Lignocellulose As Carbon Source Promotes Bacterial Synergism and Reduces Antagonism

The University of Southern Mississippi The Aquila Digital Community

Dissertations

Spring 5-1-2016

Lignocellulose as Carbon Source Promotes Bacterial Synergism and Reduces Antagonism

Yijie Deng University of Southern Mississippi

Follow this and additional works at: https://aquila.usm.edu/dissertations

Part of the Biotechnology Commons, and the Environmental Microbiology and Microbial Ecology Commons

Recommended Citation Deng, Yijie, "Lignocellulose as Carbon Source Promotes Bacterial Synergism and Reduces Antagonism" (2016). Dissertations. 331. https://aquila.usm.edu/dissertations/331

This Dissertation is brought to you for free and open access by The Aquila Digital Community. It has been accepted for inclusion in Dissertations by an authorized administrator of The Aquila Digital Community. For more information, please contact [email protected]. LIGNOCELLULOSE AS CARBON SOURCE PROMOTES BACTERIAL

SYNERGISM AND REDUCES ANTAGONISM

Yijie Deng

A Dissertation Submitted to the Graduate School and the Department of Biological Sciences at The University of Southern Mississippi in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy

Approved:

______Dr. Shiao Y. Wang, Committee Chair Professor, Biological Sciences

______Dr. Mohamed O. Elasri, Committee Member Professor, Biological Sciences

______Dr. Kevin A. Kuehn, Committee Member Associate Professor, Biological Sciences

______Dr. Glenmore Shearer Jr., Committee Member Professor, Biological Sciences

______Dr. Dmitri V. Mavrodi, Committee Member Assistant Professor, Biological Sciences

______Dr. Karen S. Coats Dean of the Graduate School

May 2016

YIJIE DENG

2016

ABSTRACT

LIGNOCELLULOSE AS CARBON SOURCE PROMOTES BACTERIAL

SYNERGISM AND REDUCES ANTAGONISM

by Yijie Deng

May 2016

Lignocellulose decomposes slowly in nature because it consists of complex polymers resistant to enzymatic degradation by most organisms. Some bacteria are capable of producing cellulolytic enzymes, but the way in which bacteria interact within a community to enhance degradation of the recalcitrant substrate is poorly understood. A better understanding of how bacterial interactions affect lignocellulose degradation would provide potential approaches to improve the efficiency of lignocellulose degradation for biofuel production.

To study whether bacterial interactions enhance lignocellulose degradation, I grew environmental bacterial isolates in mixed cultures and pure cultures. I found that bacterial synergism in mixed cultures was common in lignocellulose medium. Bacterial synergism promoted bacterial growth, metabolic activity, and the production of β-1,4-glucosidase in mixed cultures. I also found that the complexity of carbohydrates mediated bacterial interactions. The synergistic growth found in lignocellulose medium was not observed in glucose medium, suggesting that bacterial synergism was substrate-dependent. Pairwise antagonistic interactions among bacteria showed that the frequency of antagonism in carboxymethyl cellulose (CMC)-xylan medium was only half of that in glucose medium, suggesting that reliance on complex polysaccharides as

carbon source reduces bacterial antagonism. The frequency of antagonistic interactions among bacteria was not randomly distributed. Firmicutes and

Gamma-Proteobacteria were among the most antagonistic groups, whereas

Bacteroidetes and Actinobacteria were the most susceptible groups. In addition, I also found different interaction network structures between bacteria relying on glucose and CMC-xylan as carbon sources.

Overall, results from the study showed that complex polysaccharides as the main carbon source promote bacterial synergism and reduce the frequency of bacterial antagonism. They support the potential use of bacteria synergism from different bacteria combinations to enhance plant biomass degradation/conversion for biofuel production.

iii

ACKNOWLEDGMENTS

I would never have been able to finish my dissertation without the support and love from my wife and my family, the advice of my committee members, and help and support from my friends.

Firstly, I am thankful to my advisor, Dr. Shiao Y. Wang, for his continuous support throughout my graduate career, giving me this opportunity of doing research and mentoring me for my professional development. I appreciate his valuable time in helping me revise my manuscripts, collect saltmarsh samples, and look for laboratory resources. I would definitely like to thank my dissertation committee members, Dr. Kevin A. Kuehn, Dr. Dmitri V. Mavrodi, Dr. Glen

Shearer, and Dr. Mohamed O. Elasri, for their valuable time, advice, and feedback. I would like to thank MS-INBRE for allowing me to use the facility.

I am really grateful to Professor Shujing Wang and Dr. Jiangchun Hu at

Shenyang Institute Applied Ecology, Chinese Academy of Sciences for their encouragements and support. I would also like to thank my graduate and undergraduate colleagues, especially Steven Everman, Krystyn Davis, Dr.

Xunyan Ye, Kimberly Lewis, Sarah Monroe, Dr. Rong Su, Dr. Gyan Sahukhal,

Deepak Kumar, and Mathew Arnold, for their assistance, encouragements, and good wishes at many occasions.

I cannot also forget my friends, Dr. Jinghai Yang, Lixing Song, Dawson

Zhang, Dr. Zheng Sun, Dr. Huiqing Zhu, and Dr. Shaoen Wu, for their support, prayers, and encouragements throughout my graduate career. I would like to express my deep gratitude to my wife, Ellen Shengai Jin, for her constant love,

care, and sacrifice to our family. Her sincere prayers and love have been critical for me in completing this endeavor. I am thankful to my daughter, Gina Zhener

Deng, and my son, Timothy Yangen Deng, for the happiness and troubles they brought to me, and for their reminder of what is life. I am grateful for the love and support from my Father, Xuehong Deng, my Mother, Qiongying Chen, and my brother, Yichao Deng. I have been missing you for years.

Finally, I would like to thank Lord for giving me this special opportunity to establish my research philosophy and strength to endure sufferings. I am also thankful to Him for teaching me the purpose of life and giving me a hope of His promises and faith in His immeasurable love. I thank Him for giving me inspirations and passions to do research even in many difficulties. Thank you!

DEDICATION

I dedicate my dissertation to my parents, Xuehong Deng and Qiongying

Chen, for their endless sacrifices and continuous support for my studies home and overseas.

TABLE OF CONTENTS

ABSTRACT...... ii

ACKNOWLEDGMENTS ...... iv

DEDICATION ...... vi

LIST OF TABLES ...... ix

LIST OF ILLUSTRATIONS ...... x

CHAPTER

I. INTRODUCTION ...... 1

1.1 Importance of Lignocellulose Degradation 1.2 Bacterial Interactions during Lignocellulose Degradation 1.3 Major Questions to Study

II. BACKGROUND RELEVANT TO THIS STUDY ...... 6

2.1 Bacterial Communities Associated with Plant Detritus 2.2 Bacterial Interactions and Organic Matter Degradation 2.3 Mechanisms of Bacterial Interactions 2.4 Significance of This Study

III. SYNERGISTIC INTERACTIONS AMONG SALT MARSH BACTERIA IN LIGNOCELLULOSE DEGRADATION...... 21

3.1 Abstract 3.2 Introduction 3.3 Materials and Methods 3.4 Results 3.5 Discussion 3.6 Conclusion

IV. COMPLEX POLYSACCHARIDES MEDIATE ANTAGONISTIC INTERACTIONS AMONG LIGNOCELLULOLYTIC BACTERIA .……...... 42

4.1 Abstract 4.2 Introduction

vii

4.3 Materials and Methods 4.4 Results 4.5 Discussion 4.6 Conclusion

REFERENCES ………………………………………………………………………62

viii

LIST OF TABLES

Table

4.1 Identify of Bacterial Isolates for Study of Antagonistic Interactions………49

4.2 Three Bacteria Exhibiting the Most Frequent Antagonism toward Other Bacteria………………………………………………………………….53

4.3 Frequency of Antagonistic Interactions among Bacterial Groups Grown on Glucose Medium…………………………………………………………...55

4.4 Frequency of Antagonistic Interactions among Bacterial Groups Grown on CMC-xylan Medium……………………………………………...………...56

LIST OF ILLUSTRATIONS

Figure

2.1 Examples of Bacterial Cooperation in Organic Matter Decomposition ….12

3.1 Phylogenetic Relationship of Nine Bacterial Isolates Used in This Study ………………………………………………………………………….. 31

3.2 Growth Comparison of Mixed Cultures to Their Corresponding Pure Cultures in Lignocellulose Medium .…………………………….…………. 32

3.3 Comparison of Mixed Culture Growth in Lignocellulose and Glucose Media ………………………………………………………………………….. 34

3.4 Effect of Synergy on Bacterial Growth, Specific Growth Rate, and Metabolic Activity …………………………………………………………….. 36

3.5 Effect of Synergy on Bacterial Production of Lignocellulolytic Enzymes ……………………………………………………………………… 36

4.1 An Example of Bacterial Antagonistic Interactions Detected by the Burkholder Spot-on-Lawn Method ……………………………………….... 47

4.2 Antagonistic Interaction Networks among Lignocellulolytic Bacteria Growing on Glucose and CMC-xylan Media …………………...………… 51

4.3 Comparison of the Number of Antagonistic Interactions among Bacteria Grown on Glucose and CMC-xylan Media………………………….…….. 53

4.4 Relationship between bacterial antagonism and susceptibility ……...…. 54

CHAPTER I

INTRODUCTION

1.1 Importance of Lignocellulose Degradation

Lignocellulose is the most abundant biomass on Earth, and its degradation has significant impacts on agriculture, biofuel development, and the environment (Field et al. 1998; Falkowski 2000; Lynd et al. 2002). In the environment, more than 100 billion tons of plant biomass (carbon) are produced each year through photosynthesis, and eventually those materials are mineralized by microbes into CO2 released into the atmosphere (Field et al.

1998). Lignocellulose degradation is the first and rate-limiting step for plant detritus decomposition, so it determines the flow of carbon cycle and CO2 emission at both global and local levels (Field et al. 1998; Falkowski 2000; Lynd et al. 2002). In agriculture practices, decomposition of crop residues in fields is critical to soil productivity. Large amounts of crop residues provide not only a source of nutrients to soil but also alter the chemical and physical property of soil during their degradation (Kumar and Goh 1999; Hasanuzzaman and Mahmood

2014). As degradation of lignocellulosic crop residues mediates soil fertility and quality in agriculture fields, it can affect crop growth either positively or negatively

(Kumar and Goh 1999).

In industry, lignocellulose is a good feedstock to use for making next- generation biofuels. Because plant biomass is abundant and renewable, biofuels produced from lignocellulosic biomass is one of the most promising alternatives to fossil fuels (Lynd et al. 2002; Zhang 2013). Degradation of the recalcitrant

feedstock into simple sugars is a critical step before microbes can ferment sugars into biofuels (Lynd et al., 2002). However, the cost of biofuel production is relatively high due to the low efficiency in degrading lignocellulose into soluble sugars (Lynd et al. 2002). The low efficiency is mainly due to the recalcitrant structure and insoluble property of lignocellulose, which limits large-scale production of biofuel from cellulosic biomass (Lynd et al. 2002; Zhang and Lynd

2004). Approaches that are more effective in lignocellulose degradation are needed in order to reduce the cost of biofuel production. Multiple approaches have been used to improve the efficiency of this step, including efforts to produce cellulases with increased activity by means of protein engineering (Schülein

2000; Zhang, Himmel and Mielenz 2006) and the discovery of new cellulolytic bacteria or new cellulase genes (Maki et al. 2009; Ransom-Jones et al. 2012).

Another approach is to develop microbial consortia that are more efficient in degrading cellulosic biomass (Kato et al. 2005; Wongwilaiwalin et al. 2010). New approaches are needed to enhance the efficiency of cellulose degradation.

Although lignocellulose degradation is a critical process in various contexts and has been studied for a long time (Field et al. 1998; Lynd et al.

2002), some aspects concerning fundamental mechanisms of microbial degradation remain unexplored, such as how microbes interact as a way to facilitate the degradation of recalcitrant lignocellulose.

1.2 Bacterial Interactions during Lignocellulose Degradation

Lignocellulose consists of three main types of polymers including cellulose, hemicellulose, and lignin that are strongly cross-linked in plant tissues

and form a recalcitrant architecture (Lynd et al. 2002). The complex compositions of lignocellulose and its insoluble nature limit the rate and extent of its decomposition in the environment (Hatfield and Ralph 1998). The first and the rate-limiting step is the initial hydrolysis of polymers into smaller and soluble molecules that can be utilized by organisms as carbon sources. Decomposing lignocellulose into smaller subunits is essential for microbial fermentation of biofuels (Lynd et al. 2002).

The primary way to break down lignocellulose is through microorganisms such as bacteria and fungi that are able to produce enzymes that degrade cellulose, hemicelluloses and lignin (Tomme et al. 1995; Lynd et al. 2002). The mechanism by which bacteria decompose plant biomass is their production of extracellular enzymes that degrade lignocellulosic material (McCarthy 1987;

Sinsabaugh, Osgood and Findlay 1994). Although enzymes are key components in microbial degradation of lignocellulose, they alone do not seem efficient enough in the degradation of recalcitrant plant biomass. The question is what other mechanisms bacteria can use to facilitate lignocellulose degradation besides producing lignocellulolytic enzymes.

In nature, bacteria do not live in isolation but interact dynamically with other bacteria. Competition for limiting resources has been suggested as the predominant interaction type for bacteria (Hibbing et al. 2010; Foster and Bell

2012). However, the common observations of bacterial syntrophy (Wintermute and Silver 2010a), microbial consortia during lignocellulose degradation (Haruta et al., 2002; Wongwilaiwalin et al., 2010) and mixed-species biofilm (Elias and

Banin, 2012) indicate the importance of positive interactions among bacteria.

Taken together, these studies indicate the complexity of bacterial interactions in the environment. In the current study, I proposed that interactions among bacteria might be substrate-dependent and that synergistic interactions are important for bacteria to utilize complex substrate such as lignocellulose. Due to the structural complexity of lignocellulose, bacteria might work in a cooperative manner to degrade the refractory substrate by producing complementary enzymes and/or by exchanging their metabolites. The goal of the study was to test the hypotheses that detritus-associated bacteria can form consortia and synergistically degrade lignocellulose and that the way bacteria interact depends on substrate complexity with more synergism in lignocellulose medium but more antagonism in labile substrates such as glucose medium.

1.3 Major Questions to Study

To better understand the mechanism by which bacteria degrade lignocellulose, it is critical to study the process in the context of bacterial interactions. The study provides insight about lignocellulose degradation in the environment and results from the study may be useful in developing microbial systems for biofuel production from lignocellulose. The goal of the study was to investigate how bacterial interactions affect lignocellulose degradation and how carbon source complexity mediate bacterial interactions. I used environmental bacteria cultures and their mixed-species cultures to study two main questions:

1) Are there synergistic interactions among detritus-associated bacteria

during lignocellulose degradation?

2) Does the structural complexity of the carbon source mediate bacterial

interactions?

CHAPTER II

BACKGROUND RELEVANT TO THIS STUDY

In this review, I discuss bacterial communities associated with plant biomass decomposition, the importance of bacterial interactions in organic matter decomposition, and some potential mechanisms by which bacteria interact with others.

2.1 Bacterial Communities Associated with Plant Detritus

As the main form of lignocellulose in nature, plant detritus harbors numerous bacteria and fungi that are the major decomposers (Odum and Biever

1984; Hall and Meyer 1998; Buchan et al. 2002; Buchan et al. 2003; Moore et al.

2004; Newman et al. 2015; Yamashita et al. 2015). Microbial communities on detritus are highly diverse as revealed by ribosomal RNA-based molecular methods (Buchan et al. 2003; Das, Royer and Leff 2007; Haichar et al. 2007;

Gihring et al. 2009; Rietl et al. 2016).

Bacteria play important roles in degradation and fermentation of lignocellulosic carbons as well as many other ecological functions such as nitrogen fixation, sulfate reduction and methanogenesis (Azam and Malfatti 2007;

Strom 2008; Pereyra et al. 2010). Though under dry conditions bacteria only account for a small proportion of microbial biomass (< 5%) when compared to fungi (Kuehn et al. 2000), in terms of the cell number, bacteria probably overwhelm any other organisms on detritus, being about 1011 ~ 1014 per gram dry litter. Their high numbers and high growth rate indicate that bacteria also play important roles in organic matter decomposition and transformation, especially

under aquatic conditions (Buchan et al. 2003; Das, Royer and Leff 2007; Haichar et al. 2007; Gihring et al. 2009; Rietl et al. 2016).

2.1.1 The diversity of bacterial communities for decomposition

Lignocellulolytic bacteria are widespread among most bacterial phyla, including Actinobacteria, Acidobacteria, Cytophaga-Flavobacterium-

Bacteroidetes (CFB), Firmicutes, Fibrobacteres, Proteobacteria, Spirochaetes,

Tenericutes, Thermotogae, and Verrucomicrobia. An analysis of more than

5,000 bacterial genomes showed that 21 out of the 24 analyzed phyla have glycoside hydrolase genes that are associated with cellulose degradation such as genes encoding β-glucosidases, endo-cellulases and exo-cellulases

(Berlemont and Martiny 2013). From another list of collections, bacteria experimentally showing cellulose degrading activity are also widespread across many phyla (http://www.wzw.tum.de/mbiotec/cellmo.htm).

Community profiling methods, such as denaturing gradient gel electrophoresis (DGGE), terminal restriction fragment length polymorphism (T-

RFLP), and next-generation DNA sequencing, also show diverse bacterial communities on detritus (Hannen et al. 1999; Rietl et al. 2016), although the specific function of each bacterial group during lignocellulose degradation is unclear. The bacteria in Cytophaga-Flavobacterium-Bacteroidetes (CFB) group have been implicated in the degradation of lignocellulose (Kirchman 2002) due to their ability to degrade complex carbohydrates (Lydell et al. 2004). This bacterial group was often found dominant in salt-marsh sediments (Lydell et al. 2004) and on decaying blades of S. alterniﬂora at a late stage (Buchan et al. 2003). Alpha-

and Gamma-Proteobacteria are among the most abundant bacteria on the detritus of Spartina alterniflora. It was shown that more than 60% clones belonged to alpha-Proteobacteria followed by Gamma-Proteobacteria. At the genus level, the predominant bacteria were Erythrobacter-related bacteria belonged to Alpha-Proteobacteria, followed by the Bacillus-related bacteria belonged to Firmicutes (Buchan et al. 2003).

2.1.2 Identification of functionally active bacteria

It has been noticed that rRNA-based methods alone are not able to distinguish which bacteria are playing what functions in a community. Functional gene analysis is a useful way to link bacterial populations to their ecological functions (Pereyra et al. 2010; Bates et al. 2011; Fierer et al. 2012; Sessitsch et al. 2012; Vanwonterghem et al. 2014). For example, gene sequencing of ring- cleaving dioxygenase gene pcaH on decaying S. alterniflora found that

Roseobacter group, a lineage of the class Alpha-Proteobacteria, was the most abundant lignin-degrading bacteria, which was consistent with isolates from enrichment cultures (Buchan, Neidle and Moran 2001). The diversity of pcaH gene found among bacteria indicated that many bacteria actually were involved in the degradation of lignin-related compounds during detritus decomposition.

Therefore, the development of functional gene primers (e.g., cellulase gene primers targeting general cellulose-degrading fungi/bacteria) is in great need for study of the microbial decomposition of detritus.

More recently, stable isotope probing (SIP), for example using 13C -label substrate, coupled with rRNA-based molecular methods (e.g., DGGE and next-

generation DNA sequencing), has been shown to be able to identify the major bacterial taxa actively involving in detritus degradation (Dumont and Murrell

2005; Haichar et al. 2007; Gihring et al. 2009; Lee et al. 2011). Diverse bacterial groups have been identified as active decomposers of detritus in soil, including

Actinobacteria, Bacilli, Gamma-Proteobacteria, Sphingobacteria, and

Flavobacteria (Lee et al. 2011). Bacilli were detected mainly in the first stage, and Actinobacteria were detected throughout the incubation period (Lee et al.

2011). In a study on detritus-degrading bacteria in the aquatic environment, 13C-

DNA-SIP sequencing showed that Gamma-Proteobacteria, mainly Vibrionales and Alteromonadales, were important decomposers in the sediment and they increased in response to phytodetritus addition (Gihring et al. 2009).

2.2 Bacterial Interactions and Organic Matter Degradation

The high diversity and density of bacteria living in the environment underlie intensive interactions ongoing among species. Physiologically and ecologically, different bacteria are involved in a wide range of dynamic interactions (Hibbing et al. 2010; Wintermute and Silver 2010b; Lawrence et al.

2012). Interactions between one bacterial population and another can be synergistic (e.g., commensalism, mutualism and co-metabolism), antagonistic

(e.g., predation and competition), or sometimes neutral. Such interactions play critical roles in determining the structure and functions of microbial community

(Hibbing et al. 2010; Wintermute and Silver 2010b; Lawrence et al. 2012).

2.2.1 Antagonistic interactions among bacteria

In nature, most bacteria face a constant battle for limiting nutrients and space. Therefore, competition was proposed as the prevalent type of interactions among bacteria (Hibbing et al. 2010; Foster and Bell 2012). For example, it is widely known that soil bacteria can produce a variety of antibiotics that inhibit the growth of other bacteria (D’Costa, Griffiths and Wright 2007). In marine ecosystems, more than half of detritus-associated or coral-associated bacteria exhibited antagonistic activity against other bacteria from the same environments

(Long and Azam 2001; Grossart et al. 2004; Rypien, Ward and Azam 2010).

In a study on interactions among 72 tree-hole bacteria, Foster and Bell

(2012) showed that the majority of mixed cultures (two-species and up to 72- species mixtures) had lower productivity than the total productivity of individual bacteria growing on the nature substrate beech leaves. Although the authors defined bacterial cooperation strictly, ruling out many potential synergistic interactions, the results indicated that negative interactions such as competition dominated interactions among microbial species (Foster and Bell 2012).

Bacteria have developed strategies to compete for limiting resource. In a broad way, competition can be categorized into two groups, exploitative competition and interference competition, based on what strategies/mechanisms bacteria use. The most well-known competitive strategy is probably the production of antimicrobial compounds. This is an interference competition where the producers kill and/or suppress their competitors. Antimicrobials can mediate competition between different or related species, or between identical species

within a population, depending on what compounds are produced (Hibbing et al.,

2010). For example, it was found that a number of environmental bacteria (Long and Azam 2001; Rypien et al. 2010) produced board spectrum antibiotics that can kill/inhibit many distant species, whereas some bacteria (e.g., E. coli) produce bacteriocins that target other strains of the same species (Cascales et al. 2007). Exploitative competition strategies such as colonization of new niches and rapid uptake of the substrate are also crucial for bacterial interactions

(Hibbing et al. 2010). For example, some bacteria in the human mouth can produce adhesins that allow producers to colonize on tooth surface with priority.

In a model of bacteria competing for iron, the bacteria Pseudomonas with higher- affinity siderophore took up limiting iron more efficiently than Burkholderia and thus inhibited the growth of their competitors (Joshi, Archana and Desai 2006).

2.2.2 Synergistic interactions among bacteria

Although competition for resources among bacteria is ubiquitous in nature, there is substantial and increasing evidence of positive interactions among bacteria. Bacterial synergism can benefit all populations in the microbial assembly by promoting community productivity (Bell et al. 2005), enhancing nutrient utilization, or improving resistance to environmental stresses

(Wintermute and Silver 2010a, 2010b). Some examples of bacterial cooperation are shown in Figure 2.1.

Figure 2.1. Examples of bacterial cooperation in organic matter decomposition (Wintermute and Silver, 2010a). The image is reprinted with permission from Cold Spring Harbor Laboratory Press.

Most synergistic interactions involve metabolic cooperation or metabolite cross-feeding where the metabolic waste from one species feeds other species or two species exchange their by-products for greater fitness (Wintermute and

Silver 2010a). One classic example is bacterial syntrophy in organic matter decomposition under anaerobic conditions. For example, the anaerobic bacterium Desulfovibrio vulgaris requires sulfate as an electron accepter to grow, and the methanogen Methanococcus maripaludis consumes hydrogen as its electron donor. Although neither can grow alone in lactate medium containing no sulfate or hydrogen, they can achieve robust growth in the same medium when they grow together (Figure 2.1A) (Walker et al. 2009). Under this condition where neither bacteria can grow alone, the second bacteria need the waste (CO2 and

H2) from the first bacteria and make endergonic reaction possible for the first

bacteria (Walker et al. 2009; Wintermute and Silver 2010a). If sulfate is sufficient, methane can be converted into CO2 by the coculture of methanotrophs and sulfate-reducing bacteria, as shown in Figure 2.1B (Orphan et al. 2001).

Waste exchange among bacteria is sometimes critical for their survival.

Photolithotrophs cooperate with sulfate-reducing bacteria to degrade organic substrate when sulfate or sulfur is present while they both are inhibited by the toxic excess of sulfite (Figure 2.1C) (Searcy 2002; Wintermute and Silver 2010a).

Metabolic cooperation is common in animal digestive systems. For example,

Veillonella atypical alone cannot grow on sugar but utilize the by-product lactic acid from fermenter bacteria Streptococcus gordonii (Figure 2.1D). The removal of excess lactic acid by the former might protect the latter from low pH conditions.

Therefore, both bacteria benefit better if they grow together (Egland, Palmer and

Kolenbrander 2004).

Positive interactions among bacteria are common in the degradation of complex organic matter such as lignocellulose and xenobiotics (Sutherland

2001). For example, synergistic degradation of lignocellulosic material is evident in detritus decomposition (Newman et al. 2005), agriculture composting, and biofuel production (Lynd et al. 2002). Many bacteria can form consortia that exhibit greater efficiency in lignocellulose degradation than pure cultures (Haruta et al. 2002; Lynd et al. 2002; Wongwilaiwalin et al. 2010; Zuroff and Curtis 2012).

Lignocellulose is mainly composed of cellulose fibers, hemicelluloses and lignin that together form complex and recalcitrant architecture in the plant cell wall

(Lynd et al. 2002; Perez et al. 2002). The insolubility and complex structure make

lignocellulose hard to break down. Thus, the degradation requires multiple degrading enzymes using different mechanisms to hydrolyze and oxidize each chemical component in lignocellulose. Enzymatic synergism is very common in the degradation of lignocellulose. For example, the cocktail of endo-, exo- cellulases and β-1,4-glucosidase usually degrade cellulose faster than the sum of individual enzymes (Kostylev and Wilson 2012). The presence of xylanases can also promote lignocellulose degradation through the removal of the xylooligomers that inhibit cellulase activity (Qing, Yang and Wyman 2010). Enzyme synergism brings about more labile carbons (e.g., simple sugars) for all the bacteria in the consortia and as a result, they can grow better in mixed culture than alone.

Microbial consortia allow enzyme synergism to happen in nature as those degrading enzymes are generally produced by multiple bacteria species (Haichar et al. 2007). For maximal utilization of lignocellulose, different bacteria might produce enzymes that function in a cooperative manner, resulting in much higher efficiency in the degradation of lignocellulose. This is less likely to be achieved by any single bacteria. The cooperative behavior in the consortia benefits all.

Bacterial synergism is also very common in the co-metabolism of xenobiotics and other complex substrates (Sutherland 2001). For example, neither Pseudomonas sp. B13 nor Burkholderia xenovorans LB400 can completely degrade 3-chlorobiphenyl. However, when grown together in biofilm, they were able to convert 3-chlorobiphenyl to water and carbon dioxide (Nielsen et al. 2000). Pseudomonas sp. B13 first converted 3-chlorobiphenyl to chlorinated benzoate which was then utilized by B. xenovorans can converted to

water and carbon dioxide (Nielsen et al. 2000). The synergism was also evidenced by the formation of mixed-species biofilm colonies where cooperation for better substrate utilization might occur between the two bacteria.

2.3 Mechanisms of Bacterial Interactions

The complexity of nutrients is a critical factor that regulates bacterial interactions. Bacterial interactions might switch between antagonism and synergism as the available substrate changes. In the example above (Nielsen et al. 2000), when Pseudomonas sp. B13 and B. xenovorans LB400 were fed with the complex substrate 3-chlorobiphenyl, they cooperated to degrade the substrate to water and carbon dioxide and mainly form mixed-species biofilm colonies. In contrast, when they were fed with citrate which can be utilized by both species, they mainly formed separate pure-species colonies. The separate biofilm colonies suggest the dominance of competition for citrate by the two bacteria. Similarly, when aromatic compounds, either linuron or 3,4- dichloroaniline, were the sole carbon source available, Comamonas testosteroni WDL7 and Variovorax sp. WDL1 formed an intermixed biofilm consortium and they were able to completely degrade the substrate in a cooperative manner. However, when the labile substrate citrate was used, competition occurred and C. testosteroni WDL7 outcompeted and covered Variovorax sp. WDL1 (Breugelmans et al. 2008).

The finding of different interactions among bacteria during catabolism of labile and complex substrates suggest that the complexity of carbon source might trigger synergistic interactions among bacteria while labile carbon source

might trigger competition. The reason might be that the labile carbon source is used by most species so bacteria tend to compete with each other. In contrast, an individual bacterial species is often not able to completely digest complex substrates and thus multiple species might cooperate in a manner that benefits all partners. This might explain why bacterial cooperation is commonly found in the degradation of xenobiotics and lignocellulose. This cannot rule out the possibility of inter- and intra-competitions among species, but the effect is probably minor on the whole community compared to the benefit of synergism.

In addition to the complexity of nutrients, biofilm architecture allows organisms to be spatially organized such that even incompatible bacteria can coexist (Marsh and Boeden 2000). For example, Pseudomonas putida R1 and

Acinetobacter sp. C6 both can utilize benzyl alcohol as their carbon source, but whether they compete or cooperate depends on their relative physical positions in the biofilm (Christensen et al. 2002). When they were fed benzyl alcohol in a well-mixed environment, competition between the two bacteria resulted in the dominance of Acinetobacter sp. C6. When grown under condition that promoted biofilm formation with the same substrate, these two bacteria aggregated and formed a mixed species biofilm consortium. In the mixed biofilm, P. putida R1 shifted its substrate to benzoate, the product converted from benzyl alcohol by

Actinobacter sp. C6. In this case, the competition for primary substrate might be minor because they both grew faster during in the first few days and formed a thicker biofilm than individual species did (Christensen et al. 2002). Together with many examples, Molin et al. (2004) proposed that spatial organization of

bacteria, differential ability to utilize a substrate and continuous occurrence of nutrient gradients may contribute to a range of interactions within biofilm consortia.

Quorum sensing is one of the most important molecular mechanisms for regulating bacteria interactions (Visick and Fuqua 2005; Waters and Bassler

2005; Ryan and Dow 2008). Quorum sensing (QS) is a system by which bacteria communicate with other organisms including intra-species, inter-species and even inter-kingdom species (Lowery, Dickerson and Janda 2008). Both

Gram-positive and negative bacteria can produce and sense a variety of chemical signals known as QS molecules or autoinducers that coordinate gene expression in response to population density (Waters and Bassler 2005). The concentration of QS molecules increases as the bacterial population goes up because more signal molecules are produced. Once the population reaches a threshold density as does QS molecules, they trigger changes in the expression of other genes that affect the production of pigments, enzymes, and antibiotics, as well as motility and biofilm formation (Miller and Bassler 2001; Visick and

Fuqua 2005).

Bacteria use diverse quorum sensing molecules to communicate with their neighbors (Visick and Fuqua 2005). Well known quorum sensing molecule families identified so far include: 1) peptides, both linear and cyclic, generally used by Gram-positive bacteria, 2) AHLs: acylated homoserine lactones from

Gram-negative bacteria, 3) GBLs: Gamma-butyrolactones, 4) AI-2: furanosyl

borate diester (autoinducer-2), generally used by Gram-positive bacteria, and 5)

FAs: fatty acids such as cis-11-methyl-2-dodecenoic acid.

It is becoming evident that bacteria can sense signal molecules that they do not synthesize themselves and/or they can produce signal molecules that are sensed by other species (Brenner, You and Arnold 2008; Ryan and Dow 2008).

This mechanism provides an avenue for inter-species interactions. One species might produce QS molecules to induce other species to express more enzymes that this species itself cannot produce. For example, Veillonella atypical alone cannot grow on sugar but utilize the lactic acid produced by Streptococcus gordonii from the fermentation of starch (Figure 1D). In the coculture of the two bacteria in starch, V. atypical enhanced the production of amylase by S. gordonii that degrade starch. This process is mediated by diffusible QS signals

(Egland, Palmer and Kolenbrander 2004). The stimulation of enzyme production is also found in the microbial consortia that degrade lignocellulose. For example, proteomic analysis showed production of lignocellulolytic proteins in mixed cultures that are either not expressed or expressed at low levels in pure cultures

(Wongwilaiwalin et al. 2010; Adav et al. 2012). The enhanced protein expression in coculture might be attributed to quorum sensing among species. Thus, bacteria in a consortium are able to cooperate with others through their quorum sensing systems and perform more complex tasks than individual species cannot

(Brenner, You and Arnold 2008).

2.4 Significance of This Study

Lignocellulose degradation is essential to carbon cycling in the environment. Microbe-mediated degradation is the first and rate-limiting step in the decomposition of plant biomass. Although lignocellulolytic enzymes produced by microbes are required to degrade refractory lignocellulosic material, it remains unclear how bacteria interact to facilitate the process. The identity of bacteria within communities is now more readily available with the use of molecular-based methods. However, information about the degrading capability and ecological function of specific microbial species on detritus is very limited, not to mention complex interactions among species in a community. The lack of knowledge about microbial physiology and interactions among bacterial populations hinders our understanding of the mechanisms by which microbes decompose lignocellulose.

What is also needed is to better understand how bacteria interact during lignocellulose degradation so that more efficient methods can be developed to convert cellulose to simple sugars for biofuel production. Although cellulose is the most abundant renewable biomass for biofuel production, the low efficiency of cellulose degradation limits its economic potential (Lynd et al. 2002; Zhang and

Lynd 2004). Cellulose is structurally recalcitrant and insoluble. The main problem is the initial rate-limiting step during which cellulose is hydrolyzed into soluble sugars. One promising approach is to develop microbial consortia that are more efficient (Kato et al. 2005; Brenner et al. 2008; Wongwilaiwalin et al. 2010; Zhou et al. 2015). What would help is a better understanding of how bacteria interact

within consortia so that their performance in lignocellulose degradation can be optimized.

Given the importance of bacterial interactions in complex organic matter decomposition as discussed above, I studied lignocellulose degradation from the perspective of bacterial interactions. Results of the study provide insights that can perhaps be used to develop more efficient microbial systems for biofuel production. The goals of this study were to investigate: 1) how bacterial interactions affect lignocellulose degradation, 2) how complexity of the carbon source mediates bacterial interactions, and 3) whether the way bacteria interact relates to their phylogenetic classification.

CHAPTER III

SYNERGISTIC INTERACTIONS AMONG SALT MARSH BACTERIA IN

LIGNOCELLULOSE DEGRADATION

3.1 Abstract

Lignocellulose degradation by microbes is the first and rate-limiting step in plant biomass decomposition. However, it is largely unknown whether bacterial interactions influence the degradation process. Because lignocellulose is refractory to degradation, I hypothesize that indigenous bacteria can form consortia and their synergistic interactions enhance the utilization of lignocellulosic biomass. I isolated lignocellulolytic bacteria from salt marsh detritus and compared bacterial growth, metabolic activity and enzyme production of pure cultures to those of three-species mixed cultures. Synergistic growth was common in lignocellulose medium. As measured by OD595, 15 of 27

(56%) mixed cultures reached significantly higher density than their corresponding pure cultures. Bacterial synergism promoted metabolic activity in synergistic mixed cultures but not the maximal growth rate (µ). Bacterial synergism also promoted the production of β-1,4-glucosidase but not the production of cellobiohydrolase or β-1,4-xylosidase. The synergistic growth occurred frequently in lignocellulose medium but never in glucose medium, suggesting that bacterial synergism may depend on the structural complexity of the substrate. The results indicate that synergistic interactions among bacteria may be important in lignocellulose degradation in the natural environment.

3.2 Introduction

Microbe-mediated lignocellulose degradation is the first and probably the most critical step in the decomposition of dead plant matter (Fenchel, King and

Blackburn 2012). Lignocellulose consists of three main types of polymers including cellulose, hemicellulose, and lignin that are strongly cross-linked in plant tissue (Lynd et al. 2002). The physical and chemical structures of lignocellulose make it refractory to break down by most organisms (Lynd et al.

2002; Perez et al. 2002). Therefore, the first and the rate-limiting step is the initial hydrolysis of lignocellulose by some microbes into smaller and soluble carbohydrates that can be easily utilized by other microorganisms.

The mechanism by which microbial communities degrade lignocellulose in the natural environment is largely unknown. Although rRNA-based molecular methods have revealed extremely diverse communities of bacteria associated with detritus (Buchan et al. 2003; Das, Royer and Leff 2007; Haichar et al. 2007;

Gihring et al. 2009), the ecological functions performed by each population and the interactions among them remain mysterious. It is also unclear how bacterial interactions affect the process of lignocellulose degradation.

Bacteria are important decomposers of lignocellulose (Lynd et al. 2002) and their high density and close proximity of detritus particles underlie intensive interactions among species. In order to effectively utilize lignocellulosic detritus as a carbon source, bacteria might work in a coordinated manner to degrade the substrate, by producing complementary enzymes for example. This is supported by many observations of faster biomass degradation and higher conversion

efficiency by microbial consortia than pure cultures (Haruta et al. 2002;

Wongwilaiwalin et al. 2010). However, these studies sought to explore industrial applications for biofuel production and do not necessarily represent indigenous bacteria and their interactions in the natural environment. It is also unclear how species interactions affect bacterial activities in mixed cultures.

In the contrast to the common bacterial synergism found in lignocellulose degradation, bacterial competition has been suggested as the prevalent type of interaction among bacteria (Hibbing et al. 2010; Foster and

Bell 2012). I speculated that the way bacteria interact might depend on the complexity of substrate. When bacteria were grown on labile but limited nutrients, one bacteria might compete with other bacteria by producing antibiotics that inhibit the growth of others (D’Costa, Griffiths and Wright 2007;

Hibbing et al. 2010). When bacteria were grown on recalcitrant substrate such lignocellulose, they tend to interact synergistically to utilize the substrate, however.

In the present study, I hypothesized that detritus-associated bacteria can form consortia and synergistically degrade lignocellulose and that the way bacteria interact depends on the complexity of substrate with greater cooperation in the presence of recalcitrant substrate. To investigate bacterial interactions, I isolated lignocellulolytic bacteria from salt marsh detritus. Bacteria were grown in single and three-species mixed cultures in both lignocellulose medium and glucose medium. I compared the growth, enzyme production and metabolic activity of three-

species mixed cultures to those of pure cultures. The specific aims of this study are to determine: 1) the frequency of synergistic degradation among bacteria when grown in lignocellulose medium; 2) whether the occurrence of synergistic growth depends on the complexity of the carbon source; and 3) whether bacterial synergy affects the maximal specific growth rate, enzyme production and metabolic activity of bacteria in mixed cultures.

3.3 Materials and Methods

3.3.1 Culture media preparation

Three bacterial isolation media were used. They were composed of

Bushnell-Haas basal salt medium (Lo et al. 2009) amended with one of the three main components of lignocellulose: carboxymethyl cellulose (CMC) (0.5%), xylan

(0.5%) or lignin (0.3%). The basal salt medium was adjusted to pH 7.5 and 1%

NaCl, and pre-filtered through G/C filter (1.0 μm) before adding any organic nutrients. Zobell marine medium (HiMedia, Cat# M385) (with 1.5% agar for plates) was used to grow purified bacterial isolates.

Two media, containing either glucose or lignocellulosic compounds, were used to study bacterial interactions. The glucose medium contained basal salt medium with 0.3% glucose and 0.05% yeast extract. The lignocellulose medium, used to simulate recalcitrant carbon substrates found in nature, contained basal salt medium with 0.3% CMC, 0.2% xylan, 0.1% lignin and 0.05% yeast extract.

The relative proportion of CMC, xylan, and lignin, 3:2:1, mimics the typical lignocellulosic composition in grass material (Sun and Cheng 2002). All components in the lignocellulose medium are water soluble, producing a clear

solution that permitted measurement of bacterial growth directly by optical density (OD) in 96-well plates.

3.3.2 Bacterial isolation and classification

Natural detritus was collected from a salt marsh in Ocean Springs, MS,

U.S.A (30°23'32'' N 88°47'56''W). Potential lignocellulolytic bacteria were isolated from the detritus using the three isolation media mentioned above. After incubation for 10 - 14 days at 25 °C, representative single colonies were streaked on new agar plates of the same isolation media to obtain well-isolated colonies.

Pure single colonies were then transferred and grown on Zobell marine agar.

Based on the carbon source used during the initial isolation, the isolates were classified into three groups: cellulose-degrading (C), xylan-degrading (X), or lignin degrading bacteria (L).

To study bacterial interactions, nine bacteria with different colony morphology were randomly selected, three from each group. The nine bacteria were identified by 16S rRNA gene sequencing using universal primers 27F and

1492R (Weisburg et al. 1991; Ciric, Philp and Whiteley 2010). The nine sequences have been deposited in GenBank database (Accession Number

KJ158195-KJ158203). Putative taxonomic identities of the nine isolates were assigned to genus level using ribosomal database project (RDP) Bayesian classifier (Wang et al. 2007) with the minimum bootstrap confidence of 80%.

Two of the nine isolates with bootstrap confidence less than 80% at genus level were assigned to the family level. A phylogenetic analysis of the nine bacteria

was conducted by MEGA (Tamura et al. 2011) using neighbor-joining method

(Saitou and Nei 1987).

3.3.3 Culture preparation and growth measurements

The nine isolates were used to determine bacterial interactions in both

lignocellulose and glucose media. To prepare the isolates for growth

experiments, they were first grown in 3 ml Zobell marine broth in 100 mm x 16

mm polypropylene tubes. After 18 hr at 25°C with shaking (250 rpm), 0.2 ml of

each culture was used to inoculate 3 ml of either lignocellulose or glucose

medium and grown at 25°C with shaking. Those in lignocellulose medium were

grown for 30 hr, and those in glucose medium were grown for 24 hr. To conduct

growth experiments, seed cultures were prepared by diluting each culture to an

6 optical density (OD595) of 0.01 (ca. 10 colony formation units per mL) with either

lignocellulose or glucose medium.

Growth experiments using nine pure cultures and 27 mixed cultures were

conducted to test whether the complexity of carbon source affects bacterial

interactions. Each of the 27 mixed cultures was created by combining three pure

cultures, one from each of the three groups (C, X and L) described above. All

growth experiments started with cultures at an OD595 of 0.01 in clear flat-bottom

96-well plates. Growth experiments were conducted using 100 μL of glucose

medium or lignocellulose medium. For three-species mixed cultures, each

species contributed 1/3 of the starting volume. Each culture (pure or mixed) was

replicated in four wells in each of three plates; thus, a total of 12 replicates were

used for each pure culture and bacterial combination. Each 96-well plate also

included four wells filled with 100 μL sterile lignocellulose or glucose media as blank controls.

Bacterial cultures in growth experiments were grown at 25 °C without shaking. Cultures in lignocellulose medium were incubated for 48 h and those in glucose medium for 30 hr. Sterile growth media were optically clear at the start of each experiment and growth was determined by measuring the increase in OD595 of each culture (minus the blank controls) using a Synergy 2 microplate reader

(BioTek Instruments, Inc., Winooski, VT) over time. To determine maximal growth rates (µ, h-1), optical densities of the cultures during the exponential growth phase were log-transformed and the slope for each culture used.

3.3.4 Definition of synergistic growth

Synergistic growth is defined as having occurred when a mixed culture grew more densely than any of the three corresponding pure cultures. The densest among the three pure cultures is referred to as the reference culture.

Thus, a mixed culture is considered to exhibit synergism when it reached significantly higher density (OD595) than its reference culture. No distinction was made whether the higher density resulted from the enhanced growth of all three bacteria or just one or two in the mixed culture. Of interest was the fact that enhanced growth of the mixed cultures indicated more bacterial biomass production and greater degradation of lignocellulose, the ecological process of interest.

3.3.5 Enzyme production assay

The production of lignocellulolytic enzymes was measured using fluorometric assays adopted from widely used methods (Sinsabaugh et al. 1997,

2008; Marx, Wood and Jarvis 2001; Saiya-Cork, Sinsabaugh and Zak 2002). The assays use enzyme-specific substrates labeled with 4-methylumbelliferone

(MUB). The substrates are non-fluorescent, but fluorescence is observed when

MUB is released upon substrate hydrolysis by enzymes in the growth medium

(Hoppe 1983; Marx, Wood and Jarvis 2001). Enzyme activity is quantified by measuring the amount of fluorescence. 4-MUB-β-D-glucoside, 4-MUB-β-D- cellobioside, and 4-MUB-β-D-xyloside, respectively, were used to test the activities of β-1,4-glucosidase (EC.3.2.1.21), cellobiohydrolase (EC.3.2.1.91) and

β-1,4-xylosidase (EC.3.2.1.37). The first two degrade cellulose and the third degrades hemicellulose.

To perform assays, bacterial cultures growing in lignocellulose medium for

48 h were diluted five-fold with 50 mM MOPS buffer (pH 6.5), and 150 μL of each diluted culture were combined with 50 μl of one of the three 4-MUB-labeled enzyme-specific substrates (200 μM) in black 96-well plates. After incubation at

25°C in the dark for 1 h with shaking at 500 rpm, fluorescence in each well was measured using a Synergy 2 Bio-Tek microplate reader (365ex, 450em). The amount of lignocellulolytic enzyme in each culture was calculated based on the fluorescence of the sample relative to that of a standard. The standard contained

150 μL of diluted bacterial culture and 50 μL of 10 μM MUB. Blank controls contained 150 μL diluted culture and 50 μL buffer. Substrate controls contained

150 μL buffer and 50 μL substrates. The enzyme activity in each sample expressed in nmol∙hr-1∙mL-1 was calculated using the formula below:

푆퐹 − 퐵퐶퐹 − 푆퐶퐹 푛푚표푙 1 1 × (10 × 0.05 푚푙) × × 푀퐹 − 퐵퐶퐹 푚푙 0.15 푚푙 1 ℎ where SF is the sample fluorescence, BCF the blank control fluorescence, SCF the substrate control fluorescence, and MF the MUB standard fluorescence.

Results were multiplied by five to adjust for the initial five-fold sample dilution. Each sample, standard, blank control and substrate control was replicated four times.

The experiment was repeated twice.

3.3.6 Microbial metabolic activity measurement

To test whether bacterial interactions enhance microbial activity, the metabolic activity of bacterial cultures was measured using the 2,3,5- triphenyltetrazolium chloride (TTC) assay (Gabrielson et al. 2002; Burmølle et al.

2006). The colorless TTC is enzymatically reduced by metabolically active bacteria to red 1,3,5-triphenylformazan (TPF) that can be quantified by measuring sample absorbance at 490 nm (Gabrielson et al. 2002). After growth for 30 hr in lignocellulose medium, 100 μL of each culture was combined with 25

μL of substrate solution (0.05% TTC and 1.5% glucose) and incubated at 25°C in the dark. Glucose was included to improve sensitivity (da Silva et al. 2008).

Sterile culture medium with the same amount of substrate solution served as negative control. Absorbance was measured at 0 and 3 h, and the difference in absorbance (ΔA490) was used as a measure of metabolic activity. I also

-1 compared the growth-specific metabolic activity (ΔA490 ∙ OD595 ) between mixed

and pure cultures. The experiment was repeated twice using eight replicates each time.

3.3.7 Statistical analysis

One-way ANOVA and Tukey HSD test were performed to test whether mixed cultures grew better than their corresponding pure cultures. The effects of synergism on growth, maximal growth rate, and metabolic activity were analyzed by one-way ANOVA, followed by Tukey HSD test. Due to the non-normal distribution of enzyme production data, the non-parametric Kruskal-Wallis test was used to test the effect of synergism on enzyme production. All statistical tests were performed in RStudio (www.rstudio.org).

3.4 Results

3.4.1 Bacterial Isolates Classification

The nine lignocellulolytic bacteria isolated from a natural salt marsh in

Ocean Springs, MS, U.S.A. were taxonomically diverse (Figure 3.1). Two isolates were classified as Alpha-Proteobacteria including Labrenzia sp. (L1) and

Paracoccus sp. (L3). Six isolates belonged to Gamma-Proteobacteria, including two Vibrio spp. (X3 and C3 ), one Gallaecimonas sp. (L2), one Hahella sp. (C1) and two presumptively new isolates (X1 and C2) that could not be assigned to genus level. These two isolates were thus assigned to the family level and classified as Vibrionaceae bacterium (X1) and Alteromonadaceae bacterium

(C2). The ninth isolate was Jonesia sp. (X2), belonging to Actinobacteria.

Figure 3.1. Phylogenetic relationship of nine bacterial isolates used in this study. Bold text with black diamond indicates the nine bacteria isolates used. The letter and number following the name of the bacterium denote the substrate used in isolating each bacterium, C for cellulose, L for lignin and X for xylan. Bootstrap values are shown as the percentage of 1,000 replicates when greater than 50%. The horizontal bar represents nucleotide substitutions per sequence position. GenBank accession numbers are in parentheses.

Figure 3.2. Growth comparison of mixed cultures to their corresponding pure cultures in lignocellulose medium. Growth of each mixed culture (black bars) is compared to its three corresponding pure cultures (gray bars alongside) that made up this mixed culture. N = 12. Error bar is one standard deviation. Asterisks indicate significantly greater growth compared to pure cultures. One-way ANOVA (F35, 396 = 151.77, P < 0.001) followed by Tukey HSD test (P < 0.05).

3.4.2 Synergistic growth in three-species mixed cultures

Synergistic growth among three-species mixed cultures was common in lignocellulose medium (Figure 3.2). Synergistic growth is defined as having occurred when a mixed culture grew better than any of the three corresponding pure cultures. Fifteen among the 27 possible mixed cultures exhibited significantly greater growth when compared to their corresponding pure cultures

(Figure 3.2). In fact, all but one of these 15 mixed cultures reached higher densities (OD595 > 0.209) than all the pure cultures. The experiment was repeated twice more with similar results (data not shown). Hereafter, the 15 three-species combinations are designated as synergistic mixed cultures because those mixed cultures showed greater growth than their corresponding cultures that made up each of them. The remaining 12 mixed cultures are designated as non-synergistic mixed cultures.

3.4.3 Carbon source-dependent synergistic interaction

To determine whether the synergistic growth was substrate dependent, I repeated the mixed-culture experiment using glucose as the sole carbon source.

For comparing bacterial growth in glucose medium and lignocellulose medium, I plotted the growth of each mixed culture against the growth of its reference culture (the pure culture with the greatest growth among the three that made up the mixed culture) (Figure 3.3). Although generally pure cultures grown in glucose medium reached higher densities than the same cultures in lignocellulose medium, none of the 27 mixed cultures in glucose medium reached higher density than their reference cultures. All had optical densities

below the isometric line that indicates equal growth between mixed and reference cultures (Figure 3.3). This result suggests the dominance of negative interaction or competition among three species in mixed cultures when grown in glucose medium. In contrast, most of the mixed cultures (21 in 27) reached densities above the isometric line when grown in lignocellulose medium (Figure

3.3). The results show that bacterial synergism occurred frequently in lignocellulose medium but never in glucose medium, suggesting that bacterial synergistic growth was dependent on the structural complexity of the carbon source.

Figure 3.3. Comparison of mixed culture growth in lignocellulose and glucose media. Each data point is the OD595 of one mixed culture (mean ± SD, n =12) plotted against the OD595 of its reference culture (the pure culture with the greatest growth among the three that made up the mixed culture). The isometric line represents equal growth between mixed cultures and their reference cultures.

3.4.4 Bacterial growth and activity during lignocellulose degradation

To explore the potential mechanism for bacterial synergism in lignocellulose medium, I compared the maximal growth rate and metabolic activity among the three culture groups: synergistic mixed cultures (the 15 mixed cultures that exhibited synergistic growth), non-synergistic mixed cultures, and pure cultures. Synergistic mixed cultures reached the highest cell density among the three groups (F2, 33 = 26.78, P < 0.001) (Figure 3.4A) but did not grow faster during the exponential phase (Figure 3.4B). The mean growth density of synergistic mixed cultures (OD595 = 0.280) was more than twice that of pure cultures (OD595 = 0.135) and 1.6 times that of non-synergistic mixed cultures

(OD595 = 0.176) (Figure 3.4A). In terms of maximal growth rate, there was, however, no significant difference between synergistic mixed cultures (mean =

0.429 h-1) and pure cultures (mean = 0.430 h-1) (Figure 3.4B). Interestingly, although they grew less than synergistic mixed cultures (Figure 3.4A), non- synergistic mixed cultures had higher maximal growth rate (mean = 0.542 h-1) than synergistic mixed cultures and pure cultures (F2, 33 = 8.29, P = 0.001)

(Figure 3.4B).

Bacteria in synergistic mixed cultures had higher metabolic activity (mean

ΔA490 = 0.186) than those in pure cultures (mean ΔA490 = 0.096) (F2, 33 = 3.79, P =

0.033) but not significantly higher than non-synergistic mixed cultures (mean

ΔA490 = 0.132) (Figure 3.4C). Although synergistic mixed cultures had higher metabolic activity than pure cultures, there was no significant difference in the

-1 growth-specific metabolic activity (ΔA490 OD595 ) among the three groups after normalizing for cell density (data not shown).

Figure 3.4. Effect of synergy on bacterial growth, specific growth rate and metabolic activity. Gray bar is the mean value of each group. Error bars represent 95% confidence interval. Different letters indicate significant difference (P < 0.05). NC = non-synergistic mixed cultures, n = 12; PC = pure cultures, n = 9; SC = synergistic mixed cultures, n = 15.

Figure 3.5. Effect of synergy on bacterial production of lignocellulolytic enzymes. BG: β-1,4-glucosidase; CBH: cellobiohydrolase; BX: β-1,4-xylosidase. Gray bar is the mean enzyme activity of each group. Error bars represent 95% confidence interval. Different letters indicate significant difference (P < 0.05). NC = non- synergistic mixed cultures, n = 12; PC = pure cultures, n = 9; SC = synergistic mixed cultures, n = 15.

3.4.5 Bacterial production of lignocellulolytic enzymes

Bacterial synergism promoted the production of β-1,4-glucosidase (BG) but not the production of cellobiohydrolase (CBH) or β-1,4-xylosidase (BX)

(Figure 3.5). The mean BG activity among synergistic mixed cultures (133.5 nmol h-1 mL-1) was over three times that of pure cultures (43.7 nmol h-1 mL-1), and 2.6 times that of non-synergistic mixed cultures (52.3 nmol h-1 mL-1) ( χ2 =13.87, df =

2, P = 0.001) (Figure 3.5A). Activities of CBH and BX for the synergistic group were slightly higher than two other groups but not significantly different from them

(Figure 3.5B and C). When adjusted for cell density, there was no significant

-1 -1 -1 difference in production of the three enzymes tested (nmol h mL OD595 ) among synergistic mixed cultures, non-synergistic growth and pure culture groups (data not shown).

3.5 Discussion

I found that three-species mixed cultures composed of taxonomically diverse bacteria frequently exhibited synergistic growth during lignocellulose degradation. The results suggest that bacterial synergism may be important to detritus degradation in the salt marsh ecosystem where those bacteria were isolated. Whether bacterial synergism takes place appeared to depend on the chemical complexity of the carbon source. In lignocellulose medium, most of the mixed cultures exhibited synergistic growth, but none exhibited synergistic growth in glucose medium. The results support the hypothesis that the complexity of carbon source plays a role in determining bacterial interactions in the environment.

Several mechanisms that promote synergistic growth have been proposed. For example, synergistic growth can result when multiple species produce complementary enzymes and take part in metabolite cross feeding

(Wintermute and Silver 2010b; Kostylev and Wilson 2012). Due to the complexity and recalcitrance of lignocellulosic substrate, the complete degradation of lignocellulose requires multiple enzymes (Kostylev and Wilson,

2012; Van Dyk and Pletschke, 2012). Furthermore, enzyme cocktails containing cellulases, xylanases and lignin peroxidases that are produced by multiple species can significantly enhance lignocellulose degradation rate (Lynd et al.

2002; Guevara and Zambrano 2006; Kostylev and Wilson 2012; Van Dyk and

Pletschke 2012). In addition to making enzyme cocktails where enzymes act synergistically, mixed cultures can also promote the production of enzymes that produce simple sugars promoting bacterial growth. In the present study, β-1,4- glucosidase (BG) activity was more than three folds higher in synergistic mixed cultures than in pure cultures (Figure 3.5A). The higher β-1,4-glucosidase activity likely produced more glucose that contributed to the enhanced growth of mixed cultures. Thus, it appears that mixed cultures in lignocellulose medium not only produced enzymes that were complementary but also produced more active enzymes so that refractory substrates can be degraded more effectively compared to pure cultures.

Another mechanism that promotes bacterial synergism is metabolite cross feeding that allows bacteria to utilize complex substrate in a cooperative manner

(Flint et al. 2007; Wintermute and Silver 2010b). During late growth, some

species in a mixed culture may produce metabolites that are toxic to themselves but are used by others. In this case, mixed cultures can alleviate problems of feedback regulation and metabolite repression present in pure cultures (Zuroff and Curtis 2012). In the present study, synergistic mixed cultures reached higher densities than pure cultures but their maximal growth rate during exponential growth was not higher (Figure 3.4A and B). One possible explanation is that after exponential growth when the primary nutrients are depleted, bacteria in mixed cultures can continue to grow by utilizing metabolites produced by others. As a result, mixed cultures reached higher growth density than pure cultures that had no partners to exchange metabolites with. This was supported by observations of higher metabolic activity in the synergistic mixed cultures after the exponential growth (Figure 3.4C).

Although most mixed cultures (15 out of 27) in lignocellulose medium exhibited synergistic growth and degraded more lignocellulose than pure cultures, some mixed cultures (12) did not show significant synergistic growth under the same culture condition. This suggests that specific combination of bacteria may be important. Some combinations achieved enhanced growth, while others exhibited competition perhaps due to resource overlap or competing for metabolic capabilities of neighboring bacteria (Hibbing et al. 2010; Freilich et al.

2011; Elias and Banin 2012; Kinkel et al. 2014).

It was striking that none of the 27 mixed cultures exhibited synergistic growth in glucose medium. The likely explanation is that cooperation serves no purpose in the presence of a labile carbon source such as glucose, and thus, competition

is prevalent. The concept that substrate complexity regulates the type of bacterial interaction is supported by published studies. For example, Long and Azam

(2001) and Grossart et al. (2004) reported that antagonism was common among bacteria found in marine ecosystems. A possible explanation lies in the marine

(Zobell) agar used in the laboratory studies. The limited amounts of peptone and yeast extract in the medium may have spurred competition among bacteria to result in the common antagonism observed. Similarly, both Nielsen et al. (2000) and Breugelmans et al. (2008) showed that when citrate was used as the carbon source, bacteria competed for the labile nutrient and usually formed separated biofilm colonies. In contrast, when complex benzyl compounds were the only carbon source in the medium, bacteria degraded the refractory substrate more efficiently than pure cultures and usually formed a mixed-species biofilm. This suggests the possibility that bacteria grown together in the glucose-limiting medium in the present study might compete for the labile substrate by producing antimicrobial chemicals to inhibit competitors, resulting in the predominance of antagonistic interactions.

3.6 Conclusion

In conclusion, my results show that indigenous bacteria can form consortia that degrade lignocellulose synergistically. My study also demonstrates that bacterial synergism is dependent on the complexity of the substrate. When lignocellulose is the only carbon source, bacteria tend to interact synergistically to degrade the complex substrate. When glucose is the only carbon source and in a limited amount, bacteria compete for the labile substrate resulting in the

predominance of antagonistic interactions. To better understand the mechanism, additional studies on the relationship between the chemical complexity of the substrate and bacterial compositions in the consortia are needed. Currently, it is unclear how different species contribute in the consortia. Some may produce enzymes that degrade a particular component of lignocellulose more effectively, while others may produce stimulatory exudates or quorum sensing molecules that coordinate interactions among bacteria. A better mechanistic understanding may help us develop efficient microbial consortia for lignocellulose degradation and biofuel production.

CHAPTER IV

COMPLEX POLYSACCHARIDES MEDIATE ANTAGONISTIC INTERACTIONS

AMONG BACTERIA DEGRADING CELLULOSE AND XYLAN

4.1 Abstract

Bacterial competition for resources is common in nature, but positive interactions among bacteria are also evident, especially during their degradation of lignocellulose. I speculate that the structural complexity of the substrate might play a role in mediating bacterial interactions. In this study, I tested the hypothesis that dependence of bacteria on complex polysaccharides as the main carbon source reduces the frequency of antagonistic interactions among them when growing on a more labile substrate such as glucose. Results from all possible pairwise interactions among 36 bacteria isolated from salt marsh detritus showed that the frequency of antagonistic interactions was significantly lower on carboxymethyl cellulose (CMC)-xylan medium (7.6%) than on glucose medium (15.7%). The structures of the two interaction networks were also different. Although most of the antagonistic interactions observed (78%) occurred in both media, there were 22 that occurred only when bacteria were grown on

CMC-xylan, indicating that some antagonistic interactions were substrate- specific. I also found different frequencies of antagonism among phylogenetic groups. Firmicutes and Gamma-Proteobacteria were the most antagonistic, and they tend to antagonize Bacteroidetes and Actinobacteria, the most susceptible groups. The present study suggests that complex polysaccharides mediate

bacterial interactions and that bacteria community might experience dynamic interactions as their nutrient condition changes in the environment.

4.2 Introduction

In natural environments, bacteria usually compete for limited space and nutrients (Hibbing et al., 2010; Foster and Bell, 2012). One of the most common mechanisms for the bacterial competition is the production of antibiotic-like agents that inhibit the growth of other bacteria (Mangano et al. 2009; Hibbing et al. 2010; Rypien, Ward and Azam 2010). Indeed, antagonistic interactions within a community have frequently been observed among bacteria in both aquatic and terrestrial environments (Grossart et al. 2004; Mangano et al. 2009; Rypien,

Ward and Azam 2010; Prasad et al. 2011; Vetsigian, Jajoo and Kishony 2011).

Studies have shown that the majority of bacterial isolates antagonized at least one other bacteria within the same community and some isolates even antagonized most of the others (Long and Azam 2001; Grossart et al. 2004;

Mangano et al. 2009; Rypien, Ward and Azam 2010; Aguirre-von-Wobeser et al.

2013).

Although antagonistic interactions within bacterial communities are common, positive interactions have also been observed. Examples include bacterial syntrophy in complex organic matter degradation (Schink 1997) and bacterial consortia that synergistically degrade lignocellulose (Kato et al. 2005;

Wongwilaiwalin et al. 2010; Jiménez, Korenblum and van Elsas 2014). However, studies that contrast bacterial antagonism and synergy are limited; thus, factors that cause bacteria to have such opposite interactions remain unclear.

I propose that the complexity of carbon source is an important factor that mediates bacterial interactions. The antagonistic interaction was usually found when bacteria were grown using labile nutrient media such as Zobell marine agar containing peptone and yeast extract (Long and Azam 2001; Grossart et al.

2004; Rypien, Ward and Azam 2010). It is likely that bacteria compete when labile nutrients are available but in limiting quantities. On the other hand, less antagonistic interactions might be expected when bacteria rely on recalcitrant carbon sources such as lignocellulose that require multiple complementary enzymes to breakdown (Haruta et al. 2002; Wongwilaiwalin et al. 2010). It is thus of interest to know how bacterial interactions might change within a community when the complexity of the carbon source changes over time in natural environments (Jaeger et al. 1999; Marschner and Kalbitz 2003; Carrero-Colón,

Nakatsu and Konopka 2006; Mindl et al. 2007).

In the present study, I tested the hypothesis that complex polysaccharides such as cellulose or xylan reduce the frequency of antagonistic interactions among lignocellulolytic bacteria. I isolated bacteria from salt marsh detritus that can use both glucose and cellulose/xylan as carbon sources. Bacteria were identified and tested for pairwise antagonistic interactions using both media. The main goals of this study were to test whether the frequency of bacterial antagonism is dependent on the structural complexity of the substrate and to determine whether the structure of interaction networks is affected by the change of substrate complexity.

4.3 Materials and Methods

4.3.1 Bacteria isolation and culture media

Cellulose- and xylan-degrading bacteria were isolated from salt marsh detritus collected in Ocean Springs, MS, U.S.A (30°23'32'' N 88°47'56''W), according to the method previously described (Deng and Wang 2016). Briefly, three selective media were used to isolate bacteria. Each contained basal salt medium (Bushnell-Haas broth with 1% NaCl) amended with one of three complex substrates as the sole carbon source: carboxymethyl cellulose (CMC) (0.5%), xylan (0.5%) or lignin (0.3%). After incubation for 10 - 14 days at 25 °C, single colonies with different morphology were streaked on fresh agar plates to obtain well-isolated colonies. Pure single colonies were then transferred and grown on

Zobell marine agar. Because CMC, xylan or lignin was the only source of carbon in each of the isolation media, bacteria that grew were classified into three groups: cellulose-, xylan- and lignin-degrading bacteria. Many bacterial isolates, including those isolated originally on lignin medium, were able to grow on CMC and xylan as the only carbon source (data not shown).

To study bacterial antagonistic interactions, simple carbohydrate medium

(glucose medium) and complex polysaccharide medium (CMC-xylan medium) were used. The glucose medium contained the basal salt medium mentioned above with 0.3% glucose and 0.05% yeast extract. The CMC-xylan medium contained the basal salt medium with 0.2% CMC, 0.1% xylan and 0.05% yeast extract. Agar (1.5%) was included when solid media were needed.

4.3.2 Sequencing of 16S rRNA genes and bacteria identification

Among isolates that were able to grow on both glucose medium and CMC- xylan medium, 36 were randomly chosen for 16S rRNA sequencing and for the bacterial interaction study. The 16S rRNA gene of each bacteria was PCR amplified using the universal primers 27F and 1492R (Weisburg et al. 1991). The first half (approximately 700 – 900 bp) of each PCR product was sequenced by using primer 27F for bacteria identification (Lo Giudice et al. 2007). Sequences were submitted to GenBank under the following accession numbers: KT356815 to KT356850. Putative taxonomic identities were assigned to bacteria isolates using RDP Bayesian classifier (Wang et al. 2007) with the minimum bootstrap confidence of 80%.

4.3.3 Antagonistic interaction assay

Bacteria isolates were first grown in 3 ml Zobell marine broth in 100 x 16 mm polypropylene tubes. After 18 hr at 25°C with shaking (250 rpm), 0.15 ml of each culture was used to inoculate 3 ml of either glucose or CMC-xylan medium to obtain seed cultures. Bacterial cultures were incubated with shaking (250 rpm) at 25°C for 20 hrs (glucose medium) or 30 hrs (CMC-xylan medium). Bacterial isolates grown in each medium were used in antagonistic assays on either glucose or CMC-xylan agar plates.

Figure 4.1. An example of bacterial antagonistic interactions detected by the Burkholder spot-on-lawn method. Antagonism is detected by the presence of growth inhibition or halo around a colony. The red color is due to pigments produced by the bacteria.

Pairwise antagonistic interactions among bacteria were determined using the Burkholder “spot-on-lawn” method (Burkholder, Pfister and Leitz 1966;

Cordero et al. 2012). Briefly, all seed cultures were adjusted to a final optical density (OD595) of 0.5 with either fresh glucose or CMC-xylan medium. To make the bacterial lawn, each seed culture was spread evenly on either glucose plate or CMC-xylan plate using two sterile cotton swabs saturated with the culture.

Subsequently, 1 µl of each bacterial culture (OD595 = 0.5) was transferred onto the bacterial lawn in a grid pattern using a slot-pin replicator (Cat# VP408S, V&P

Scientific, Inc.). Sterile media were also spotted onto bacterial lawns as negative controls. Up to 25 bacterial cultures (referred to as sender bacteria) were tested per Petri dish, leaving enough space between cultures to prevent interference

among sender bacteria. Presence or absence of growth inhibition was determined by looking for the presence of halos on agar plates three days after incubation at 25°C for glucose plates and after five days for CMC-xylan plates

(for example, see Figure 4.1). To ensure that interactions were assigned reliably, all pairwise interactions (36 x 36 or 1,296 interactions) were tested at least three times. The bacteria used to grow bacterial lawns are defined as receiver bacteria because they receive signals produced by the sender bacteria spotted on the lawns (Aguirre-von-Wobeser et al. 2013).

4.3.4 Interaction network and analysis

For data visualization, interaction networks for bacteria grown on the two media were constructed from binary interaction data (0 for no interaction and 1 for an interaction) using the “igraph” network analysis package (Csárdi and

Nepusz 2006) in RStudio (www.rstudio.org). Interaction density (D) was calculated as the number of detected interactions (linkages) divided by all possible pairwise interactions (linkages/N2), where N is the number of bacterial isolates (36 in this case). Thus, the interaction density indicates the frequency of interactions in a network (Vetsigian, Jajoo and Kishony 2011). The sender degree (out-degree) of a bacterium is the number of isolates it inhibits, while the receiver degree (in-degree) is the number of other isolates that inhibit the bacterium (Perez-Gutierrez et al., 2013). Therefore, a bacterium with a large sender degree is more antagonistic while one with a large receiver degree is more susceptible to growth inhibition by other bacteria.

Table 4.1

Identify of Bacterial Isolates for Study of Antagonistic Interactions

GenBank Phylum Accession / Class Isolates Genus Conf * Family # JDC15 Alteromonus 67% Alteromonadaceae KT356815 JDC17 Hahella 100% Hahellaceae KT356816 SDX11 Halomonas 74% Halomonadaceae KT356840 JSL1 Microbulbifer 100% Alteromonadaceae KT356832

JSC26 Microbulbifer sp. 100% Alteromonadaceae KT356830

- JDX4 Pseudoxanthomonas 100% Xanthomonadaceae KT356825 SDL6 Pseudomonas 97% Pseudomonadaceae KT356837 SDL8 Pseudomonas 96% Pseudomonadaceae KT356838

Gamma SDX10 Vibrio 100% Vibrionaceae KT356839

Proteobacteria JDX1 Vibrio 100% Vibrionaceae KT356822 SDX13-3 Vibrio 66% Vibrionaceae KT356841 SDX17 Vibrio 100% Vibrionaceae KT356844 SDX6 Vibrio 100% Vibrionaceae KT356845 Gammaproteobacte- SSL4-1 Gallaecimonas 100% KT356848 ria incertae sedis

JDC20 Thalassospira 100% Rhodospirillaceae KT356818 JDC5 Erythrobacter 100% Erythrobacteraceae KT356821

- JDX18 Sphingopyxis 100% Sphingomonadaceae KT356824 SDL1-1 Paracoccus 100% Rhodobacteraceae KT356835

Alpha SDL2 Rhizobium 100% Rhizobiaceae KT356836 SSL1-1 Labrenzia 93% Rhodobacteraceae KT356847

Proteobacteria SSL7 Labrenzia 100% Rhodobacteraceae KT356849

JDX4 -1 Bacillus 100% Bacillaceae KT356826 JDXE1 Bacillus 100% Bacillaceae KT356827 JDXE19 Bacillus 100% Bacillaceae KT356828 SDX14-3 Bacillus 100% Bacillaceae KT356842

Firmicutes Firmicutes SDXE4Firmicutes Paenibacillus 100% Paenibacillaceae KT356846

- JDC19 Leeuwenhoekiella 100% Flavobacteriaceae KT356817

JDC4 Zunongwangia 100% Flavobacteriaceae KT356820 JSC29 Flavobacterium 100% Flavobacteriaceae KT356831

etes

Bacteroi d SDC20-1 Algoriphagus 100% Cyclobacteriaceae KT356834

Table 4.1 (continued).

GenBank Phylum Accession / Class Isolates Genus Conf * Family #

JDC27-1 Demequina 99% Demequinaceae KT356819 JDX11 Jonesia 100% Jonesiaceae KT356823 JSC24 Micrococcus 100% Micrococcaceae KT356829 JSX2 Gordonia 100% Nocardiaceae KT356833 SDX16-2 Curtobacterium 100% Microbacteriaceae KT356843

Actinobacteria SSX1 -2 Micromonospora Actinobacteria 99% Micromonosporaceae KT356850

* Confidence level refers to the bootstrap confidence of the bacterial identity at the genus level according to the RDP

Bayesian classifier. Confidence level greater than 80% is considered as a reliable assignment of the taxon.

4.4 Results

4.4.1 Classification of cellulose/xylan-degrading bacteria

Thirty-six isolates that were able to grow using CMC and/or xylan as sole carbon sources were accurately assigned to the family level with 100% confidence. All but three could be assigned a genus with high confidence (>

93%). Three isolates, JDC15, SDX11, and SDX13-3, are likely novel species

(Table 4.1). The 36 isolates were classified into five phyla or classes, including

14 Gamma-Proteobacteria (39%), seven Alpha-Proteobacteria (19%), six

Actinobacteria (17%), five Firmicutes (14%), and four Bacteroidetes (11%) (Table

4.1). At the genus level, Vibrio spp. (5 isolates) were the most abundant, followed by Bacillus spp. (4 isolates). Except for SDX14-3 and JDX4-1, none shared identical 16S rRNA gene sequences.

Figure 4.2. Antagonistic interaction networks among lignocellulolytic bacteria growing on glucose medium (top) and on CMC-xylan medium (bottom). Each circle represents a bacterial isolate. The size of the circle is proportional to the number of the linkages (interactions) the bacterial isolate has with other bacteria. Arrows indicate the direction of interactions from the sender bacteria (bacteria spotted on the bacterial lawn producing chemical signals) to receiver bacteria forming the lawn and receiving signals. Red: Firmicutes; Cyan: Gamma-Proteobacteria; Pink: Actinobacteria; Yellow: Alpha-Proteobacteria; Green: Bacteroidetes.

4.4.2 Interaction network analysis

Interaction networks constructed using pairwise interaction data among the 36 bacteria isolates show antagonistic interactions occurring more frequently when bacteria grew on glucose medium than on CMC-xylan medium (Figure

4.2). There were 204 antagonistic interactions among bacteria grown on glucose medium with an interaction density of 0.157, indicating that 15.7% of all possible interactions were antagonistic. The frequency of antagonistic interactions was lower when the bacteria grew on CMC-xylan medium with 98 antagonistic interactions and an interaction density of 0.076. For each isolate, the number of antagonistic interactions on CMC-xylan medium was about one-half of that on glucose medium and confirmed to be statistically different using a paired t-test

(t35 = 4.935, P < 0.001).

In addition to different frequencies of negative interactions among bacteria in the two media, the two interaction network structures were also different between bacteria grown on simple and refractory substrates (Figures 4.2 and

4.3). Among the 98 antagonistic interactions observed on the CMC-xylan medium, 76 (78%) occurred also on the glucose medium, but 22 did not. Among the 204 antagonistic interactions observed on the glucose medium, 128 (63%) did not take place on the CMC-xylan medium. The occurrence of growth inhibitions that occurred only on one medium supports the notion of different interaction network structures among bacteria on the two media.

Figure 4.3. Comparison of the number of antagonistic interactions among bacteria grown on glucose and CMC-xylan media.

Table 4.2

Three Bacteria Exhibiting the Most Frequent Antagonism toward Other Bacteria

Glucose medium CMC-xylan medium Isolates Total Inhibiting Inhibited Total Inhibiting Inhibited

interactions others by interactions others by Vibrio sp. 33 33 0 27 27 0 JDX1 Hahella sp. 24 19 5 15 10 5 JDC17 Bacillus sp. 25 23 2 15 11 4 JDXE19

It is interesting to note that the three most antagonistic isolates (Vibrio sp.

JDX1, Hahella sp. JDC17 and Bacillus sp. JDXE19) inhibiting the largest number of other isolates (Table 4.2 and Figure 4.2) also had the most interactions

(linkages) with others. They alone accounted for 82 of 204 antagonistic interactions (40%) in glucose medium and 57 of 98 (58%) in CMC-xylan medium.

The dominance of the three bacteria suggests that they may be potential keystone species in the salt marsh bacterial community I sampled.

Figure 4.4. Relationship between bacterial antagonism and susceptibility. Each data point represents the sender degree versus receiver degree of a bacterial isolate. Sender degree is the number of bacterial isolates inhibited by this isolate. Receiver degree is the number of bacterial isolates to which the isolate is susceptible to in terms of growth inhibition.

In general, there was an inverse relationship between bacterial antagonizing ability and susceptibility. Most bacteria inhibited fewer than 10 others (Figure 4.4). Those that inhibited other bacteria more frequently (high sender degree) were themselves less susceptible to inhibition, while those that had low sender degree usually were more susceptible to others (high receiver degree). This was more obvious when bacteria grew on glucose medium (Figure

4). Interestingly, bacteria grown on CMC-xylan medium were generally less

susceptible to inhibition by others with most bacteria being inhibited by five or fewer other isolates (receiver degree ≤ 5) (Figure 4.4). My results show that highly antagonistic bacteria were generally more resistant to inhibition by others while susceptible bacteria were less antagonistic to others.

Table 4.3

Frequency of Antagonistic Interactions among Bacterial Groups Grown on

Glucose Medium

As Receiver

Bacterial Groups Act. Bact. Firm. Alpha. Gamma.

Act. 6% 4% 0% 0% 1%

Bact. 13% 0% 0% 4% 0% Firm. 50% 50% 8% 31% 22% Alpha. As Sender 10% 14% 0% 14% 6% Gamma. 36% 34% 13% 35% 15%

Sender refers to bacterial isolates spotted on bacterial lawns sending chemical signals. Receiver refers to bacterial isolates spread on plates for the lawn and receiving signals. Frequency = A / (n1 x n2) x 100%, where A is the number of antagonistic interactions between two groups and n1 and n2 are the numbers of bacterial isolates as senders and receivers, respectively. The scale of data is colored from yellow (low frequency) to dark-orange (high frequency). Act:

Actinobacteria; Bact: Bacteroidetes; Firm: Firmicutes; Alpha: Alpha-Proteobacteria; Gamma: Gamma-Proteobacteria.

4.4.3 Interaction patterns among bacterial phylogenic groups

Among the sender bacterial groups, Firmicutes were the most antagonistic followed by Gamma-Proteobacteria (Table 4.3). When grown on glucose medium, Firmicutes antagonized members of Actinobacteria and Bacteriodetes at a frequency of 50%. Growth inhibition by Gamma-Proteobacteria toward 55

members of other bacterial groups was also frequently observed in glucose medium. Gamma-Proteobacteria displayed growth inhibition against members of

Actinobacteria, Bacteroidetes or Alpha-Proteobacteria with the frequencies of 36,

34, and 35%, respectively (Table 4.3). As receiver bacteria, Firmicutes were the least susceptible to inhibition by others with no growth inhibitions by

Actinobacteria, Bacteroidetes or Alpha-Proteobacteria (Table 4.3). Only 8% of the pairwise interactions among Firmicutes and 13% between Firmicutes and

Gamma-Proteobacteria were antagonistic.

Table 4.4

Frequency of Antagonistic Interactions among Bacterial Groups Grown on CMC- xylan Medium

As Receiver

Bacterial Groups Act. Bact. Firm. Alpha. Gamma.

Act. 3% 0% 0% 0% 0%

Bact. 0% 0% 0% 0% 0%

Firm. 27% 35% 8% 11% 13%

As Sender Alpha. 2% 4% 0% 2% 1%

Gamma. 14% 14% 10% 14% 11%

receivers, respectively. The scale of data is colored from yellow (low frequency) to dark-orange (high frequency). Act:

Actinobacteria; Bac: Bacteroidetes; Firm: Firmicutes; Alpha: Alpha-Proteobacteria; Gamma: Gamma-Proteobacteria.

In CMC-xylan medium, similar patterns of group-group interactions were observed but with a lower frequency of antagonistic interactions compared to growth on glucose medium. As sender bacteria, Firmicutes and Gamma-

Proteobacteria were the most antagonistic among the five bacterial groups

(Table 4.4). As is for glucose medium, Firmicutes were also the most inhibitory toward members of Bacteroidetes and Actinobacteria. Actinobacteria and

Bacteroidetes were the least antagonistic but the most susceptible to growth inhibition by other bacteria (Table 4.4).

4.5 Discussions

Bacterial communities encounter dynamic changes in nutrient availability including carbon sources (Jaeger et al. 1999; Marschner and Kalbitz 2003;

Carrero-Colón, Nakatsu and Konopka 2006; Mindl et al. 2007). For example, simple carbohydrates such as glucose and fructose may be more available from plants during the growing season while more complex polysaccharides such as cellulose and xylan in the winter time (Hocking 1989; Lawlor 1995; Kritzberg,

Langenheder and Lindström 2006; Caffall and Mohnen 2009). Both simple carbohydrates and complex polysaccharides are important carbon and energy sources for heterotrophic bacteria. However, because simple sugars are utilized more rapidly than complex polysaccharides, the composition of carbon sources changes over time. The goal of the present study was to determine whether the chemical complexity of the carbon source affects bacterial interactions and thus

potentially influences the structure and activity of bacterial community in natural environments.

Overall, I found that complex carbohydrates reduce the frequency of antagonistic bacterial interactions. The frequency of antagonistic interactions among bacteria growing on glucose was more than twice that of bacteria relying on CMC-xylan as the source of carbon. Why do bacteria inhibit each other more frequently in glucose than in CMC-xylan medium? One possible reason is that glucose can be utilized readily so bacteria produce antibiotic-like compounds to inhibit the growth of competing bacteria. This might explain the high frequency of antagonistic interactions found among marine bacteria growing on labile nutrient media such as marine agar containing peptone and yeast extract (Long and

Azam 2001; Grossart et al. 2004; Rypien, Ward and Azam 2010). On the other hand, CMC-xylan is difficult to break down, requiring multiple lignocellulolytic enzymes. Therefore, bacteria dependent on CMC and xylan as carbon sources may spend more energy producing enzymes required to degrade the substrate and thus less energy toward the production of antibiotic-like compounds. I think that the chemical complexity of metabolic substrates is a factor in determining energy allocation among bacteria and thereby their interactions in the community. When bacteria must rely on recalcitrant substrates, they might channel more energy toward enzyme production and survival, and less toward the production of compounds that affect other bacteria.

My study indicates that labile carbon sources tend to cause competition/antagonism among bacteria, but complex carbon sources reduce

antagonism and might even trigger bacterial synergism. Bacterial synergism is commonly found among bacteria growing on lignocellulose where they degrade the refractory substrate cooperatively (Kato et al. 2005; Wongwilaiwalin et al.

2010; Jiménez, Korenblum and van Elsas 2014). Similarly, over half of the bacterial mixed cultures growing in lignocellulose medium showed synergistic growth but none in glucose medium (Deng and Wang 2016). These findings suggest that the complexity of substrate affects the way bacteria interact with less antagonism in recalcitrant substrates.

I also found that some of the pairwise antagonistic interactions were substrate-specific. While 76 antagonistic interactions occurred on both media, 22 interactions occurred only in CMC-xylan medium and 128 in glucose medium only (Figure 4.3). The substrate specificity might be due to the presence of different sugars in the two media. Sánchez et al. (2010) and Singh et al. (2014) reported that antibiotic production by bacteria is affected by the concentration and types of sugar available. Sugars such as carboxylated glucose and xylose are available when CMC and xylan are hydrolyzed ,whereas glucose alone is available in the other medium and at a much higher concentration. Such differences may explain the substrate specificity of the antagonistic interactions observed. In natural environments, changes in substrate composition over time may affect bacterial interactions in the community.

Antagonism was not uniformly distributed among bacteria but seems to be a pattern related to their phylogenic groups (Table 4.3). Gamma-Proteobacteria and Firmicutes (mostly Bacillus spp.) were the most antagonistic and they tended

to inhibit Actinobacteria and Bacteroidetes. Bacteroidetes was the most sensitive group. My results are consistent with those of previous studies (Long and Azam

2001; Grossart et al. 2004). Surprisingly, I found Actinobacteria among the most sensitive, with 10% to 50% being inhibited by other groups (Table 4.3). In an earlier study, Actinobacteria were found to be among the most antagonistic bacteria (Grossart et al. 2004) and regarded as good antibiotic producers especially Streptomycetes (Hopwood 2006). In the present study, they were antagonistic to only a few other bacteria (Table 4.3 and 4.4) probably due to a difference in bacterial species included in this study. It is also interesting to note that within-group antagonistic interactions were generally infrequent (Table 4.3 and 4.4). My results are consistent with previous studies that also showed that closely related bacteria are less likely to inhibit each other (Vetsigian, Jajoo and

Kishony 2011; Cordero et al. 2012). I also found that in general there is an inverse relationship between bacterial antagonism and susceptibility (Figure 4.4).

Highly antagonistic bacteria were more resistant to inhibition by others while susceptible bacteria were generally less antagonistic to others. This is in agreement with a previous study on interactions among Gamma-Proteobacteria showing that antagonistic bacteria are highly resistant to inhibitions by others, whereas susceptible bacteria are in general not inhibitory to others (Aguirre-von-

Wobeser et al. 2013).

4.6 Conclusion

In conclusion, my study provides evidence that the structural complexity of carbohydrate substrates mediates the frequency of antagonistic interactions

among bacteria. Complex carbohydrates reduce competitions among lignocellulolytic bacteria while labile carbon sources promote competition or antagonism. Bacteria did not interact randomly with others, but there seem to be to interaction patterns related to bacterial phylogenetic groups. Antagonistic interactions were not uniform among bacterial groups with Gamma-

Proteobacteria and Firmicutes the most antagonistic and Actinobacteria and

Bacteroidetes the most susceptible. As nutrients are utilized and substrate composition changes over time, interactions among bacteria within the community might change accordingly and thus affect the community structure and its functions in the environment.

REFERENCES

Adav SS, Ravindran A, Cheow ESH et al. Quantitative proteomic analysis of

secretome of microbial consortium during saw dust utilization. J Proteomics

2012;75:5590–603.

Aguirre-von-Wobeser E, Soberon-Chavez G, Eguiarte LE et al. Two-role model

of an interaction network of free-living gamma-proteobacteria from an

oligotrophic environment. Env Microbiol 2013;16:1366–77.

Azam F, Malfatti F. Microbial structuring of marine ecosystems. Nat Rev

Microbiol 2007;5:782–91.

Bates ST, Berg-Lyons D, Caporaso JG et al. Examining the global distribution of

dominant archaeal populations in soil. ISME J 2011;5:908–17.

Bell T, Newman JA, Silverman BW et al. The contribution of species richness

and composition to bacterial services. Nature 2005;436:1157–60.

Berlemont R, Martiny AC. Phylogenetic distribution of potential cellulases in

bacteria. Appl Environ Microbiol 2013;79:1545–54.

Brenner K, You L, Arnold FH. Engineering microbial consortia: a new frontier in

synthetic biology. Trends Biotechnol 2008;26:483–9.

Breugelmans P, Barken KB, Tolker-Nielsen T et al. Architecture and spatial

organization in a triple-species bacterial biofilm synergistically degrading the

phenylurea herbicide linuron. FEMS Microbiol Ecol 2008;64:271–82.

Buchan A, Neidle EL, Moran MA. Diversity of the ring-cleaving dioxygenase gene

pcaH in a salt marsh bacterial community. Appl Env Microbiol

2001;67:5801–9. 62

Buchan A, Newell SY, Butler M et al. Dynamics of bacterial and fungal

communities on decaying salt marsh grass. Appl Environ Microbiol

2003;69:6676–87.

Buchan A, Newell SY, Moreta JIL et al. Analysis of internal Transcribed spacer

(ITS) regions of rRNA genes in fungal communities in a southeastern U.S.

salt marsh. Microb Ecol 2002;43:329–40.

Burkholder PR, Pfister RM, Leitz FH. Production of a pyrrole antibiotic by a

marine bacterium. Appl Microbiol 1966;14:649–53.

Burmølle M, Webb JS, Rao D et al. Enhanced biofilm formation and increased

resistance to antimicrobial agents and bacterial invasion are caused by

synergistic interactions in multispecies biofilms. Appl Environ Microbiol

2006;72:3916–23.

Caffall KH, Mohnen D. The structure, function, and biosynthesis of plant cell wall

pectic polysaccharides. Carbohydr Res 2009;344:1879–900.

Carrero-Colón M, Nakatsu CH, Konopka A. Microbial community dynamics in

nutrient-pulsed chemostats. FEMS Microbiol Ecol 2006;57:1–8.

Cascales E, Buchanan SK, Duche D et al. Colicin Biology. Microbiol Mol Biol Rev

2007;71:158–229.

Christensen BB, Haagensen JAJ, Heydorn A et al. Metabolic Commensalism and

Competition in a Two-Species Microbial Consortium. Appl Environ Microbiol

2002;68:2495–502.

Ciric L, Philp JC, Whiteley AS. Hydrocarbon utilization within a diesel-degrading

bacterial consortium. FEMS Microbiol Lett 2010;303:116–22.

Cordero OX, Wildschutte H, Kirkup B et al. Ecological populations of bacteria act

as socially cohesive units of antibiotic production and resistance. Science

(80- ) 2012;337:1228–31.

Csárdi G, Nepusz T. The igraph software package for complex network research.

InterJournal Complex Syst 2006;1695:1–9.

D’Costa VM, Griffiths E, Wright GD. Expanding the soil antibiotic resistome:

exploring environmental diversity. Curr Opin Microbiol 2007;10:481–9.

Das M, Royer T V, Leff LG. Diversity of fungi, bacteria, and Actinomycetes on

leaves decomposing in a stream. Appl Environ Microbiol 2007;73:756–67.

Deng Y-J, Wang SY. Synergistic growth in bacteria depends on substrate

complexity. J Microbiol 2016;54:23–30.

Dumont MG, Murrell JC. Stable isotope probing - linking microbial identity to

function. Nat Rev Microbiol 2005;3:499–504.

Van Dyk JS, Pletschke BI. A review of lignocellulose bioconversion using

enzymatic hydrolysis and synergistic cooperation between enzymes -

Factors affecting enzymes, conversion and synergy. Biotechnol Adv

2012;30:1458–80.

Egland PG, Palmer RJ, Kolenbrander PE. Interspecies communication in

Streptococcus gordonii - Veillonella atypica biofilms: signaling in flow

conditions requires juxtaposition. Proc Natl Acad Sci U S A

2004;101:16917–22.

Elias S, Banin E. Multi-species biofilms: living with friendly neighbors. FEMS

Microbiol Rev 2012;36:990–1004.

Falkowski P. The global carbon cycle: a test of our Knowledge of Earth as a

system. Science (80- ) 2000;290:291–6.

Fenchel T, King GM, Blackburn TH. Degradation of organic polymers and

dydrocarbons. In: Fenchel T, King GM, Blackburn TH (eds.). Bacterial

Biogeochemistry. Third. Boston: Academic Press, 2012, 49–57.

Field CB, Behrenfeld MJ, Randerson JT et al. Primary production of the

biosphere: integrating terrestrial and oceanic Components. Science (80- )

1998;281:237–40.

Fierer N, Leff JW, Adams BJ et al. Cross-biome metagenomic analyses of soil

microbial communities and their functional attributes. Proc Natl Acad Sci U S

A 2012;109:21390–5.

Flint HJ, Duncan SH, Scott KP et al. Interactions and competition within the

microbial community of the human colon: links between diet and health. Env

Microbiol 2007;9:1101–11.

Foster KR, Bell T. Competition, not cooperation, dominates interactions among

culturable microbial species. Curr Biol 2012;22:1845–50.

Freilich S, Zarecki R, Eilam O et al. Competitive and cooperative metabolic

interactions in bacterial communities. Nat Commun 2011;2:589.

Gabrielson J, Hart M, Jarelov A et al. Evaluation of redox indicators and the use

of digital scanners and spectrophotometer for quantification of microbial

growth in microplates. J Microbiol Methods 2002;50:63–73.

Gihring TM, Humphrys M, Mills HJ et al. Identification of phytodetritus-degrading

microbial communities in sublittoral Gulf of Mexico sands. Limnol Oceanogr

2009;54:1073–83.

Lo Giudice A, Brilli M, Bruni V et al. Bacterium-bacterium inhibitory interactions

among psychrotrophic bacteria isolated from Antarctic seawater (Terra Nova

Bay, Ross Sea). FEMS Microbiol Ecol 2007;60:383–96.

Grossart H-P, Schlingloff A, Bernhard M et al. Antagonistic activity of bacteria

isolated from organic aggregates of the German Wadden Sea. FEMS

Microbiol Ecol 2004;47:387–96.

Guevara C, Zambrano MM. Sugarcane cellulose utilization by a defined microbial

consortium. FEMS Microbiol Lett 2006;255:52–8.

Haichar F el Z, Achouak W, Christen R et al. Identification of cellulolytic bacteria

in soil by stable isotope probing. Environ Microbiol 2007;9:625–34.

Hall RO, Meyer JL. The trophic significance of bacteria in a detritus-based

stream food web. Ecology 1998;79:1995–2012.

Hannen EJ Van, Mooij W, Agterveld MP Van et al. Detritus-dependent

development of the microbial community in an experimental system:

qualitative analysis by denaturing gradient gel electrophoresis. Appl Environ

Microbiol 1999;65:2478–84.

Haruta S, Cui Z, Huang Z et al. Construction of a stable microbial community with

high cellulose-degradation ability. Appl Microbiol Biotechnol 2002;59:529–

34.

Hasanuzzaman M, Mahmood H. Litter production and nutrient return through leaf

litter of selected cropland agroforest tree species in southwestern

bangladesh. Intergovernmental Panel on Climate Change (ed.). Agric For

2014;60:221–33.

Hatfield RD, Ralph J. Cell wall structural foundations: Molecular basis for

improving forage digestibilities. Crop Sci 1998;39:27–37.

Hibbing ME, Fuqua C, Parsek MR et al. Bacterial competition: surviving and

thriving in the microbial jungle. Nat Rev Microbiol 2010;8:15–25.

Hocking PJ. Seasonal dynamics of production, and nutrient accumulation and

cycling by Phragmites asutralis (Cav.) Trin. ex Stuedel in a nutrient-enriched

swamp in Inland Australia. I. Whole Plants. Mar Freshw Res 1989;40:421–

44.

Hoppe H. Significance of exoenzymatic activities in the ecology of brackish

water: measurements by means of methylumbelliferyl-substrates. Mar Ecol

Prog Ser 1983;11:299–308.

Hopwood DA. Soil to genomics: the Streptomyces chromosome. Annu Rev

Genet 2006;40:1–23.

Jaeger, C. H. I, Lindow SE, Miller W et al. Mapping of sugar and amino acid

availability in soil around roots with bacterial sensors of sucrose and

tryptophan. Appl Envir Microbiol 1999;65:2685–90.

Jiménez D, Korenblum E, van Elsas J. Novel multispecies microbial consortia

involved in lignocellulose and 5-hydroxymethylfurfural bioconversion. Appl

Microbiol Biotechnol 2014;98:2789–803.

Joshi F, Archana G, Desai A. Siderophore cross-utilization amongst rhizospheric

bacteria and the role of their differential affinities for Fe. Curr Microbiol

2006;53:141–7.

Kato S, Haruta S, Cui ZJ et al. Stable coexistence of five bacterial strains as a

cellulose-degrading community. Appl Environ Microbiol 2005;71:7099–106.

Kinkel LL, Schlatter DC, Xiao K et al. Sympatric inhibition and niche

differentiation suggest alternative coevolutionary trajectories among

Streptomycetes. ISME J 2014;8:249–56.

Kirchman DL. The ecology of Cytophaga–Flavobacteria in aquatic environments.

FEMS Microbiol Ecol 2002;39:91–100.

Kostylev M, Wilson D. Synergistic interactions in cellulose hydrolysis. Biofuels

2012;3:61–70.

Kritzberg ES, Langenheder S, Lindström ES. Influence of dissolved organic

matter source on lake bacterioplankton structure and function--implications

for seasonal dynamics of community composition. FEMS Microbiol Ecol

2006;56:406–17.

Kuehn KA, Lemke MJ, Suberkropp K et al. Microbial biomass and production

associated with decaying leaf litter of the emergent macrophyte Juncus

effusus. Limnol Oceanogr 2000;45:862–70.

Kumar K, Goh KM. Crop residues and Management practices: Effects on soil

quality, soil nitrogen dynamics, crop yield, and nitrogen recovery. Adv Agron

1999;68:197–319.

Lawlor DW. Photosynthesis, productivity and environment. J Exp Bot

1995;46:1449–61.

Lawrence D, Fiegna F, Behrends V et al. Species interactions alter evolutionary

responses to a novel environment. Ellner SP (ed.). PLoS Biol

2012;10:e1001330.

Lee CG, Watanabe T, Sato Y et al. Bacterial populations assimilating carbon

from 13C-labeled plant residue in soil: Analysis by a DNA-SIP approach. Soil

Biol Biochem 2011;43:814–22.

Lo Y-C, Saratale GD, Chen W-M et al. Isolation of cellulose-hydrolytic bacteria

and applications of the cellulolytic enzymes for cellulosic biohydrogen

production. Enzym Microb Technol 2009;44:417–25.

Long RA, Azam F. Antagonistic Interactions among Marine Pelagic Bacteria.

Appl Environ Microbiol 2001;67:4975–83.

Lowery CA, Dickerson TJ, Janda KD. Interspecies and interkingdom

communication mediated by bacterial quorum sensing. Chem Soc Rev

2008;37:1337–46.

Lydell C, Dowell L, Sikaroodi M et al. A population survey of members of the

phylum Bacteroidetes isolated from salt marsh sediments along the East

Coast of the United States. Microb Ecol 2004;48:263–73.

Lynd LR, Weimer PJ, van Zyl WH et al. Microbial Cellulose Utilization:

Fundamentals and Biotechnology. Microbiol Mol Biol Rev 2002;66:506–77.

Maki M, Leung KT, Qin W. The prospects of cellulase-producing bacteria for the

bioconversion of lignocellulosic biomass. Int J Biol Sci 2009;5:500–16.

Mangano S, Michaud L, Caruso C et al. Antagonistic interactions between

psychrotrophic cultivable bacteria isolated from Antarctic sponges: a

preliminary analysis. Res Microbiol 2009;160:27–37.

Marschner B, Kalbitz K. Controls of bioavailability and biodegradability of

dissolved organic matter in soils. Geoderma 2003;113:211–35.

Marsh PD, Boeden GHW. Microbial community interactions in biofilms. Symp

Gen Microbiol 2000:167–98.

Marx MC, Wood M, Jarvis SC. A microplate fluorimetric assay for the study of

enzyme diversity in soils. Soil Biol Biochem 2001;33:1633–40.

McCarthy AJ. Lignocellulose-degrading actinomycetes. FEMS Microbiol Lett

1987;46:145–63.

Miller MB, Bassler BL. Quorum sensing in bacteria. Annu Rev Microbiol

2001;55:165–99.

Mindl B, Anesio AM, Meirer K et al. Factors influencing bacterial dynamics along

a transect from supraglacial runoff to proglacial lakes of a high Arctic

glacieri. FEMS Microbiol Ecol 2007;59:307–17.

Molin S, Tolker-Nielsen T, Hansen SK. Microbial interactions in mixed-species

biofims. In: Ghannoum M and O’Toole GA (ed.). Microbial Biofilms.

Washington, D.C.: ASM Press, 2004, 206–21.

Moore JC, Berlow EL, Coleman DC et al. Detritus, trophic dynamics and

biodiversity. Ecol Lett 2004;7:584–600.

Newman MM, Liles MR, Feminella JW. Litter Breakdown and Microbial

Succession on Two Submerged Leaf Species in a Small Forested Stream.

PLoS One 2015;10:e0130801.

Nielsen AT, Tolker-Nielsen T, Barken KB et al. Role of commensal relationships

on the spatial structure of a surface-attached microbial consortium. Environ

Microbiol 2000;2:59–68.

Odum EP, Biever LJ. Resource quality, mutualism, and energy partitioning in

food chains. Am Nat 1984;124:360–76.

Orphan VJ, House CH, Hinrichs KU et al. Methane-consuming archaea revealed

by directly coupled isotopic and phylogenetic analysis. Science

2001;293:484–7.

Pereyra LP, Hiibel SR, Prieto Riquelme M V. et al. Detection and quantification of

functional genes of cellulose- degrading, fermentative, and sulfate-reducing

bacteria and methanogenic archaea. Appl Environ Microbiol 2010;76:2192–

202.

Perez J, Munoz-Dorado J, de la Rubia T et al. Biodegradation and biological

treatments of cellulose, hemicellulose and lignin: an overview. Int Microbiol

2002;5:53–63.

Perez-Gutierrez R-A, Lopez-Ramirez V, Islas A et al. Antagonism influences

assembly of a Bacillus guild in a local community and is depicted as a food-

chain network. ISME J Multidiscip J Microb Ecol 2013;7:487–97.

Prasad S, Manasa P, Buddhi S et al. Antagonistic interaction networks among

bacteria from a cold soil environment. FEMS Microbiol Ecol 2011;78:376–85.

Qing Q, Yang B, Wyman CE. Xylooligomers are strong inhibitors of cellulose

hydrolysis by enzymes. Bioresour Technol 2010;101:9624–30.

Ransom-Jones E, Jones D, McCarthy A et al. The Fibrobacteres: an important

phylum of cellulose-degrading bacteria. Microb Ecol 2012;63:267–81.

Rietl AJ, Overlander ME, Nyman AJ et al. Microbial community composition and

extracellular enzyme activities associated with Juncus roemerianus and

Spartina alterniflora vegetated sediments in Louisiana saltmarshes. Microb

Ecol 2016;71:290–303.

Ryan RP, Dow JM. Diffusible signals and interspecies communication in bacteria.

Microbiology 2008;154:1845–58.

Rypien KL, Ward JR, Azam F. Antagonistic interactions among coral-associated

bacteria. Environ Microbiol 2010;12:28–39.

Saitou N, Nei M. The neighbor-joining method: a new method for reconstructing

phylogenetic trees. Mol Biol Evol 1987;4:406–25.

Saiya-Cork KR, Sinsabaugh R., Zak D. The effects of long term nitrogen

deposition on extracellular enzyme activity in an Acer saccharum forest soil.

Soil Biol Biochem 2002;34:1309–15.

Sánchez S, Chávez A, Forero A et al. Carbon source regulation of antibiotic

production. J Antibiot (Tokyo) 2010;63:442–59.

Schink B. Energetics of syntrophic cooperation in methanogenic degradation.

Microbiol Mol Biol Rev 1997;61:262–80.

Schülein M. Protein engineering of cellulases. Biochim Biophys Acta - Protein

Struct Mol Enzymol 2000;1543:239–52.

Searcy DG. Nutritional syntrophies and consortia as models for the origin of

mitochondria. In: Seckbach J (ed.). Symbiosis: Mechanisms and Model

Systems. Dordrecht: Springer Netherlands, 2002, 163–83.

Sessitsch A, Hardoim P, Döring J et al. Functional characteristics of an

endophyte community colonizing rice roots as revealed by metagenomic

analysis. Mol Plant Microbe Interact 2012;25:28–36.

da Silva WJ, Seneviratne J, Parahitiyawa N et al. Improvement of XTT assay

performance for studies involving Candida albicans biofilms. Braz Dent J

2008;19:364–9.

Singh AK, Rautela R, Cameotra SS. Substrate dependent in vitro antifungal

activity of Bacillus sp strain AR2. Microb Cell Fact 2014;13:67.

Sinsabaugh RL, Findlay S, Franchini P et al. Enzymatic analysis of riverine

bacterioplankton production. Limnol Oceanogr 1997;42:29–38.

Sinsabaugh RL, Lauber CL, Weintraub MN et al. Stoichiometry of soil enzyme

activity at global scale. Ecol Lett 2008;11:1252–64.

Sinsabaugh RL, Osgood MP, Findlay S. Enzymatic models for estimating

decomposition rates of particulate detritus. J North Am Benthol Soc

1994;13:160–9.

Strom SL. Microbial ecology of ocean biogeochemistry: A community

perspective. Science (80- ) 2008;320:1043–5.

Sun Y, Cheng J. Hydrolysis of lignocellulosic materials for ethanol production: a

review. Bioresour Technol 2002;83:1–11.

Sutherland IW. The biofilm matrix--an immobilized but dynamic microbial

environment. Trends Microbiol 2001;9:222–7.

Tamura K, Peterson D, Peterson N et al. MEGA5: molecular evolutionary

genetics analysis using maximum likelihood, evolutionary distance, and

maximum parsimony methods. Mol Biol Evol 2011;28:2731–9.

Tomme P, Warren RAJ, Gilkes NR. Cellulose Hydrolysis by Bacteria and Fungi.

In: Poole RK (ed.). Advances in Microbial Physiology. Vol Volume 37.

Academic Press, 1995, 1–81.

Vanwonterghem I, Jensen PD, Ho DP et al. Linking microbial community

structure, interactions and function in anaerobic digesters using new

molecular techniques. Curr Opin Biotechnol 2014;27:55–64.

Vetsigian K, Jajoo R, Kishony R. Structure and evolution of Streptomyces

interaction networks in soil and in silico. PLoS Biol 2011;9:e1001184.

Visick KL, Fuqua C. Decoding microbial chatter: cell-cell communication in

bacteria. J Bacteriol 2005;187:5507–19.

Walker CB, He Z, Yang ZK et al. The electron transfer system of syntrophically

grown Desulfovibrio vulgaris. J Bacteriol 2009;191:5793–801.

Wang Q, Garrity GM, Tiedje JM et al. Naive bayesian classifier for rapid

assignment of rrna sequences into the new bacterial taxonomy. Appl Environ

Microbiol 2007;73:5261–7.

Waters CM, Bassler BL. Quorum Sensing: Cell-to-Cell Communication in

Bacteria. Annu Rev Cell Dev Biol 2005;21:319–46.

Weisburg WG, Barns SM, Pelletier DA et al. 16S ribosomal DNA amplification for

phylogenetic study. J Bacteriol 1991;173:697–703.

Wintermute EH, Silver PA. Dynamics in the mixed microbial concourse. Genes

Dev 2010a;24:2603–14.

Wintermute EH, Silver PA. Emergent cooperation in microbial metabolism. Mol

Syst Biol 2010b;6:1–7.

Wongwilaiwalin S, Rattanachomsri U, Laothanachareon T et al. Analysis of a

thermophilic lignocellulose degrading microbial consortium and multi-species

lignocellulolytic enzyme system. Enzyme Microb Technol 2010;47:283–90.

Yamashita S, Masuya H, Abe S et al. Relationship between the decomposition

process of coarse woody debris and fungal community structure as detected

by high-throughput sequencing in a deciduous broad-leaved forest in Japan.

Martins LO (ed.). PLoS One 2015;10:e0131510.

Zhang YH, Lynd LR. Toward an aggregated understanding of enzymatic

hydrolysis of cellulose: noncomplexed cellulase systems. Biotechnol Bioeng

2004;88:797–824.

Zhang YHP, Himmel ME, Mielenz JR. Outlook for cellulase improvement:

Screening and selection strategies. Biotechnol Adv 2006;24:452–81.

Zhang Y-HP. Next generation biorefineries will solve the food, biofuels, and

environmental trilemma in the energy-food-water nexus. Energy Sci Eng

2013;1:27–41.

Zhou K, Qiao K, Edgar S et al. Distributing a metabolic pathway among a

microbial consortium enhances production of natural products. Nat

Biotechnol 2015;33:377–83.

Zuroff TR, Curtis WR. Developing symbiotic consortia for lignocellulosic biofuel

production. Appl Microbiol Biotechnol 2012;93:1423–35.