requires strict anaerobic conditions (Minamisawa 2004). Thus elimination of trace O2 is the key to growth of obligately anaerobic clostridia and their physiologic metabolisms. In our experiments, L-cysteine worked as a rapid oxygen scavenger. We hypothesized that the removal of oxygen in medium by the addition of L-cysteine improved the competitive ability of oxygen-sensitive anaerobic bacteria with other bacterial species in batch experiments. On the other hand, spore germination is an indispensable step for the outgrowth of spore- forming bacteria. It is a multistep process involving an initial trigger reaction where the germinant binds to an active site on the spore wall (Duncan and Foster, 1968; Smoot and
Pierson, 1982). Therefore, L-cysteine could also have influenced community composition by promoting spore germination (Montville et al., 1985).
In our previous work, glucose concentrations affected the composition of bacterial community composition in continous flow reactors (Luo et al., manuscript in preparation). In
72
the present study using batch cultures containing L-cysteine and 5 g/L glucose, the dominant
Enterobacter spp. appearing in the second serial culture were gradually replaced by
Clostridium spp. in the fourth and sixth cultures. As observed in the bioreactor study, 5g/L glucose appeared to be a lower limit for rapid growth of Clostridium spp. with capacity for
H2 production under these conditions. When glucose concentration exceeded 5g/L, the bacterial communities in batch cultures with L-cysteine were all dominated by Clostridium spp. Differences in community composition between media with and without L-cysteine at
7.5 g/L and 10 g/L glucose were associated with different peak gas production and H2 yields from glucose. Below 5.0 g/L glucose, the bacterial communities were dominated by
Enterobacteraceae. The only common feature shared by cultures grown with and without L- cysteine was the greatly reduced lag time following the second serial transfer accompanied by an increase in hydrogen production. This phenomenon appeared to be related to acclimation of bacteria to the medium.
The awareness of the effects of oxidation-reduction potential and glucose concentration on community composition offers insights to manipulating the fermentation conditions for H2 production to make it more optimal. The proper adjustment of culture parameters that favor the dominance of bacteria with high capacity for hydrogen production could underpin further biohydrogen research and assist in the isolation of dominant hydrogen producers.
ACKNOWLEDGEMENTS
We thank Sang Eun Oh for technical advice. This research was funded by NSF grant
73
BES 01-24674 and by NSF (IGERT) grant DGE-9972759, which supported the Penn
State Biogeochemical Research Initiative for Education.
74
REFERENCES
1. Altschul SF, Gish W, Miller W, Myers EW and Lipman DJ (1990) Basic local
alignment search tool. J.Mol.Biol. 215:403–410
2. Bahl H and Durre P (2001) Clostridia: Biotechnology and Medical Application,
Wiley-VCH Verlag GmbH, Weinheim, pp147
3. Biebl, H (1999) Clostridium acetobutylicum. In Encyclopedia of food
microbiology; Robison,R., Batt, C., Patel, P., Eds.; Academic press: New York.
4. Brosseau JD, Zajic JE 1982. H2-gas production with Citrobacter intermedius and
Clostridium pateurianum. J. Chem. Tech. Biotechnol 32:496-502
5. Chang F-Y and Lin C-Y (2004). Biohydrogen production using an up-flow anaerobic
sludge blanket reactor. Int J Hydrogen Energy. 29:33-9
6. Chin KJ, Hahn D, Hengstmann U, Liesack W, Janssen PH 1999. Characterization
and identification of numerically abundant culturable bacteria from the anoxic bulk
soil of rice paddy microcosms. Appl.Environ.Mcirobiol. 65(11): 5042-5049.
7. Chen CC, Lin CY, Chang JS (2001) Kinetic of hydrogen production with continuous
anaerobic cultures utinilizing sucrose as the limiting substrate.
Appl.Microbiol.Biotechnol. 57:56-64.
8. Gould GW (1970) Germination and the problem of dormancy. J.Appl.Bact 33:34-49.
9. Duncan, CL and Foster EM (1968) Effect of sodium nitrite, sodium chloride, and
sodium nitrate on germination and outgrowth of anaerobic spores. Appl. Microbiol.
16: 406-411.
10. Ginkel SV, Sung S, Lay JJ (2001) Biohydrogen production as a function of Ph and
75
substrate concentration. Envrion.Sci.Technol 35:4726-30
11. Hallenbeck PC, Benemann JR (2002) Biological hydrogen production:
fundamentals and limiting processes. Int.J. Hydrogen Energy 27:118-93.
12. Hussy, Hawkes FR, Dinsdale R and Hawkes DL (2003) Continuous fermentative
hydrogen production from a wheat starch co-product by mixed-microflora.
Biotechnol.Bioeng 84(6): 619-26
13. Iyer P, Bruns MA, Zhang H, Van Ginkel S and Logan BE (2004) H2-Producing
bacterial communities from a heat-treated soil inoculum Appl.Microbiol.Biotechnol
66: 166-173.
14. Jone DT and Woods DR (1986) Acetone-butanol fermentation revisited. Micro.
Rev. 50(4): 484-524
15. Kalia VC, Jain SR, Kumar A and Joshi AP (1994). Fermentation of bio-waste to H2
by Bacillus licheniformis. World J Microbiol Biotechnol 29:280-7
16. Kim JO, Kim YH, Ryu JY, Song BK, Kim IH, Yeom SH (2004). Immobilization
Methods for continuous hydrogen gas production biofilm formation versus granulation.
Proc Biochem. 40:1331-7
17. Kim SH, Han SK and Shin HS (2006) Effect of substrate concentration on
hydrogen production and 16S rDNA-based analysis of the microbial community in a
continuous fermenter. Process Biochemistry. 41: 199-207.
18. Kumar GR, Vatsala TM (1989) H2 production from glucose by Citrobacter freundii.
Indian J Exp Biol 27:824-825.
19. Kumar N, Das D (2001) Continuous H2 production by immobilized Enterobacter
76
cloacae IIT-BT 08 using lignocellulosic materials as soild matrices. Enzyme Microb
Technol 29:280-287.
20. Fabiano B and Perego P (2002). Thermodynamic study and optimization of
hydrogen production by Enternobacter aerogenes. Int.J.Hyfrogen.Energy. 27: 149-59
21. Fang HHP, Zhang T and Liu H (2002a) Microbial diversity of mesophilic hydrogen-
producing sludge. Appl.Microbiol.Biotechnol 58:112-118
22. Fang HHP, Zhang T, Liu H (2002b) Characterisation of a hydrogen-
producing granular sludge. Biotechnol.Bioeng 78:44-52.
23. Lane DJ (1991) 16S/23S rRNA sequencing. In: Stackebrandt E, Goodfellow M (eds)
Nucleic acid techniques in bacterial systematics. Wiley, New York, pp 115–175
24. Lay JJ, Lee YJ and Noike T (1999). Feasibility of biological hydrogen production
from organic fraction of municipal solid waste. Water Res: 33(11):2579-86.
25. Lay JJ. 2001.Biohydrogen generation by mesophilic anaerobic fermentation of
microcrystalline cellulose. Biotechnol Bioeng.. 74: 280-287.
26. Logan B E, Oh, S.-E, Kim I S and Van Ginkel S (2002) Biological hydrogen
production measured in batch anaerobic respirometers. Environ. Sci.Technol.36: 2530-
2535.
27. Mitchell, W.J. (1998) Physiology of carbohydrate to solvent conversion by clostridia,
Adv.Microbiol.Physiol. 39,31-130.
28. Minamisawa K, Nishioka K, Miyaki T, Ye Bin, Miyamoto T, You M, Saito A,
Saito A, Barraquito WL, Teaumroog N, Sein Than and Sato Tadashi. (2004)
Anaerobic Nitrogen-Fixing Consortia Consisting of Clostridia Isolated from
77
Gramineous Plants. Appl.Environ.Microbiol 70: 3096-3102.
29. Mizuno O, Ohara T, Shinya M and Noika T (2000) Characteristics of H2 production
from bean curd manufacturing waste by anaerobic microflora. Water. Sci.Technol 42:
345-350.
30. Mizuno O, Dinsdale R, Hawkes FR, Hawkes DL, Noile T (2002). Enhancement of
hydrogen production from glucose by nitrogen gas sparaging. Bioresource Technol
31. Montville T.J, Jones SB, Conway L.K and Sapers GM (1985) Germination of
spores from Clostridium botulinum B-aphis and Ba410. Appl.Environ.Mcirobiol.
50(4): 795-800.
32. Nielsen AT, Amandusson H, ,Bjorklund R, Dannetun H, Ejlertsson J, Ekedahl L-
G Sevennsson BH. (2001) Hydrogen production from organic waste Int.J. Hydrogen
Energy. 26:547-50.
33. Oh SE, Lyer P, Bruns MA and Logan BE (2004) Biological hydrogen production
using a membrane bioreactor. Biotechnol.Bioeng 87(1): 119-27.
34. Park W, Hyun SH, Oh SE, Logan BE, Kim IS (2005) Removal of headspace CO2
increases biological hytorgen production. Envrion.Sci.Technol 39(12): 4416-20.
35. Rachman MA, Nakashimada Y, Kakizono T and Nishio N (1997) Enhanced
hydrogen production in altered mixed acid fermentation of glucose by Enterobacter
aerogenes. J.Ferment.Bioeng., 83,358-363
36. Ruzicka, M (1996) The effect hydrogen on acidogenic glucose cleavage. Water Res.
30(10): 2447-2451.
37. Sambrook J, Frutsch EF and Maniatis T 1989) Moleculae cloning: A laboratory
78
manual, 2nd edn. Cold Spring Harbor Laboratory Press, Plinview, N.Y.
38. Smoot LA, Pierson (1982) Inhibiton and control of bacterial spore germination.
J.Food.Prot. 45:84-92.
39. Solomon BO, Zeng A-P, Biebl H, Schlieker H, Posten C, Deckwer W-D 1995.
Comparison of the energetic efficiencies of H2 and oxychemicals formation in
Klebsiella pneumoniae and Clostridium butyricum during anaerobic growth on glycerol.
J.Biotechnol 39:107-117.
40. Song YG, Logan BE (2004) Effect Of O2 exposure on perchlorate reduction by
Dechlorosoma sp. KJ. Water.Res. 38(6):1626-1632
41. Taguchi F, Mizukami N, Saito-Taki T and Hasegawa K (1995) Hydrogen
production from continuous fermentation of xylose during growth of Clostridium sp.
Strain no.2. Can. J . Mcrobiol.,41,536-540.
42. Tortura GJ, Funke BR, Case CL (1998) Microbiology, 6th ed., New York, USA:
Benjamin/Cumming,.
43. Van Ginkel S, Sung S and Lay JJ. (2002) Biohydrogen Production as a Function of
pH and Sustrate Concentration. Environ. Sci. Technol. 35: 4726-4730.
44. Van Ginkel S and Logan Bruce (2005) Increased biological hydrogen production with
reduced organic loading. Water Research. 39:3819-3826.
45. Walden CW and Hentgens DJ (1975) Differential effects of oxygen-reduction
potential on the multiplication of three species of anaerobic intestinal bacteria.
Appl.Microbiol.: 30(5):781-785
46. Yokoi H, Maeda Y, Hirose J, Hayashi S, and Takasaki Y (1997). H2 production by
79
immobilized cells of clostridium butyricum on porous glass beads. Biotechnol. Lett.
11: 431-433
47. Yokoi H, Tokushige T, Hirose J, Hayashi S, Takasaki Y 1998. H2 production from
starch by a mixed culture of Clostridium butyricum and Enterobacter aerogenes.
Biotechnol. Lett.. 20(2): 143-147.
80
10g/L 7.5g/L 5.0g/L 2.5g/L
Marker F2 F4 F6 F2 F4 F6 F2 F4 F6 F2 F4 F6 1353bp 1078bp
872bp
603bp
Figure 3-1: RISA profiles showing the bacterial community fingerprints of batch
cultures without L-cysteine. The glucose concentrations and the number of serial transfers
indicated at the top. The size of DNA bands are indicated by the arrows.
10g/L 7.5g/L 5.0g/L 2.5g/L
Marker F2 F4 F6 F2 F4 F6 F2 F4 F6 F2 F4 F6 1353bp 1078bp
872bp 603bp
Figure 3-2: RISA profiles showing the bacterial community fingerprints of batch
cultures with L-cysteine. The glucose concentrations and the number of serial transfers are
indicated at the top. The size of DNA bands are indicated by the arrows 81
500
450
400
350 300 2.5 g/L glucose 250 5.0 g/L glucose 7.5 g/L glucose 200 10 g/L glucose 150
Gas production (ml) Gas production 100 50 0 0 50 100 150 200 250 300 350 400
Time (hour) Figure 3-3: The cumulative biogas production by the batch cultures without L-
cysteine 550 500 450 400 350 300 2.5 g/L glucose 5.0 g/L glucose 250 7.5 g/L glucose
200 10 g/L glucose 150 Gas production (ml) Gas production 100 50 0 0 50 100 150 200 250 300 Time (hour)
Figure 3-4: The cumulative biogas production by the batch cultures with L- cysteine 82
90 70
80 60 70 50 60
50 40
40 30 Biogas (mL) Hydrogen (%) 30 20 20 Biogas from 2.5g/L-6 10 10 Hydrogen (%)
0 0 301 306 311 316 321 326 Time (h)
Figure 3-5: H2 percentage in the biogas produced from batch culture at 2.5 g/L
glucose during bacterial growth phase
83
2.5
2.5g/L 7.5g/L 5.0 g/L 10g/L 2
/mol Glucose) Batch 2 1.5 Batch Batch Batch culture 4 Batch Batch culture 1 culture 2 culture 3 culture 5 culture 6
1
0.5
Hydrogen yield (mol-H Hydrogen yield 0 123456 Batch cultures
Figure 3-6 H2 yield from the batch cultures without L-cysteine
2.5 2.5g/L 7.5g/L 5.0 g/L 10g/L Batch Batch Batch Batch Culture 6 Culture 5 2 Batch Batch Culture 3 Culture 4 Culture 1 Culture 2 /mol Glucose) 2 1.5
1
0.5
Hydrogen yield (mol-H Hydrogen yield 0
123456 Batch cultures
Figure 3-7 H2 yield from the batch cultures with L-cysteine
84
Table 3-1 Summary of peak gas production, final pH and H2 percent in headspace when batch cultures with L-cysteine exhibited no further accumulation of biogas.
2.5 g/L Glucose 5.0 g/L Glucose 7.5 g/L Glucose 10 g/L Glucose
Gas H2 Gas H2 Gas H2 Gas H2 Productin Final pH Percent Productin Final pH Percent Productin Final pH Percent Productin Final Percent (ml) (%) (ml) (%) (ml) (%) (ml) pH (%)
Batch culture 1 39.5 5.70 39.9 164 5.24 51.24 219 4.70 53.66 437 4.80 55.33
Batch culture 2 59 5.58 49.6 196 5.26 53.98 362 4.75 54.17 458 4.60 56
Batch culture 3 77 5.66 55.31 224 4.89 53.82 303 4.90 53.83 500 3.90 55.7
Batch culture 4 67 5.42 58 164 4.40 53.6 352 4.56 56.1 489 4.19 55.7
Batch culture 5 85 5.22 57.2 224 4.78 53.7 350 4.49 54.3 521 4.23 52.8
Batch culture 6 91 4.93 60.3 232 4.80 53.4 374 4.52 55 518 4.28 57.25 85
Chapter 4
Strain LYH2, a N2-fixing, ferric iron-reducing Clostridium sp. isolated from a silty clay loam soil
Luo Y.H. and Bruns M.A.
(To be submitted for publication in International Journal of Systematic and
Evolutionary Microbiology)
86
ABSTRACT
An endospore-forming bacterial isolate was obtained from silty clay loam soil and
grown in H2-producing batch cultures in medium containing glucose and L-cysteine. Based on its 16S rRNA sequence, the isolate was designated Clostridium sp. strain LYH2 having closest affiliation to Cluster I Clostridium representative C. pasterurianum (97% similarity).
Clostridium sp. FRB1 was also closely affiliated with Strain LYH2. The pH range for the growth of strain LYH2 was 4.6-7.8 with no distinct optimum between pH 6.0-7.8. The temperature range for the growth of stain LYH2 was 15oC-55oC with no distinct optimum
between 35oC-45oC . The G+C content of LYH2 was 25 mol%. LYH2 grew in nitrogen-free
medium at pH 6.2 under an N2 atmosphere and reduced acetylene at rates approximately 1.3-
-1 1.4 nmol (mg dry weight). Ethanol, butyrate, acetate, H2, and CO2 were end products of
glucose fermentation. In the presence of glucose as carbon and energy sources, thiosulfate,
sulfur and ferric iron could be reduced as electron acceptors. Thus strain LYH2 was
characterized as N2-fixing, ferric iron-reducing, and H2-producing fermentative chemo-
organotroph.
87
INTRODUCTION
Clostridia are gram-positive, spore-forming, rod-shaped bacteria. Polymers and monosaccharides derived from the degradation of polymers are important substrates for clostridia in nature as they can provide carbon compounds and energy for growth. Due to the wide distribution of polymers such as cellulose, xylan, starch, and pectin in plant biomass, soils contain a range of carbohydrates, from monosaccharides to large polymers, and are therefore favorable habitats for clostridia. Many clostridial species can convert carbohydrate into hydrogen as a fermentation by-product via hydrogenase (Mitchell 1998, Rogers 1986,
Vignais et al. 2001). Fermentation by H2-producing Clostridium species could provide a relatively inexpensive source of renewable energy. Biohydrogen production from the fermentation of synthetic glucose-containing wastewater has been studied using heat-treated soil as the source of microflora in continuous flow bioreactors. (Iyer et al. 2004, Oh et al.
2004). A large number of H2-producing Clostridium species were also reported to be isolated from different kinds of soils (Evvyernie D et al. 2000; Jeong H et al. 2005; Kataoka N et al.
1998; Kuhner et al. 2000, Ohmiya K 2005; Ohmiya K et al. 2006; Wilde E et al. 1997).
Clostridium spp. especially predominate in grassland and waterlogged soils or soil aggregates where moisture conditions and organic substrates are available but oxygen supply is severely restricted. Thus, soil appears to be an important source for H2-producing Clostridium species. 88 Most Clostridium species are obligate anaerobes, although tolerance to oxygen varies widely (Cato 1986). Therefore, the utilization of reducing agents such as L-cysteine should stimulate the growth of Clostridium species due to the elimination of oxygen from the medium (Song and Logan 2004). L-cysteine can also work as a spore germinant (Gould 1970,
Montville et al. 1985). Thus, the addition of L-cysteine to maintain a lower oxidation- reduction potential during initial cultivation is a potential method for the isolation of H2- producing Clostridium species from soil. In this paper, we report the isolation and characterization of Clostridium sp. strain LYH2 from a silty clay loam soil (fine, mesic, typic hapludalfs ). The addition of L-cysteine to the isolation medium appears to have enhanced its growth by acting as an oxygen scavenger and spore germinant.
MATERIALS AND METHODS
Sampling and isolation
Soil (Hagerstown silt loam, fine, mixed, mesic typic hapludalfs) was sampled from
harvested tomato plots to 5 cm depth at the Penn State University Park campus. Soil was
dried, ground to a uniform powder, and mixed thoroughly by sieving. Soil was then heated at
105oC for 2 hours to remove nonspore-forming methanogens which would consume
hydrogen (Lay 2001; Biebl 1999; Logan et al. 2002; Van Ginkel et al.2002). The mineral
medium used for the culture of spore-forming hydrogen producers from soil contained the
following components (g/L): NH4Cl (0.88); K2HPO4 (0.44); KH2PO4 (0.44); MgSO4.7H2O
(0.36); FeCl3 (0.05); NiCl2.6H2O (0.0476); CaCl2.2H2O (0.0662); ZnCl2 (0.023); CoCl.6H2O
o (0.021); CuCl2 (0.01), MnCl2 (0.03). 0.5g heated soil (105 C for 2 hr in convection oven) 89 was added to 100 ml autoclaved mineral medium with 10g/L glucose concentration and L- cysteine was added as an oxygen scavenger at a final concentration of 0.5 g/L The culture was incubated at 30oC under anaerobic conditions until it became turbid. Organisms were isolated on mineral medium agar with 0.5 g/L L-cysteine using serial dilutions of the sample
and spread-plate method (Madigan et al., 2003). Plates were then incubated at 30oC under
anaerobic conditions until colonies appeared on the plates. Colonies were streaked to ensure purity and an individual purified colony was chosen for further study.
Growth parameters
Anoxic BactoTM tryptic soy broth (TSB medium) was used for pH and temperature
range studies. The pH of medium was adjusted with sterile anaerobic stock solutions of 1 M
HCl or 0.5M NaOH to obtain pH values between 3.7 and 8.7. The medium was dispensed
aseptically into 50 mL polypropylene centrifuge tubes (40ml medium per tube). All
manipulations of medium and inoculation were carried out in an anaerobic glove box. Log-
phase culture was transferred to medium in the tubes at a ratio of 2.5% (v/v). The tubes were
then removed to a GasPak jar containing an activated anaerobic system envelope with
palladium catalyst inside (Becton Dickinson). The GasPak jar filled with tubes was placed at
30oC. The optical densities of samples were measured periodically by spectrophotometer
(PerkinElmer) at 600 nm. The temperature range for growth was tested in anoxic batch
bottles from 5oC to 60oC.
Reference cultures 90 Clostridium acidisoli ATCC BAA-167, Clostridium akagii ATCC BAA-166, and
Clostridium pasteurianum ATCC 6013 were used as reference cultures. Cultures were maintained in Tryptic Soy Broth medium.
Electron microscopy
Cells were grown in mineral medium at 30oC for scanning electron microscopy
(SEM). They were fixed in 1.5% glutaraldehyde in 0.1 M sodium cacodylate buffer (pH7.4) and were negatively stained with 2% (w/v) aqueous uranyl acetate (Valentine et al.,1968).
For transmission electron microscopy (TEM), cells were fixed in glutaraldehyde-OsO4 and
prepared by standard protocols (Traub et el., 1976). Thin sections were stained with uranyl
acetate and lead acetate (Reynold, 1963). Both the SEM and TEM micrographs were taken at
the Penn State Electron Microscope Facility.
G+C content
Microgram quantities of genomic DNA were obtained from late log phase cultures
using Gram Positive DNA Isolation Kit (Gentra Systems), with a modified step for cell lysis:
after the cell lysis solution was added to the cell pellet, resuspended cells were incubated at
50oC for 30 min instead of at 80oC for 5 min. The genomic DNA was then dialyzed using
Slide-A-Lyzer dialysis cassette (Perbio Science). The absorbancies of genomic DNA at 260
nm for thermal denaturation curves were measured with a spectrophotometer (Spectronic
instruments). The G+C content of purified genomic DNA was determined by its melting
temperature as described by Mandel & Marmur (1968). Escherichia coli ATCC 23848
served as control. 91 Alternative electron acceptors
Sulfur-containing electron acceptors were tested in mineral media with 0.5 g/L L- cysteine. Sodium thiosulfate (20mM), sodium sulfate (20mM), sulfite (2mM) and elemental sulfur (2% w/v) were added as electron acceptors. Tests for Fe(III) reduction were performed in the same mineral medium as above but with amorphous iron(III) oxide as an electron acceptor and 5% w/v glucose as electron donor and energy source. The medium was incubated anaerobically at 30oC temperature. The dissolved sulfide in cultures was determined with the method described by Cord-Ruwisch (1985). HCl-extractable Fe (II) in the culture was periodically assayed with ferrozine as described by Lovley and Phillips
(1987).
Fixation of N2 and reduction of acetylene
The N-free medium used in nitrogen fixation experiments contained the following components (g/L): MgSO4.7H2O (0.0493); FeCl3.6H2O (0.0541); MnSO4.H2O (0.0034);
ZnSO4.7H2O (0.00058); CuSO4.5H2O (0.00050); CoCl2.6H2O (0.00048); K2HPO4 (1.132);
-6 Biotin (2.5×10 ); NaCO3 (3.0). The fixation of nitrogen by strain LYH2 was assessed with the method described by Paerl (1998). Briefly, LYH2 strain was cultured in N-free medium with 5g/L glucose concentration in serum bottles. The aqueous to vapor phase in serum bottles was approximately 60:40 (v:v). The serum bottles were sealed with butyl rubber stoppers and LYH2 was allowed to grow in medium anaerobically with N2 as the sole source
of nitrogen. Prior to gas sampling, the headspace of the serum bottle was flushed with argon
to remove N2 and acetylene was injected immediately at approximately 15% of headspace 92 volume. After one hour of incubation, the amount of ethylene derived from reduced acetylene by nitrogenase was measured using a gas chromatograph (Hewlett Packard
HP6890).
Analysis of fermentation end products
Strain LYH2 was inoculated into 125 ml mineral medium containing 0.5 g/L L- cysteine in serum bottles with 5 % w/v glucose as carbon source. The headspace of the serum bottles was about 35 ml. The biogas production was monitored by a respirometer (Challenge
Environmental System AER-200 Respirometer, Fayeteville, AR). Once biogas production ceased as determined by respirometry, liquid samples were filtered using a 0.2um pore diameter (Whatman, Florham Park, NJ) and immediately diluted twofold with 50% formic acid. Samples were stored at -20oC prior to analysis. Acetone, methanol, ethanol, n-propanol,
n-butanol, acetate, propionate and butyrate in liquid phase were measured using an Agilent
GC-FID (Palo Alto, CA) using pure chemicals as standards. Headspace hydrogen was
measured with a gas chromatograph equipped with a thermal conductivity detector and a
molecular sieve column (Molesieve 5A 80/100 6’X1/8X0.085, Alltech) with nitrogen as the
carrier gas. Carbon dioxide in headspace was measured using a gas chromatograph of the
same model (Porapak Q column; Alltech) with helium as the carrier gas.
Analysis of 16S rRNA gene sequence.
DNA was extracted from strain LYH2 using UltraClean Microbial DNA Isolation
Kit, according to the manufacture’s protocol (Mo Bio). PCR amplification of the 16S rRNA
gene was performed using a Perkin Elmer Gene Amp PCR, System 2400 with two bacteria- 93 specific 16S rDNA primer sets: (1) 27f (16S rRNA) and 907f (16S rRNA), (2) 926f (16S rRNA) and 115r (23S rRNA). The nearly complete 16S rDNA was obtained from the alignment of 16S rDNA sequence fragments derived from PCR amplification with these two primer sets. Cloning of the fingerprint DNA bands followed the TOPO TA Cloning kit protocol (Invitrogen). Cloned DNA bands were sequenced at the Penn State Nucleic Acid
Facility on an ABI 370 sequencer. Its closest relative was identified by database search with
BLAST. The nucleotide sequence of the 16S rDNA was manually aligned with reference sequences of various members of the genus Clostridium using CLUSTALX (1.81) software.
Reference sequences were obtained from GenBank databases (Benson et al., 1999). A phylogenetic tree was constructed by the neighbor-joining method (Saitou and Nei, 1987), based on approximately 1500 nucleotide positions (Escherichia coli positions 27–1,510).
RESULTS
G+C content
The G+C content of the genomic DNA of strain LYH2 was calculated to be 25%,
based on its melting temperature. This G+C value is significantly different from those of
Clostridium acidisoli (31.4%) and Clostridium akagii (30.7%). It is, however, close to the
G+C content of Clostridium pasteurianum (26-28%).
Electron acceptors
Among sulfur-containing electron acceptors, strain LYH2 used thiosulfate and sulfur,
but not sulfate or sulfite as electron acceptors. Strain LYH2 grew anaerobically in medium 94 with glucose as the organic carbon source and amorphous iron (III) oxide as external electron acceptor. After 36-62 hours of incubation, brown amorphous iron (III) oxide was converted to a black, solid material and 17.2-19.38 mmol/L accumulated Fe (II) was observed in the culture (Fig 4-5).
N2 fixation
Cultured of LYH2 wrtr maintainable in nitrogen-free mineral medium at pH 6.2 under an N2 atmosphere; no growth was observed in the nitrogen-free medium under an argon
atmosphere. Cells of LYH2 grown in medium with N2 as the sole source of nitrogen reduced
acetylene at rate approximately 1.3-1.4 nmol min-1 (mg dry weight)-1.
Fermentation end products
In mineral medium (pH 6.2) with 0.5 g/L glucose as carbon source, glucose was
fermented to acetate, butyrate, ethanol, H2 and CO2 by strain LYH2. Acetone, methanol, n- propanol, n-butanol and propionate were not detected in the culture (Table 4-1). This suggested that strain LYH2 is an acetogenic Clostridium species. The experiments were
replicated three times.
Phylogenetic analyses
The phylogenetic analysis of the 16S rRNA sequence showed that strain LYH2 was a
member of the low G+C Gram-positive bacteria and a member of cluster I of the Clostridium
subphylum (Fig 4-1) (Collins et al. 1994, Stackebrandt and Hippe 2001). Its closest relatives
in GenBank are Clostridium sp. FRB1 and Clostridium pasteurianum (97% similarity). Its 95 16S rRNA gene sequence similarities to Clostridium acidisoli and Clostridium akagii are
96% and 95%, respectively. That is, the 16S rRNA gene sequence of LYH2 is no more than
97% identical to any other sequence in the gene bank, indicating that LYH2 could belong to new clostridial species (Rossello-Mora and Amann, 2001).
Morphological and cytological characteristics
Strain LYH2 forms rod-shaped cells 2.6-8.2 μ in length x 0.6 μ width (Fig 4-2a and
4-2b). Sporulated cells can be observed upon microscopic examination. The cells consisted of single rods and multicellular chains (up to eight cells) or aggregates which gave liquid cultures a flocculated appearance (Fig 4-2c). On agar at pH 6.2 and temperature 30oC, LYH2 formed round colonies. Young colonies were white, opaque and convex. Older colonies became brown.
Optimal growth conditions
Strain LYH2 differed from C. acidisoli and C. akagii in its pH and temperature ranges for growth in TSB medium. The pH range for LYH2 was 4.6-7.8 with no distinct optimum between pH 6.0-7.8 (Fig 4-3). The temperature range for LYH2 was 15-55oC with no optimum between 35-45oC (Fig 4-4). Reported temperature ranges for C. acidisoli CK58T
and C. akagii CK74T were 5-30oC and 5-37oC respectively. The pH ranges for C. acidisoli
CK58T and C. akagii CK74T were 3.7-7.1 and 3.6-6.9. C. acidisoli and C. akagii are two of only a few Clostridium species which can grow below pH 4 (Kuhner et al. 2000). Although
Strain LYH2 and C. pasteurianum grow moderately well at 25oC, no distinct optimum was 96 exhibited for Strain LYH2 between 35oC and 45oC. At 45oC, however, C. pasteurianum grew
poorly (Cato et al. 1986).
Description of Clostridium sp. strain LYH2
Cells are rod-shaped, 2.6-8.2 x 0.6 μ, spore-forming, and Gram-positive. Cultures contain single cells and cells in short chains. Growth occurs at pH 4.6 -7.8 with no distinct optimum between pH 6.0-7.8. Temperature range for growth is 15-55oC with no optimum between 35-45 oC. Use thiosulfate, sulfur and ferric iron (III) as electron acceptors Sulfate
and sulfite cannot act as electron acceptors. Glucose is fermented to acetate, butyrate,
ethanol, H2 and CO2. The genomic DNA has a G+C content of 25 mol%. Sequence analysis
of the 16S rRNA gene shows that strain is closely related to Clostridium sp. FRBI and
Clostridium pasteurianum. Clostrdium sp. stain LYH2 was isolated from a silty clay loam
soil (fine, mesic, typic hapludalfs ) in Penn State University Park campus where tomatoes
were planted.
DISCUSSION
Phylogenetic analysis of 16S rRNA gene sequences showed that C. acidisoli, C.
akagii, C. pasteurianum and Clostridium sp. FRBI were the closest cultured relatives of
Strain LYH2. Percent nucleotide identities of16S rRNA gene sequences of strain LYH2 and
C. acidisoli and C. akagii are less than 97%, indicating that strain LYH2 may be a different
species. from Clostridium acidisoli and Clostidium akagii. Strain LYH2 also showed
morphological and physiological differences from C. acidisoli and C. akagi. Strain LYH2
could be distinguished from Clostridium acidisoli and Clostidium akagii in pH and 97 temperature ranges for growth, optimum pH range and temperature ranges for growth, G+C content, and liquid end products of fermentation. Phylogenetically, strain LYH2 is more closely related to Clostridium pasteurianum and Clostridium sp. FRBI (97% similarity). The
GC content of strain LYH2 (25 mol%) is close to that of Clostridium pasteurianum (26-28 mol%) (Cummind and Johnson 1971). Acetate, butyrate and ethanol can be observed in the liquid end product profiles of fermentation by both strain LYH2 and Clostridium pasteurianum (Ehrlich et al., 1971). However, strain LYH2 differed from Clostridium pasteurianum in pH range and temperature ranges for growth, and in utilization of electron acceptors. Clostridium pasteurianum can use sulfite as electron acceptor (McCready et al.
1976). In contrast, strain LYH2 is capable of reducing sulfur rather than sulfite. Strain LYH2 displayed a very strong ability to reduce ferric iron during the fermentation of glucose,
Clostridium pasteurianum was reported as a Clostridium species with very weak capability of reducing ferric Fe(III) (Ottow 1971). Although Clostridium butyricum (Park et al., 2001) and Clostridium beijerinckii (Dobbin et al., 1999) are known to reduce Fe(III) when glucose is provided as carbon source in the fermentation, Clostridium species other than Clostridium butyricum and Clostridium beijerinckii were indicated to be present in soils in Ottow’s report of the role of iron-reducing clostridia in formation of gleyed soils (Ottow 1971).
Clostridium sp. FRBI is another strain of Clostridium with a reported capacity to reduce
Fe(III). The comparisons of strain LYH2 with its close relatives were based on Clostridium strains that were presently available. To determine whether isolate LYH2 is a novel species or a novel strain of C. pasteurianum, DNA:DNA hybridization tests are needed.
98 ACKNOWLEDGEMENTS
The authors are grateful to David Jones for the analysis of fermentation end products,
Jennifer Loveland-Curtze for technical support in the determination of G+C content, and
Kathleen M. Brown and Amy R. Annino for their assistance in the measurement of acetylene reduction. This research was funded by NSF grant BES 01-24674 and by NSF (IGERT) grant
DGE-9972759, which supported the Penn State Biogeochemical Research Initiative for
Education. 99
REFERENCES
1. Benson DA, Boguski MS, Lipman DJ, Ostell J. Ouellette BFF, Rapp BA, Wheeler
D (1999). GenBank. Nucleic Acids Res 27:12-7
2. Biebl, H (1999) Clostridium acetobutylicum. In Encyclopedia of food microbiology;
Edited by Robison,R., Batt, C., Patel, P.. New York: Academic press.
3. Cato E.P, George W.L and Finegold SM (1986) Genus Clostridium Prazmowski
1880, 23AL. In Bergy’s Manual of systematic Bacteriology, vol.2, pp.1141-1200.
Edited by P.H.A Sneath, N.S Mair, M.E Sharp & J.G. Holt Baltimore: William
&Wilkins.
4. Collins MD, Lawson P.A, Willems A, Cordoba J.J, Fernandez-Garayzabal J,
Garcia P, Cai J, Hippe H & Farrow JAE (1994). The phylogeny of the genus
Clostridium: proposal of five new genera and eleven new species combination. Ins J
Syst Bacteriol 44,812-826.
5. Cord-Ruwisch Ralf. (1985) A quick method for the determination of dissolved and
precipitated sulfides in cultures of sulfate-reducing bacteria. Journal of Microbiology
Methods. 4:33-36
6. Cummins CS and Johnson JL. (1971) Taxonomy of the clostridia: wall composition
and DNA homologies in Clostridium butyricum and other butyric acid-producing
clostridia. J.Gen.Microbiol. 67:33-46.
7. Dobbin PS, Carter JP, San Juan CGS, von Hobe M, Powell AK and Richardson 100 DJ. (1999) Dissimilatory Fe(III) reduction by Clostridium beijerinckii isolated from
freshwater sediment using Fe(III) maltol enrichment, FEMS Microbil.Lett. 176:131-
138.
8. Ehrlich GG, Goerlitz DF, Bourell JH, Eisen GV and Godsy EM. (1981) Liquid
chromatographic procedure for fermentation product analysis in the identification of
anaerobic bacteria. Appl.Environ.Microbiol. 42:878-885
9. Evvyernie D,Yamazaki S, Morimoto K, Karita S, Kimura T, Sakka K, Ohmiya
K.(2000) Identification and characterization of Clostridium paraputrificum M-21, a
chitinolytic, mesophilic and hydrogen-producing bacterium. Journal of Bioscience and
Bioengineering. 89(6): 596-601
10. Gould G.W (1970) Germination and the problem of dormancy. J.Appl.Bact 33:34-49.
11. Hoch GE and Westlake DWS. (1958) Fixation of N2 by extraction from Clostridium
pasteurianum. Fed Proc. 17: 243.
12. Johnson, J.L, and Francis B.S (1975) Taxonomy of the clostridia: ribosomal
ribonucleic acid homologies among the species. J.Gen.Microbiol 88:229-244.
13. Jeong H, Yi H, Sekiguchi Y, Muramatsu M, Kamagata Y, Chun J (2004)
Clostridium jejuense sp. nov., isolated from soil. International Journal of Systematic
and Evolutionary Microbiology. 54(5): 1465-1468
14. Iyer P, Bruns M.A., H Zhang, G.S Van, Logan BE (2004) H2-Producing bacterial
communities from a heat-treated soil inoculum. Applied Microbiology and
Biotechnology 66:166-173
15. Kataoka N, Tokiwa Y (1998) Isolation and characterization of an active mannanase-
producing an aerobic bacterium, Clostridium tertium KT-5A, from lotus soil. Journal 101 of Applied Microbiology. 84(3): 357-367
16. Kuhner C.H, Matthies C, Acker G, Schmittroch M, Gossner A.S, Drake H.(2000)
Clostridium akagii sp.nov. and Clostridium acidisoli sp.nov.: adcid-tolerant, N2-fixing
clostridia isolated from acidic forest soil and litter, Int J Syst Evol Microbiol 50:873-81
17. Lay, JJ (2001) Biohydrogen generation by mesophilic anaerobic fermentation of
microcrystalline cellulose. Biotechnol Bioeng. 74(4), 280-287
18. Logan B E, Oh, S-E, Kim I S and Van Ginkel S (2002) Biological Hydrogen
Production Measured in Batch Anaerobic Respirometres. Environ. Sci.Technol. 36:
2530-2535.
19. Lovely DR and Phillips EJP (1987) Rapid assay for microbially reducible ferric iron
in aquatic sediments. Appl. Environ. Microbiol. 53: 1536-1540.
20. Madigan MT, Martinko JM and Parker J (2003) Microbial growth. In Brock
Biology of Microorganism, pp.146-147. Edited by Madigan M.T., Martinko J.M. and
Parker J. New Jersey: Pearson Education.
21. Mandel, M & Marmur, J (1968). Use of ultraviolet absorbance-temperature profile
for determining the guanine plus cytosine content of DNA. Methods Enzymol 12B,
195–206.
22. McCready RGL, Laishley EJ and Krouse HR(1976) The use of stable sulfur
isotope labeling to elucidate sulfur metabolism by Clostridium pasteurianum.
Arch.Microbiol. 109:315-317.
23. Mitchell WJ. Biology and physiology. (2001) In: Bahl H, Purre P, editors. Clostridia.
Weinheim, Germany: Wiley-VCH. P.49-104.
24. Montville TJ, Jones SB, Conway LK and Sapers GM (1985) Germination of 102 spores from Clostridium botulinum B-aphis and Ba410 App.Environ.Microbiol.
50(4):795-800.
25. Stackebrandt E and Hippe H (2001) Toxonomy and Systematics, In Clostridia:
Biotechnology and Medical Application, Edited by H. Bahl and P. Durre, pp 19-49,
Wiley-VCH, New York.
26. Traub, Wh., Acker, G & Kleber, I. (1976). Ultrastructural surface alterations of
Serratia marcescens after exposure to polymyxin B and/or fresh human serum.
Chemotherapy 22, 104-113.
27. Oh S.E, lyer P, Bruns M.A and Logan BE (2004). Biological hydrogen production
using a membrane bioreactor. Biotechnology and Bioengineering. 87(1): 119-127
28. Ohmiya K (2005) Effective utilization of biomass by using anaerobic bacterial
functions. Seibutsu-Kogaku-Kaishi. 83(1): 7-17
29. Ohmiya K, Sakka K, Kimura T (2006) Anaerobic bacterial degradation for the
effective utilization of biomass. Biotechnology and Bioprocess-Engineering.10(6):
482-493
30. Ottow JCG 1971. Iron reduction and gley formation by nitrogen-fixing clostridia.
Oecologia. 6(2): 164-175.
31. Park HS, Kim BH, Kim HS, Kim GT, Kim M, Chang IS, Park YH and Chang HI
(2001) A novel electrochemically active and Fe(III)-reducing bacterium
phylogenetically related to Clostridium butyricum isolated from a microbial fuel cell.
Anaerobe 7:297-306.
32. Paerl HW (1998) Microbially mediated nitrogen cycling. In Techniques in microbial
ecology. pp. 3-30. Edited by Burlage R.S., Atlas R., Stahl D., Geesey G., and Sayler 103 G..; New York: Oxford University Press.
33. Reynolds, ES (1963) The use of lead citrate at high pH as an electron opaque stain in
electron microscopy. J Cell Bio 7, 208.
34. Rogers P, (1986) Genetics and biochemistry of Clostridia relevant to development of
fermentation processes, in: Advances in Applied Microbiology (A.I.Laskin, Ed.),
Academic Press, New York, pp.1-89.
35. Rosselleo -Mora R and Amann R (2001) The species concept for prokaryotes. FEMS
Microbial Revs. 25:39-67
36. Saitou N and Nei M (1987). The neighbor-joining method: a new method for
reconstruction phylogenetic trees. Mol Bio Evol. 4:405-25.
37. Song YG, Logan BE. (2004) Effect of O2 exposure on perchlorate reduction by
Dechlorosoma sp. KJ. Water.Res. 38(6):1626-1632.
38. Valentine , RC, Shapiro, BM & Stadtman, ER (1968) Regulation of glutamine
synthetase. XII. Eletron microscopy of the enzyme from Escherichia coli. Biochemistry
7, 2143-2152.
39. Van Ginkel S, Sung S and Lay J J (2002) Biohydrogen Production as a Function of
pH and Substrate Concentration. Environ. Sci. Technol. 35: 4726-4730.
40. Vignais P.M, Billoud B and Meyer J (2001) Classification and phylogeny
of hydrogenases. FEMS Microbiology Reviews. 25:455-501.
41. Wilde E, Collins MD, Hippe H (1997) Clostridium pascui sp. nov., a new glutamate-
fermenting sporeformer from a pasture in Pakistan. International Journal-of-
Systematic-Bacteriology. 47(1): 164-170
104
ClostridiumC.pasteurianum pasteurianum M23930 (M23930)
99
C.FRB1Clostridium AY925092 sp. FRBI (AY915092) 99
LYH2 100 ClostridiumC.akagii AJ237755 akagii (AJ237755)
86 81 ClostridiumC.acidisoli CK74 acidisoli AJ237756CK74 (AJ237756)
72 ClostridiumC.acetobutylicum acetobutylicum ATCC 824 AE0ATCC 824 (AE007543)
ClostridiumX68316 C argentinense argentinense (x68316)
ClostridiumC.sporogenes sporogenes ( X68189 ) (X68189) 59 90 ClostridiumC.tetani X74770 tetani (X74770)
100 C.beijerinckiiClostridium beijerinckiiDSM 791 X68179DSM 791 (X68179)
ClostridiumC.proteolyticum proteolyticum DSM 3090 DSM X734 3090 (X73448)
ClostridiumC.cellulosi L09177 cellulosi (L09177) 100 100 ClostridiumC.thermocellum thermocellum DSM 1237 L0917DSM 1237 (L09173)
BacillusBacillus subtilissubtilis AF549498(AF549498) 0.01
Fig 4-1: Neighbor-joining dendrogram illustrating phylogenetic relationship between strain LYH2 and its related Clostridium species based on nearly complete 16S rRNA gene sequences. The tree was constructed using neighbor-joining algorithm with 100 bootstrappings. Numbers at branching points refer to bootstrap values > 50%. Accession numbers of 16S rRNA gene sequences are given in parentheses. The scale bar represents a 1% difference in nucleotide sequence.
105
(b) (a)
WL
CM
(C)
s
Fig 4-2: Transmission electron micrographs of strain LYH2.
(a) Negatively stained preparation.
(b) Thin section. Abbreviations: WL wall layer; CM, cytoplasmic membrane.
(c) Thin section showing septum formation at two division points (S)
Bars: 1µm(a), 0.2µm(b), 5µm (c)
106
1.2
1.0
0.8 600 0.6 OD
0.4
0.2
0.0
0 10 20 30 40 50 60 70 Time (h) Time (h)
Fig 4-3: Effect of pH on the growth of LYH2 in TSB medium. Initial pH of medium:
§,3.7,4.2,8.3 and 8.7; , 4.6; ,5.4; ,6.0;{,6.5;W,6.8; ,7.1; ,7.4;+,7.8.
107
1.2
1.0
0.8
600 0.6 OD 0.4
0.2
0.0 0 5 10 15 20 25 Time (h)
Fig 4-4: Optimal temperature for the growth of LYH2 in TSB medium (pH 6.2) :
o o o o o o o z, 25 C, 30 C, 50 C and 55 C ; , 35 C; ,40 C;{, 45 C.
108
LYH2 Control
Fig 4-5: After 54-62 hour of incubation at 30oC, non-magnetic, brown amorphous iron
(III) oxide in 5g/L glucose culture was converted by LYH2 to a black, solid material of
less volume, which was strongly attracted to a magnet and contained a significant amount of Fe(II).
109
Table 4-1 Glucose-dependent product profile of LYH2.
Products formed Medium Glucose concentration (pH) (g/L) Butyrate Acetate Ethanol H2 CO2 (ppm) (ppm) (ppm) (%) (%)
Mineral medium (6.2) 5.0 1235.1±109.7 523.6±54.1 60.70±8.55 57.5±0.3 35.2±1.1
Data are given as mean±SD, n=3
110
Chapter 5
General Conclusions
111 Hydrogen is regarded as a feasible alternative to fossil fuel. The biological generation of hydrogen by microalgae and bacteria is an important way to obtain hydrogen from renewable resources. Biohydrogen production is divided into two main categories: photosynthetic hydrogen production and fermentative hydrogen production.
Among them, fermentative hydrogen production from wastewater containing carbohydrate has been extensively investigated because it can serve the dual role of renewable energy production and waste treatment. Therefore, the hydrogen production by fermentative H2-producing bacteria is a very attractive sustainable technology.
To increase hydrogen production by fermentation, it is necessary to optimize design parameters such as substrate concentration, hydraulic retention time, oxidation- reduction potential for the fermentation. We hypothesized that specific growth factors can affect the composition of bacterial communities during H2 production and that these
factors influence the hydrogen yields of the communities. Our study focused on the effect
of glucose concentration and L-cysteine on the composition of H2-producing bacterial
communities in a bioreactor and in batch cultures inoculated with heat-treated soil. In
addition, a new strain of H2-producing Clostridium sp. was isolated from mixed soil
inoculum using L-cysteine as oxygen scavenger.
In continuous-flow bioreactors, our work indicated that H2-producing bacterial
communities were dominated by Clostridium spp. when glucose concentrations were
greater than 2.5 g/L (5.0, 7.5 and 10g/L glucose). The bacterial community in the reactor
operated with 2.5 g/L glucose was more diverse than at other glucose concentration, consisting of representatives of Clostridiaceae, Enterobacteriaceae and
Acidominococcaceae. The lowest glucose concentration for supporting effective 112 competition of Clostridium spp. over other species appeared to be 5g/L. Our study also implied that hydraulic retention time could affect the H2-producing bacterial community compositions.
In batch experiments, differences in glucose concentration affected the compositions of H2-producing bacterial communities in ways that were similar to those
observed in the bioreactor study. When glucose concentrations were greater than 2.5 g/L
(5.0, 7.5 and 10g/L glucose), batch cultures were eventually dominated by Clostridium
species. Enterobacter spp., however, dominated bacterial communities at 2.5 g/L glucose.
Initial oxidation-reduction potential and number of successive transfers were two other
factors that affected microbial community composition in batch cultures. With higher
numbers of successive transfers, there was a shift in dominant populations from
Enterobacter spp. to Clostridium spp. in 5g/L batch cultures with 0.5 g/L L-cysteine. The
addition of L-cysteine could improve the competition of oxygen-sensitive anaerobic
bacteria with other species by eliminating trace oxygen present in the medium.
An isolate designated Clostridium sp. strain LYH2 was obtained using L-cysteine
3+ 2+ as reducing agent and spore germinant. Strain LYH2 fixes N2, reduces Fe to Fe , produces hydrogen by fermentation of glucose, and grows across a broad range of temperature and pH. Since LYH2 tolerates a broad range of temperature and pH, the feasibility of using LYH2 for industrial production of biohydrogen is improved due to the possibility that temperature and pH can be adjusted to reduce growth by other microbes.
In addition, the capability of LYH2 to reduce Fe3+ to Fe2+ may be used in fuel cells to
produce electricity from renewable sources. The recovery of strain LYH2 increases the 113 number of potential target microbes and genes which could be used in genetic engineering approaches to increase biohydrogen yield.
The results obtained from this research suggested how environmental factors affect hydrogen production by influencing the composition of H2-producing bacterial
communities. The identification of key growth factors is very important for large-scale
hydrogen production. For examples, our studies implied that glucose concentrations
above 5g/L can improve the competition of some H2-producing Clostridium species with
other bacteria. L-cysteine is a useful oxygen scavenger which can be used for the
germination of spore-forming H2-producing bacteria to improve H2 production yields.
The identification and optimization of key growth factors will be important in moving
toward industrial hydrogen production.
Although our studies showed that glucose concentration and addition of L-cysteine
can affect hydrogen production, much remains to be learned about whether there are other factors that could enhance hydrogen production. Based on the successful recovery of LYH2 in this study, attempts can continue to be made to isolate H2-producing bacteria
from other types of soils by using L-cysteine as reducing agent. For example, grassland
and waterlogged soils are known to be rich in Clostridium spp. Future research should
also focus on the performance of pure cultures under industrial conditions and the
sustainability of reactor-based biohydrogen production. Meanwhile, it would be
interesting to know how LYH2 strain could contribute to hydrogen production based on
the degradation of cellulose and hemicellulose, and if LYH2 could be combined with
other Clostridium spp. that can degrade biopolymers. Addressing these questions will aid
in making further progress on biohydrogen production from renewable materials. 114
Appendix A
Hydrogen Production from Bioreactor
115
Table A-1 Summary of continuous flow reactor operation
Source: Van Ginkel S.W., Logan B./ Water Research 39 (2005) 3819-3826
116
Appendix B
16S rRNA Probes
117 The experiments with fluorescent in situ hybridization (FISH) tests on bioreactor material were based on the recovery and its 16S rRNA sequence information. The first probe designed for LYH1 strain was complementary to E. coli 16S rRNA positions 973-
990 (Fig B-1).
The first LYH1 probe: 16 S rRNA sequences Target 16 S rRNA positions ( E.coli numbering) Strain LYH1 probe 3’-TATCGCATCTCTATGCAC-5’ Strain LYH1 target 5’-ATAGCGTAGAGATACGTG-3’ 973-990 Clostridium sp. FRB1 5’-ATAGCGTAGAGATACGTG-3’ C.pasteurianum strain CK74 5’-ATAGCGTAGAGATACGTG -3’ C.akagii 5’-ATTACCTGTA_ATAAGGG -3’ C.acidisoli 5’-ATAACGTAGAGATACGCG-3’ C.novyi 5’-ATTACTCGTAACTGAGGA-3’ C.sporogenes 5’-ATAGCCTAGAGATAGGTG-3’ C.acetobutylicum ATCC 5’-ATTAGTCCGTAATGGATG-3’ C.collagenovorans 5’-ATTAGTCCGTAATGGATG -3’ C.cellovorans 5’-ATTACCCGTAACTGGGGA-3’ C.tyrobutylicum 5’-ATAACCTAGAGATAGGCG -3’ C.subterminale 5’-ATTA _ CCTGTAATGAGGG-3’ C.argentinense 5’-ATTA _ CCTGTAATGAGGG-3’ C.tetani 5’-ACCTCTCTTAACTGAGGA-3’ C.cellulosi 5’-ATCCGGAAGAGATTCTGG-3’ C.thermocellum_DSM1237 5’-CAGCTCTAGAGAT_ _ _ AG-3’ C.proteolyticum_DSM309 5’-AGTCGGTGGAAACACTGA -3’ C.beijirinckii_DSM791 5’-ATTACCCTTAATCGGGGG-3’ B.subtilis 5’-CAATCCTAGAGATAGGAC-3’
Fig B-1 Fig 2-7: Alignments of the LYH1 probe sequence, its target site, and sequences
of corresponding sites in reference organisms. 17 related Clostridia, Bacillus subtilis were selected as reference organisms. The target sequence is displayed in the upper row; mismatches between target and sequences of corresponding sites in reference organisms are shadowed.
118 Because this probe hybridized to all cells in bioreactor samples grown at10g/L glucose, I reported that this bioreactor community consisted of nearly all LYH1 cells (Fig A-2).
EUB:5’-GCTGCCTCCCGTAGGAGT-3’ LYH1 probe: 5’-CACGTATCTCTACGCTAT-3’
(Labeled by Alexa Fluor 488, positive control probe) (Labeled by Alexa Fluor 555, specific probe)
A B
C D
Non EUB:5’-GCTGCCTCCCGTAGGAGT-3’
(Labeled by Alexa Fluor 488, negative control probe)
Fig B-2 Confocal Scanning Laser Microscope images of sample culture at 10g/L
glucose concentration with 10-h HRT: (A) The hybridization of sample cells with
universal EUB probe labeled by Alexa 488. (B) The hybridization of sample cells with
LYH1-specific probe labeled by Alexa 555. (C) The hybridization of sample cells with
NON EUB probe labeled by Alexa 488. (D) Morphology of sample cells examined by
confocal scanning laser microscopy.
Since the first probe designed for strain LYH1 was complementary to the
sequence in a non variable area of its 16S rRNA (973-990). The sequences of closely 119 related Clostridium spp. were similar to LYH1 in this region (Fig A-1). These clostridia also could have hybridized to the first probe. Therefore, I developed a second probe for
LYH1, which was complementary to E. coli positions 174-196, and which was confirmed to have mismatches with closely related Clostridium spp (Fig 2-7). I have carried out new
FISH experiments with the new LYH1 probe (Cy5 fluochrome) and a Cy3-labeled universal probe for bacteria. About 90% percent of the bacterial cells in microscope fields from bioreactor samples at 10 g/L glucose hybridized with the LYH1 probe (Fig 2-8).
This indicated that the bacterial community that grew in the 10g/L glucose feed was not a pure population. In sample from the bioreactor fed with 2.5 g/L glucose, only 26 percent of bacterial cells hybridized to LYH1 probe (Fig 2-9) . The varied bacterial communities may be responsible for the varied hydrogen yields in chemostat bioreactors at different glucose loading rates.
VITA
Luo Yonghua
EDUCATION
Expected 08 /07 Ph.D., Soil Science, The Pennsylvania State University
06/97 M.S., Microbiology, Wuhan Institute of Virology of Chinese Academy of Sciences.
06/94 B.S., Biology, Hunan Normal University
TEACHING EXPERIENCE
Fall 2006 Teaching Assistant, Soil Ecology, The Pennsylvania State University
WORK EXPERIENCE
2004-Present Graduate Teaching and Research Assistant, Department of Crop and Soil Sciences, The Pennsylvania State University
1997-2004 Research Assistant, Guangdong Institute of Microbiology
HONORS AND REWARDS:
1997 Diao Scholarship(Second Prize) of Chinese academy of Sciences
2002 Guang Dong Scientific Technology Award.
SELECTED PUBLICATIONS (of 10 total):
Yonghua L, Jun G. 2002. Initial location of nifHDK gene in W80-2 strain. J.Micro.
22 (2): 55-57
Yonghua L, Jun G. 2002. Molecular detectation of homology of hydrogen. J.Micro.
22 (3): 11-13