R EVIEW: The rumen microbiome:

Transcription

1 The Professional Animal Scientist 30 (2014): American Registry of Professional Animal Scientists R EVIEW: The rumen microbiome: Composition, abundance, diversity, and new investigative tools Frédérique Chaucheyras-Durand * 1 and Faisury Ossa * Lallemand Animal Nutrition, 19 rue des Briquetiers, Blagnac cedex, France; INRA UR 454 Microbiology, CR Clermont-Ferrand/Theix, Saint-Genès Champanelle, France; and Lallemand Animal Nutrition, National Research Council Canada (NRCC), 6100 Avenue Royalmount, Montreal, QC, H4P 2R2, Canada ABSTRACT Ruminants are able to degrade and use fibrous feed as a source of energy and nutrients because of the presence of complex anaerobic microbiota in the rumen, composed mainly of bacteria, fungi, and ciliate protozoa. Ruminal microorganisms play different roles in feed digestion and act synergistically to ferment plant structural and nonstructural carbohydrates and proteins. This review reports the latest assessment of microbiota diversity in the rumen ecosystem and summarizes the molecular techniques and the newly available omic technologies, based on DNA and RNA sequence analysis, which allow for new insights into the structure and functions of these complex microbial communities. Key words: rumen, microbiota, ruminant digestion, diversity, DNA/ RNA-based technique 1 Corresponding author: fchaucheyrasdurand@lallemand.com INTRODUCTION Herbivores, omnivores, and carnivores have in common the presence of microbial life in their gastrointestinal tract. Gut microbial communities (Figure 1), composed of bacteria, ciliate and flagellate protozoa, anaerobic fungi, and viruses, play a vital role in nutritional, physiological, immunological, and protective functions of the host. The rumen is one of the most extensively studied gut ecosystems, because of the importance of ruminants for human nutrition and the major role played by rumen microbes in nutrition and health of the ruminant animal. Indeed, ruminants, in a symbiotic relationship with the microorganisms in the rumen, degrade and use fibrous feed as a source of energy and nutrients. Ruminal microorganisms play different roles in feed digestion and act synergistically to ferment plant structural and nonstructural carbohydrates and proteins. This review provides an overview of microbial abundance and diversity in the rumen. Moreover, this paper provides an outline of the current molecular approaches used to gain a better understanding of the structure and function of the microbial ecosystem. RUMEN MICROBIAL ABUNDANCE AND DIVERSITY Conventional culture-based techniques such as isolation, enumeration, and nutritional characterization have provided significant information on the diversity of the rumen microbiota. In fact, more than 200 species of bacteria and at least 100 species of protozoa and fungi inhabiting the rumen have been identified by culture-based techniques. Nevertheless, over the last 10 years the development of highthroughput sequencing techniques has allowed for a considerable increase in knowledge of the microbial diversity of the rumen ecosystem. Indeed, even if culture-based techniques are successful in isolating key representatives of rumen bacteria, archaea, and fungi, they are not well suited to characterizing the overall microbial diversity, because a vast majority of rumen species are not yet culturable. Recent studies indicate that when quantified

2 2 Chaucheyras-Durand and Ossa Figure 1. Rumen microbial consortium. B = bacteria; P = ciliate protozoa; S = fungal sporocyst; F = plant fiber. Scanning electron microscope photograph, B. Gaillard-Martinie, INRA UR454 Microbiology, France. Copyright INRA not for commercial purpose. by real-time PCR, some uncultured bacteria are as abundant as major cultured bacteria found in the rumen, suggesting that uncultured bacteria may play an important role in ruminal fermentation (Kim et al., 2011). The diversity within the rumen can be strongly affected not only by diet composition (Chaucheyras-Durand et al., 2012) but probably also by host genetics (Benson et al., 2010) and environmental factors (Uyeno et al., 2010). Rumen Bacteria and Archaea The diversity of bacteria and Archaea in the rumen of predominantly domesticated livestock has been studied through a metaanalysis of all curated 16S ribosomal RNA (rrna) gene sequences deposited in the Ribosomal Database Project database as of 2010, amounting to 13,478 bacterial and 3,516 archaeal sequences (Kim et al., 2011). From this analysis, the diversity of bacterial and archaeal species in the rumen is estimated to be approximately 7,000 and 1,500 species, respectively. The bacterial sequences were assigned to 5,271 operation taxonomic units, which represented 19 existing phyla, with Firmicutes (~56%), Bacteroidetes (~31%), and Proteobacteria (~4%) being the predominant. More than 90% of the Firmicutes sequences were related to genera within the class Clostridia. Within the Clostridia, Lachnospiraceae, Ruminococcaceae, and Veillonellaceae were the largest families. The predominant genera included Butyrivibrio, Acetivibrio, Ruminococcus, Succiniclasticum, Pseudobutyrivibrio, and Mogibacterium. Within the class Bacilli, Streptococci were dominant. In the Bacteroidetes phylum, the majority of sequences were assigned to class Bacteroidia, and Prevotella was the most predominant genus, the species Prevotella ruminicola, Prevotella brevis, Prevotella bryantii, and Prevotella albensis being among the most commonly found. All 5 classes of Proteobacteria were represented in the database of rumen sequences, with a predominance of Proteobacteria. It has been suggested that, as in humans, a shared core community exists between different cows because around 35 genera could be shared across 90 to 100% of the 16 cows studied in Jami and Mizrahi (2012a). The abundance of rumen bacteria has been reported to be up to individuals per gram of rumen content. Taxa abundance may greatly vary across individuals. Jami and Mizrahi (2012b) used quantitative PCR (qpcr) to quantify the abundance of 13 key functional species of the rumen in 16 cows. They found that some species, such as Eubacterium ruminantium, a biohydrogenating species with hemicellulolytic activities, were very stable in abundance across individuals, whereas others, such as Megasphaera elsdenii, a lacticacid-using species, seemed to exhibit a much greater degree of variation between animals, with values spreading over 3 orders of magnitude. In the study of Kim et al. (2011), almost all the archaeal sequences were assigned to the phylum Euryarchaeota. Based on the analysis of public databases, more than 90% of sequences of rumen archaea are affiliated with methaneproducing genera Methanobrevibacter (>60%), Methanomicrobium (~15%), and a group of uncultured rumen archaea commonly referred to as rumen cluster C (~16%). Within the genus Methanobrevibacter, 2 major groups M. gottschalkii (contains M. gottschalkii, M. thaueri, and M. millerae) and M. ruminantium (M. ruminantium and M. olleyae) appear dominant. In terms of abundance, depending on the target gene (rrs or mcra), qpcr data report values between 10 8 to gene copies per gram of rumen contents (Mosoni et al., 2011). Rumen Protozoa Microscopy has been the method of choice in identifying and enumerating protozoal populations in rumen samples and is still a reliable method. With this method, protozoa are generally enumerated at 10 5 to 10 6 cells per gram of rumen contents, and it has found Entodinium to be the dominant genus in the rumen, which can account for up to 95% of the total population in animals fed high-grain diets. Besides microscopy, protozoal diversity representing many different genera and species has been assessed through molecular techniques that are becoming more prevalent in ecological studies of ruminal protozoa (Skillman et al.., 2006; Tymensen et al., 2012). For example, using a DNA fingerprinting method with various protozoa-specific PCR primers, Sylvester et al. (2004) showed the effect of diets on protozoal diversity in the rumen and duodenum and identified as major protozoal species Epidinium caudatum, Entodinium caudatum, and Isotricha prostoma. Other protozoal genera include Dasytricha, Ostracodinium, Diplodinium, Diploplastron, Eudiplodinium, Epidinium, Ophryoscolex, and Polyplastron. Rumen Fungi Anaerobic fungi have been isolated from rumen contents and feces of numerous herbivores including ruminant and monogastric animals and have been assigned recently to a new and separate phylum, Neocallimastigomycota (Griffith et al., 2010; Liggenstoffer et al., 2010). Six genera, Neocallimastix, Piromyces, Anaeromyces, Caecomyces, Orpinomyces, and more recently Cyllamyces, have been

3 Molecular approaches for understanding the rumen microbial ecosystem 3 recognized, and up to now, 18 species of anaerobic rumen fungi have been described on the basis of their thallus morphology and their zoospore ultrastructure. However, through high-throughput sequencing technology, several new uncultured taxa were revealed. Indeed, the phylogenetic diversity of the gut anaerobic fungi was investigated in 30 different herbivore species using the internal transcribed spacer region 1 rrna region as the phylogenetic marker (Liggenstoffer et al., 2010). A total of 267,287 sequences representing all known anaerobic fungal genera were obtained in this study. Sequences affiliated with the genus Piromyces were the most abundant, representing 36% of the obtained sequences. Sequences affiliated with the genera Cyllamyces and Orpinomyces were the least abundant, representing 0.7 and 1.1% of the total sequences obtained, respectively. Furthermore, 38.3% of the sequences obtained did not cluster with previously identified genera and formed 8 phylogenetically distinct novel anaerobic fungal lineages. The abundance of the rumen fungal community has been estimated to be around 10% of the total microbial biomass, with large variations according to diet and individual (Krause et al., 2013). Rumen Bacteriophages Bacteriophages are abundant ( particles per gram) in the rumen ecosystem, but the diversity of these viruses as well as their ecological role are poorly understood. The first comprehensive metagenomic analysis of the bovine rumen virome was reported recently, in which 28,000 different viral genotypes were identified (Berg-Miller et al., 2012). The majority (~78%) of sequences did not match any previously described viruses. Prophages outnumbered lytic phages by approximately 2:1, with the most abundant bacteriophage and prophage types being associated with members of the dominant rumen phyla (Firmicutes, Bacteroidetes, and Proteobacteria). ESTABLISHMENT OF THE RUMEN MICROBIOTA AT BIRTH Our current knowledge of the changes in rumen ecology from birth to adulthood is based primarily on cultivation studies and thus provides only a general understanding of the microbial successions that occur with age. Over the past 20 years, the most informative studies have been performed in France by INRA research teams who raised newborn lambs in sterile isolators under different conditions to describe the sequence of rumen colonization (Fonty et al., 1983, 1987; Chaucheyras-Durand et al., 2010). More recently, a pyrosequencing approach was used to assess bacterial composition in ruminants of ages 1 d to 2 yr (Jami et al., 2013). Results confirmed previous studies and showed that the rumen microbial communities are established very soon after birth, in a very precise sequence, as reflected by a decrease in aerobic and facultative anaerobic taxa and an increase in anaerobic ones (Figure 2). Some rumen bacteria that are essential for mature rumen function (i.e., cellulolytic bacteria) can be detected from the first day after birth, well before arrival of solid plant material. The diversity and within-group similarity increases with age, suggesting a more diverse but homogeneous and specific mature community, as compared with the more heterogeneous and less diverse primary community. The bacterial community may be transmitted from the dam via birth canal, teat surface, skin, or saliva and thereby depends greatly on maternal interactions. Using animal models, i.e., lambs harboring more or less complex microbial communities in the rumen, the importance of maintaining a complex and diversified microbiota to ensure optimal development of the rumen wall and papillae, efficient digestive function (as assessed by feed intake, concentrations of VFA and ammonia, digestibility of the feed), as well as to sustain animal growth Figure 2. Phylum-level composition of the rumen microbiota from birth to 2 yr of age in ruminants, using pyrosequencing (adapted by permission from Macmillan Publishers Ltd.: ISME Journal, Jami et al., 2013, copyright 2013). has been demonstrated (Hobson and Fonty, 1997). USE OF MOLECULAR TECHNIQUES AND OMIC APPROACHES TO UNDERSTAND RUMEN MICROBIAL FUNCTION Analysis of the 16S/18S rrna Gene Sequence During the last 15 years, the advent of routine gene sequencing technologies and the availability of large public databases for comparative analysis have allowed for rapid identification of new bacterial isolates on the basis of their 16S/18S rrna (rrs) gene sequences. Because of evolutionary conservation of 16S/18S rrna and their encoding genes (rdna), techniques based on 16S/18S rrna/rdna can be used to enumerate targeted microbes within a complex ecosystem, in addition to their use in molecular characterization and for establishing a classification scheme, which predicts natural evolutionary relationships without the need to cultivate the organisms (Zoetendal et al., 2004; Deng

4 4 Chaucheyras-Durand and Ossa et al., 2008; Figure 2). Construction of 16S rrna gene clone libraries of rumen bacteria and sequencing of these clones has revealed a vast diversity of bacterial genera and species that have never been characterized, mainly because there were no cultured representatives (Tajima et al., 2001; Larue et al., 2005; Ozutsumi et al., 2005). It is estimated that less than 20% of the rumen microbiome can be cultured using standard techniques (Krause et al., 2013). Sequencing of rrs in rumen or cecum samples of ruminants identified Prevotella as an important genus in the community (Brulc et al., 2009; Pitta et al., 2010). These studies also showed that the cultivated cellulolytic bacterial genera Ruminococcus and Fibrobacter were not among the most abundant members of the community and identified the presence of various other fiber-degrading genera. Whitford et al. (2001) detected 41 sequences of 16 rrna genes from the bovine rumen and stated that M. ruminantium and M. stadtmanae were the predominant species in the rumen; a second study found that Methanobrevibacter phylotypes were most frequently detected in the ovine and bovine rumen (Wright et al., 2004, 2007). In addition, previously uncultured species belonging to the Methanobacteriaceae and Methanosarcinaceae families were also identified in the bovine rumen (Whitford et al., 2001). Furthermore, in-depth rrs sequencing shows that the community structure of rumen bacteria is affected by changes in diet composition, allowing researchers to identify the prevalent core members of the community but also rare community members that could be associated with feeding practices (Callaway et al., 2010; Pitta et al., 2010; Shanks et al., 2011). The community structure in the lower gastrointestinal tract is, as expected, different from that of the rumen (Callaway et al., 2010), but is also greatly influenced by diet (Callaway et al., 2010; Durso et al., 2010; Shanks et al., 2011). The complexity of the feed seems to favor diversity. Rumen communities associated with bermudagrass diets, rich in structural carbohydrates and secondary compounds, were more diverse than were those associated with growing wheat forage (Pitta et al., 2010), and the presence of highly degradable carbohydrates in the diet such as starch decreased bacterial diversity in feces (Shanks et al., 2011). Quantification of the 16S/18S rrna Gene Absolute or relative abundance of a given organism has been assessed by fluorescence in situ hybridization, blot hybridization, or qpcr targeting the 16S/18S rrna gene. For instance, blot hybridization allowed researchers to determine the proportions of the main cellulolytic species Fibrobacter succinogenes, Ruminococcus flavefaciens, and Ruminococcus albus among the total bacterial community. They appeared to represent less than 10% of the total bacterial community (Michalet-Doreau et al., 2002; Stevenson and Weimer, 2007). However, recent data involving fluorescence in situ hybridization detection suggest that these bacteria account for about 50% of the total active cellulolytic bacteria (Kong et al., 2012). Recently, rrna probes were designed and applied for evaluating the effect of heat stress on the rumen microbial composition of Holstein heifers (Uyeno et al., 2010). This study showed that exposure to heat and humidity does affect the population levels of specific bacterial groups in the ruminal microbial community, notably Streptococcus and Fibrobacter genera, which increased and decreased, respectively. Studies using fluorescence in situ hybridization based enumeration have established the predominant specific microorganisms in rumen samples. For instance, Yanagita et al. (2000) found that Methanomicrobium mobile represented 54% of the total methanogens in rumen sheep, and Mackie et al. (2003) detected Oscillospira spp. in different ruminants. Several approaches to qpcr have been developed to monitor microbial species in a very sensitive manner. Using competitive PCR, the 16S rrna genes of a variety of bacterial species were quantified in rumen samples (Koike and Kobayashi, 2001). Competitive PCR studies identified bacterial populations from animals fed various diets at various times per day. These studies indicated F. succinogenes as the predominant cellulolytic bacteria (Kobayashi et al., 2008). Real-time PCR targeting the 16S rrna gene also overcame the hybridization disadvantages and has been used successfully on nucleic acids extracted from rumen contents to monitor microbial populations in the rumen. Tajima et al. (2001) monitored changes in 12 bacterial species in the rumen during diet transition. Ouwerkerk et al. (2002) used real-time qpcr to enumerate the population of Megasphaera elsdenii, an important lactate-using ruminal bacterium. This method was used as a tool for tracking probiotically introduced M. elsdenii in the rumen. Klieve et al. (2003) applied qpcr to determine the populations of M. elsdenii YE34 and B. fibrisolvens YE44 in the rumen of cattle fed a high-grain diet. This study showed that M. elsdenii supplementation establishes a lactic acid using bacterial population in the rumen of grain-fed cattle 7 to 10 d earlier than in uninoculated cattle. This approach, and the publication of many primer pairs for exploring other rumen bacterial populations (Anaerovibrio lipolytica, Butyrivibrio fibrisolvens, Eubacterium ruminantium, Prevotella albensis, P. brevis, P. bryantii, P. ruminicola, Ruminobacter amylophilus, genus Prevotella) using different qpcr assays, can provide useful information for evaluating the effect of dietary treatments on the rumen microbiome, such as changes associated with acidogenic diets in dairy cows (Khafipour et al., 2009). DNA Fingerprinting Techniques To study population structure and dynamics, genetic fingerprinting techniques have been developed. These PCR-based techniques involve extrac-

5 Molecular approaches for understanding the rumen microbial ecosystem 5 tion of nucleic acids, amplification of rrna/rdna, and analysis of PCR products by fingerprinting techniques. The more commonly used techniques to monitor community shifts and compare differences and similarities in the community of different rumen microbial groups are restriction fragment length polymorphisms, denaturing gradient gel electrophoresis (DGGE), temperature gradient gel electrophoresis, temporal temperature gradient gel electrophoresis, and single-strand conformation polymorphism. Using 16S rdna restriction fragment length polymorphisms, Jarvis et al. (2001) suggested that S. bovis diversity may depend on the diet. Studies using 16S rdna DGGE/temperature gradient gel electrophoresis found that the corn-fed animals displayed more diverse and abundant bacterial populations than did the hay-fed animals (Kocherginskaya et al., 2001). Mackie et al. (2003) used DGGE to monitor the uncultured bacterium Oscillospira spp. in different ruminants. The results suggested that the presence of Oscillospira species in various rumen ecosystems depends on diet and that the highest counts are associated with the feeding of fresh-forage diets to cattle and sheep. Sylvester et al. (2004) and Regensbogenova et al. (2004) demonstrated the utility of DGGE in profiling protozoal communities in the rumen and duodenum by employing different protozoa-specific primers. They were able to show the effect of diets on protozoal diversity and identified the major protozoal species (Epidinium caudatum, Entodinium caudatum, and Isotricha prostoma) by sequencing rdna fragments recovered from predominant DGGE bands (Sylvester et al., 2004). Using the DGGE targeting mcra gene, Mosoni et al. (2011) studied the effect of defaunation on methanogen diversity and on their abundance with qpcr. Interestingly, neither the diversity nor the abundance of the dominant methanogen population was affected by defaunation, although methane production was significantly reduced. The results revealed that further molecular investigation is needed to understand the microbial mechanisms leading to a reduction in methane emission by ruminants (Mosoni et al., 2011). It is clear that the combined application of molecular techniques will enable researchers to gain access to a complete description of the genetic diversity of the gastrointestinal tract (Figure 3). In fact, fundamental biological processes can now be studied by applying the full range of omic technologies [genomics, transcriptomics, proteomics, metabolomics, and beyond (Figure 4)]. Indeed, information on the microbial rumen ecosystem can be gathered through transcriptome (whole set of genes that are converted into mrna molecules), proteome (all proteins found in a given cell), metabolome (all metabolism products and intermediates in a cell), interactome (biologically active metabolism products able to interact with a given protein), and phenome (all observable characteristics of an organism) investigations. The combined use of high-throughput DNA sequencing and omics disciplines offers 2 things. It offers the potential to obtain a complete understanding of the lifestyle of a specific microbe and to assess its genetic potential in a comparative and functional manner. Whole Genome Sequencing De novo assemblies of microbial genomes have been accomplished with new high-throughput sequencing technologies, which are referred to as next-generation (or second-generation) sequencing, including systems such as Roche/454, Illumina/Solexa, Applied Biosystems/SOliD, and Helicos BioSciences. These methods enable rapid characterization of targeted sequences and cost much less than traditional, first-generation, Sanger sequencing techniques. The combination of nucleic-acid sequencing technologies and annotation platforms has made the genome sequencing of individual microorganisms both more affordable and more available to the research community. This has led to a marked increase in the number of rumen microorganisms whose genomes have been sequenced. There are currently 20 publicly available genome sequences from rumen microorganisms (Table 1; McSweeney and Mackie, 2012). Genome sequences of strains belonging to Fibrobacter succinogenes, Ruminococcus albus, Ruminococcus flavefaciens, Prevotella ruminicola, and Prevotella bryantii are now available, and from these genome sequences, 183 putative carbohydrate active enzymes have been found for F. succinogenes, and more than 140 for both Ruminococcus species (Flint et al., 2012). The power of genome sequencing is, therefore, apparent from this case; several fold more enzymes were identified in the F. succinogenes genome than were known from all previous studies of this organism. Next-generation sequencing also has the potential to provide new insights into the entire genome of the rumen environment, including genes of the methanogenic community that are present at very low levels. Six genome sequences belonging to 4 species are available for the family Methanobacteriaceae. Among them, M. ruminantium is usually found to be the predominant species in the rumen, whereas M. smithii and M. stadtmanae are common but less abundant. Methanobrevibacter ruminantium is the first methanogen from the rumen to have a completely assembled genome sequence (Leahy et al., 2010), and sequencing of the genomes of other rumen methanogens is under way. Methanobrevibacter ruminantium appears to contain a greater diversity of genes encoding surface adhesion-like proteins than either M. smithii or M. stadtmanae. The surface adhesion-like proteins of M. ruminantium contain a cell anchoring domain. The discovery of these proteins has provided candidate targets on the cell surface for vaccine development that may inhibit rumen methanogens, and thus could mitigate ruminant methane emissions (Leahy et al., 2010). There is little information available on the genomic make-up of rumen anaerobic fungi and ciliate protozoa, and

6 6 Chaucheyras-Durand and Ossa Figure 3. Molecular methods used alone or in combination to examine rumen microbial community structure [adapted from Zoetendal et al., 2004, J Nutr (2004;134: )]. r = ribosomal; RT-PCR = reverse-transcription PCR; DGGE = denaturing gradient gel electrophoresis; TGGE = temperature gradient gel electrophoresis; TTGE = temporal temperature gradient gel electrophoresis. to date only one publication reports the analysis of the lignocellulolytic machinery in the genome of the Orpinomyces sp. strain, a rumen fungal strain. The study revealed an extremely rich repertoire with evidence of horizontal gene acquisition from multiple bacterial lineages (Youssef Figure 4. Omics technologies involving the genome, transcriptome, proteome, and metabolome (adapted from McSweeney and Mackie, 2012, Source: Food and Agriculture Organization of the United Nations, 2012, McSweeney and Mackie, Microorganisms and ruminant digestion: State of knowledge, trends and future prospects. Reproduced with permission.) Seq = sequencing. et al., 2013), supporting the essential role that these microorganisms play in plant biomass degradation. Recently, a Rumen Microbial Genomics Network (RMG.Network@ agresearch.co.nz) has been established that coordinates and accelerates the sequencing and development of rumen microbial genomics approaches for easier access to methods, genome sequences, and metagenome data relevant to the rumen microbial community. In 2011, the Rumen Microbiology group (AgResearch, New Zealand) obtained support from the US Department of Energy Joint Genome Institute through their Community Sequencing Programme for a major sequencing project titled The Hungate 1000: A catalogue of reference genomes from the rumen microbiome. The Hungate 1000 project aims to produce a reference set of 1000 rumen microbial genome sequences from cultivated rumen bacteria and methanogenic Archaea, together with representative cultures of rumen anaerobic fungi and ciliate protozoa that represent the broad diversity of organisms in the rumen and which are of functional significance (McSweeney and Mackie, 2012). These sequencing projects have different purposes such as discovering new genes encoding fiber-degrading activities for enhancing animal production, for feedstock depolymerization, and for biofuel production. In addition, genome information will be used to support international efforts to develop methane mitigation and rumen adaptation technologies, as well as to initiate research aimed at understanding rumen function to find a balance between food production and greenhouse gas emissions. These projects are also adding functional omics technologies to gain a better understanding of gene function in these organisms. Metagenomics In fact, the sequence information, which can be linked to defined species of known function, is an essential prerequisite for interpretation of

7 Molecular approaches for understanding the rumen microbial ecosystem 7 Table 1. Publically available genome sequences of rumen bacteria and archaea (adapted from McSweeney and Mackie, 2012; Morgavi et al., 2013) 1 Family Organism Reference Fibrolytic bacteria Lachnospiraceae Butyribivrio proteoclasticus B316 Kelly et al. (2010) Eubacteriaceae Eubacterium cellulosolvens 6 Prevotellaceae Prevotella bryantii B 1 4 Purushe et al. (2010) Prevotella ruminicola 23 Purushe et al. (2010) Fibrobacteraceae Fibrobacter succinogenes S85 Suen et al. (2011b) Ruminococcaceae Ruminococcus albus 7 Suen et al. (2011a) Ruminococcus albus 8 Ruminococcus flavefaciens FD-1 Berg Miller et al. (2009) Ruminococcus flavefaciens 007C Methanogens Methanobacteriaceae Methanobrevibacter ruminantium M1 Leahy et al. (2010) Other bacteria Pasteurellaceae Actinobacillus succinogenes 130Z Basfia succiniciproducens MBEL55E Mannheimia succiniciproducens Hong et al. (2004) Peptococcaceae Desulfotomaculum ruminis DSM 2154 Spring et al. (2012) Desulfovibrionaceae Desulfovibrio desulfuricans ssp. desulfuricans str. ATCC Coriobacteriaceae Slackia heliotrinireducens DSM20476 Pukall et al. (2009) Spirochaetaceae Treponema saccharophilum DSM 2985 Helicobacteraceae Wolinella succinogenes DSM 1740 Baar et al. (2003) Veillonellaceae Megasphaera elsdenii DSM Marx et al. (2011) Lactobacillaceae Lactobacillus ruminis RF3 1 Source: Food and Agriculture Organization of the United Nations, 2012, McSweeney and Mackie, Micro-organisms and ruminant digestion: State of knowledge, trends and future prospects. Reproduced with permission. metagenomic data sets. Metagenomics is the culture-independent genomic analysis of microbial communities and comprises the functional and sequence-based analysis of the collective microbial genomes contained in an environmental sample. Metagenomics provides the potential to capture and study the entire microbiome (the predominant genomes). This approach creates DNA clone libraries and ultimately generates a catalog of genomic information that can be used to gain insights into the predominant genes, determining how they are regulated, and their contribution to ecosystem function (Morgavi et al., 2013), which may provide information for environmental, health, and productivity benefits. Through sequence-based metagenomic studies of the rumen microbiome, biomass-degrading genes and genomes have been characterized. In a study by Hess et al. (2011), 268 gigabases of metagenomic DNA from microbes adherent to plant fiber (switchgrass) incubated in cow rumen were sequenced and analyzed. Putative carbohydrate-active genes (27,755) were identified, and 90 candidate proteins were expressed, of which 57% were active against cellulosic substrates. In addition, 15 uncultured microbial genomes were reconstructed. An activity-based metagenomic study of a bovine ruminal protozoal enriched cdna expression library identified 4 novel genes possibly involved in cellulose and xylan degradation (Findley et al., 2011). These data sets provide a substantially expanded catalog of genes and genomes participating in the degradation of cellulosic biomass. Metagenomic studies have emphasized differences of rumen microbiota between animals consuming the same diet. For instance, Brulc et al. (2009) studied the rumen metagenome of 3 steers and found that the ruminal fiber-adherent microbial populations of one steer had a microbiome and a metagenome that were remarkably different from the other 2 steers. Functional metagenomics and other high-throughput technologies have been used to screen enzymes involved in lignocellulose bond degradation of fibrous ruminant feed (Beloqui et al., 2006) and to identify hydrolytic enzymes of biotechnological interest using specific substrates, in particular enzymes involved in the deconstruction of structural plant polysaccharides (Morgavi et al., 2013). The first study in this direction by Ferrer et al. (2005) identified 9 endoglucanases, 12 esterases, and 1 cyclodextrinase from a dairy cow rumen metagenomic library. The same group (Ferrer et al., 2012) applied a functional metagenomic approach to total DNA isolated from a fiber-adherent ruminal microbial community of a cow and identified a new and original putative glycoside hydrolase, exhibiting a multifunctional phenotype. Through functional metagenomics, many new

8 8 Chaucheyras-Durand and Ossa enzymes have been discovered (Ferrer et al., 2005, 2007), and a bacterial artificial chromosome library has been constructed from the rumen of a dairy cow. This bacterial artificial chromosome is being used to screen for novel enzyme activities (Zhu et al., 2007). Rumen metagenome libraries have also been used to screen for other bioactivities, including novel lipases (Liu et al., 2009; Bayer et al., 2010), polyphenol oxidase (Beloqui et al., 2006), and an enzyme capable of degrading 3,5,6-trichloro-2-pyridinol, a degradation product of the organophosphorus insecticide chlorpyrifos (Math et al., 2010). Functional metagenomics are moving to target functional analysis of the microbes in the rumen ecosystem and to measure gene expression of the whole rumen microbiome. By measuring the expression of genes, the activity of functionally different microbial groups has been described. Béra-Maillet et al. (2009) identified gene expression of glycoside hydrolase genes, using reverse-transcription qpcr in the rumen contents of a conventional sheep and of a gnotobiotic lamb (harboring a microflora containing F. succinogenes as sole cellulolytic microorganism). Similar gene expression approaches under in vitro or in vivo conditions have been conducted to determine the effect of diet on gene expression and to understand better the nutritional consequences for the animal (Krause et al., 2005; Guo et al., 2008). Functional gene analysis and metagenomic catalog information are used in association with bioinformatics to transform DNA-sequence data via translation to amino-acid sequences. For example, the DNA sequence data from a rumen sample can be binned (grouped according to similarity and frequency of occurrence) and annotated using various gene identification and database tools (e.g., ContigExpress, Phylopythia, MEGAN, and MG-Rast) to classify the genes within known enzyme families or clusters of orthologous genes and metabolic pathways. By using this process, a catalog of the genetic capacity of the rumen ecosystem can be made and the path toward identifying novel genes or genes of unknown function from the environment (Brulc et al., 2009; Hess et al., 2011; Berg Miller et al., 2012) can be determined. Transcriptomics Transcriptomic (gene expression) measurements, often referred to as gene expression microarrays or gene chips, constitute another gene-based technology. Transcriptomics allow an estimation of the expression of genes under the prevailing conditions, which in the end determines whether an activity or phenotype is actually present in the rumen. This path has been widely used in animal genomics and has enabled researchers to monitor, on a broad scale, the effects of pathogens on host cells and tissues, aiming to gain insight into the molecular mechanisms that are involved in host pathogen interactions. For example, numerous gene expression studies on mastitis have been performed. Recently, a metaanalysis study was performed based on 6 independent studies of infections with mammary gland pathogens, including samples from different animal species challenged with different pathogens (S. aureus, E. coli, and S. uberis), or infected macrophages or dendritic cells. Results showed that the cattle-specific response was characterized by alteration of the immune response and by modification of lipid metabolism (Genini et al., 2011). Microarray analyses also have been used to determine rumen epithelial responses to high-grain diets, which lead to rumen acidosis. Results of this study offer molecular targets that may be useful in the treatment and prevention of ruminal acidosis (Steele et al., 2011). An emerging technology is metatranscriptomics, which is used to assess what genes are globally expressed in a microbial ecosystem by analysis of all the messenger RNA in a sample, and this has been applied recently in the rumen of muskoxen (Ovibos moschatus). This work was focused on plant cell wall degrading enzymes and found a much higher proportion of cellulase enzymes than previously identified, providing several candidate genes coding for potentially valuable lignocellulolytic enzymes (Qi et al., 2011). Proteomics and Metabolomics Proteomics (protein expression, the large-scale analysis of proteins) is another interesting approach to characterize gene function, and by building functional linkages between proteins, it can help to understand biological and regulatory cascades. Dunne et al. (2012) demonstrated that Butyrivibrio proteoclasticus modulates the secretion of hemicellulose-degrading enzymes, which supported the notion that this organism makes an important contribution to polysaccharide degradation in the rumen. It is now possible to study the simultaneous expression of more than 1,000 proteins using mass spectrometry coupled with various separation methods. Furthermore, through whole metabolic profile monitoring in living biological systems, metabolomics (production of metabolites) may be particularly useful in unraveling the biochemical consequences of disease and describing altered physiological states induced by genetic mutations, drugs, diet, or stress. For example, an NMR-based approach was recently applied to assess system-wide metabolic responses as expressed in the urine and serum of a large cohort of sheep subjected to road transport. The profiling of 1 H-NMR spectra revealed that the transported animals experienced altered gut and energy metabolism, muscle catabolism, and possibly a renal response (Li et al., 2011). In a study by Klein et al. (2012), the NMR metabolomic analysis of milk produced by dairy cows revealed that the milk glycerophosphocholine-tophosphocholine ratio could be identified as a prognostic biomarker for risk of ketosis. It is clear that these non-nucleicacids based approaches will complement genomic approaches. Genomics decodes the sequence information of

9 Molecular approaches for understanding the rumen microbial ecosystem 9 an organism and provides a parts catalogue, whereas proteomics and metabolomics attempt to elucidate the functions and relationships of the individual components and predict the outcomes of the modules they form at a higher level. For example, with metabolomics it should be possible to identify signaling molecules, excreted in the gastrointestinal tract, used for communication between host cells and different microbes, and the effect of these signaling molecules on the bacterial or host physiology can then be studied using genomic and proteomic approaches (Zoetendal et al., 2004). It is, however, important to point out the possible methodological drawbacks of these approaches. The handling and treatment of biological samples is critical when characterizing the composition of the gut microbiota for different ecological conditions. Up to now there is no general consensus on temperature or duration of storage of digestive contents dedicated to DNA/RNA-based studies. Storage at room temperature, at 20 C, at 80 C, under liquid nitrogen, or freeze-drying are being used for preservation of microbial DNA. For instance, pyrosequencing of the 16S rrna gene was used by Carroll et al. (2012) to characterize the microbiota in human fecal samples stored either at room temperature or 80 C and showed that fecal samples exposed to room or deep freezing temperatures for up to 24 h and 6 mo, respectively, exhibited a microbial composition and diversity that shared more identity with its host of origin than any other sample. However, DNA begins to fragment when exposed to room temperature for more than 24 h (Cardona et al., 2012). The integrity of total RNA is a critical parameter for metatranscriptomic analyses. Degradation of RNA also compromises results of downstream applications, such as reverse-transcription qpcr or microarray studies. Cardona et al. (2012) compared different storage conditions of human stool samples. Their results suggested that the best storing condition to extract high-quality RNA for metatranscriptomic analyses is to keep the stool samples at room or low temperature for no more than a few hours (<24 h) after collection. Alternatively, samples can be kept at 20 C for longer periods as long as defrosting is prevented until the extraction of RNA starts. The use of preservative agents (RNase inhibitor) can be necessary to avoid degradation over time, but the efficiency of such treatment may depend on the nature of the sample and, for example, variability in RNA quality observed in samples mixed with an RNase inhibitor solution could reflect an insufficient homogenization of hard stools (Cardona et al., 2012). Most of these culture-independent techniques rely on DNA/RNA extraction, and the risk of extraction bias exists. Indeed, it is important to be able to extract a DNA pool that will be representative of the whole community, and given that the rumen harbors prokaryotic and eukaryotic organisms, grampositive and gram-negative bacteria, the cell lysis step has to be efficient enough to break down resistant cell walls (such as those of gram-positive bacteria) without completely destroying the content of cells with a more fragile membrane (i.e., gram-negative bacteria). Moreover, the rumen content is a heterogeneous environment where planktonic microorganisms can be easily accessible in the liquid phase but particle-associated microorganisms can be embedded in solid digesta. When PCR steps are used, primer specificity is a key factor that has to be taken into account because a lack of specificity will lead to a bias in detection or in quantification of the target organism or gene. For example, it may be difficult to design primers that will be specific to all members of a very diverse group, such as Firmicutes. The PCR conditions are also of great importance, and the presence of chemical inhibitors, inaccurate primer/target hybridization, or annealing temperatures may induce false amplification. With the sequencing approaches, issues concerning affiliation of sequences can arise and can lead to misinterpretation of the data. Indeed, different sequence affiliations may be obtained according to various pipelines relating to sequence databases of different composition and quality. Moreover, most sequencing technologies provide information at the phylum, order, or family level, some at the genus level, and it is not yet possible to have a description at the species level. As a wide functional diversity exists even at the genus level, the biological interpretation of the sequencing data can be limited. Storage of metadata generated by bioinformatic processes can also be a big issue and a limiting factor for researchers. CONCLUSION In the context of the huge complexity of the rumen ecosystem, new nucleic acid based technologies that are culture independent can be employed to examine rumen microbial ecology and diversity to determine not only which species are present but also what role they play in the ecosystem. These technologies can potentially contribute to a higher level of knowledge of rumen function and will allow for a better understanding of the roles of microorganisms in achieving high productivity, decreasing environmental pollutants, and limiting contamination of the food chain. LITERATURE CITED Baar, C., M. Eppinger, G. Raddatz, J. Simon, C. Lanz, O. Klimmek, R. Nandakumar, R. Gross, A. Rosinus, H. Keller, P. Jagtap, B. Linke, F. Meyer, H. Lederer, and S. C. Schuster Complete genome sequence and analysis of Wolinella succinogenes. Proc. Natl. Acad. Sci. USA 100: Bayer, S., A. Kunert, M. Ballschmiter, and T. Greiner-Stoeffele Indication for a new lipolytic enzyme family: Isolation and characterization of two esterases from a metagenomic library. J. Mol. Microbiol. Biotechnol. 18: org/ / Beloqui, A., M. Pita, J. Polaina, A. Martínez- Arias, O. V. Golyshina, M. Zumárraga, M. M. Yakimov, H. García-Arellano, M. Alcalde, V. M. Fernández, K. Elborough, J. M. Andreu, A. Ballesteros, F. J. Plou, K. N. Timmis, M. Ferrer, and P. N. Golyshin Novel polyphenol oxidase mined from

10 10 Chaucheyras-Durand and Ossa a metagenome expression library of bovine rumen: Biochemical properties, structural analysis, and phylogenetic relationships. J. Biol. Chem. 281: Benson, A. K., S. A. Kelly, R. Legge, F. Ma, S. J. Low, J. Kim, M. Zhang, P. L. Oh, D. Nehrenberg, K. Hua, S. D. Kachman, E. N. Moriyama, J. Walter, D. A. Peterson, and D. Pomp Individuality in gut microbiota composition is a complex polygenic trait shaped by multiple environmental and host genetic factors. Proc. Natl. Acad. Sci. USA 107: pnas Béra-Maillet, C., P. Mosoni, A. Kwasiborski, F. Suau, Y. Ribot, and E. Forano Development of a RT-qPCR method for the quantification of Fibrobacter succinogenes S85 glycoside hydrolase transcripts in the rumen content of gnotobiotic and conventional sheep. J. Microbiol. Methods 77: mimet Berg Miller, M. E., D. A. Antonopoulos, M. T. Rincon, M. Band, A. Bari, T. Akraiko, A. Hernandez, J. Thimmapuram, B. Henrissat, P. M. Coutinho, I. Borovok, S. Jindou, R. Lamed, H. J. Flint, E. A. Bayer, and B. A. White Diversity and strain specificity of plant cell wall degrading enzymes revealed by the draft genome of Ruminococcus flavefaciens FD-1. PLoS ONE 4:e dx.doi.org/ /journal.pone Berg Miller, M. E., C. J. Yeoman, N. Chia, S. G. Tringe, F. E. Angly, R. A. Edwards, H. J. Flint, R. Lamed, E. A. Bayer, and B. A. White Phage bacteria relationships and CRISPR elements revealed by a metagenomic survey of the rumen microbiome. Environ. Microbiol. 14: org/ /j x. Brulc, J. M., D. A. Antonopoulos, M. E. Berg Miller, M. K. Wilson, A. C. Yannarell, E. A. Dinsdale, R. E. Edwards, E. D. Frank, J. B. Emerson, P. Wacklin, P. M. Coutinho, B. Henrissat, K. E. Nelson, and B. A. White Gene-centric metagenomics of the fiberadherent bovine rumen microbiome reveals forage specific glycoside hydrolases. Proc. Natl. Acad. Sci. USA 106: dx.doi.org/ /pnas Callaway, T. R., S. E. Dowd, T. S. Edrington, R. C. Anderson, N. Krueger, N. Bauer, P. J. Kononoff, and D. J. Nisbet Evaluation of bacterial diversity in the rumen and feces of cattle fed different levels of dried distillers grains plus soluble using bacterial tag-encoded FLX amplicon pyrosequencing. J. Anim. Sci. 88: org/ /jas Cardona, S., A. Eck, M. Cassellas, M. Gallart, C. Alastrue, J. Doré, F. Azpiroz, J. Roca, F. Guarner, and C. Manichanh Storage conditions of intestinal microbiota matter in metagenomic analysis. BMC Microbiol. 12: org/ / Carroll, I. M., T. Ringel-Kulka, J. P. Siddle, T. R. Klaenhammer, and Y. Ringel Characterization of the fecal microbiota using high-throughput sequencing reveals a stable microbial community during storage. PLoS ONE 7:e journal.pone Chaucheyras-Durand, F., E. Chevaux, C. Martin, and E. Forano Use of yeast probiotics in ruminants: Effects and mechanisms of action on rumen ph, fibre degradation, and microbiota according to the diet. Pages in Probiotic in Animals. E. C. Rigobelo, ed. InTech, Rijeka, Croatia. Chaucheyras-Durand, F., S. Masséglia, G. Fonty, and E. Forano Influence of the composition of the cellulolytic flora on the development of hydrogenotrophic microorganisms, hydrogen utilization, and methane production in the rumens of gnotobiotically reared lambs. Appl. Environ. Microbiol. 76: AEM Deng, W., D. Xi, H. Mao, and M. Wanapat The use of molecular techniques based on ribosomal RNA and DNA for rumen microbial ecosystem studies: A review. Mol. Biol. Rep. 35: Dunne, J. C., D. Li, W. J. Kelly, S. C. Leahy, J. J. Bond, G. T. Attwood, and T. W. Jordan Extracellular polysaccharide-degrading proteome of Butyrivibrio proteoclasticus. J. Proteome Res. 11: dx.doi.org/ /pr200864j. Durso, L. M., G. P. Harhay, T. P. Smith, J. L. Bono, T. Z. Desantis, D. M. Harhay, G. L. Andersen, J. E. Keen, W. W. Laegreid, and M. L. Clawson Animal-to-animal variation in fecal microbial diversity among beef cattle. Appl. Environ. Microbiol. 76: AEM Ferrer, M., A. Beloqui, O. V. Golyshina, F. J. Plou, A. Neef, T. N. Chernikova, L. Fernández-Arrojo, I. Ghazi, A. Ballesteros, K. Elborough, K. N. Timmis, and P. N. Golyshin Biochemical and structural features of a novel cyclodextrinase from cow rumen metagenome. Biotechnol. J. 2: Ferrer, M., A. Ghazi, A. Beloqui, J. M. Vieites, N. López-Cortés, J. Marín-Navarro, T. Y. Nechitaylo, M. E. Guazzaroni, J. Polaina, A. Waliczek, T. N. Chernikova, O. N. Reva, O. V. Golyshina, and P. N. Golyshin Functional metagenomics unveils a multifunctional glycosyl hydrolase from the family 43 catalysing the breakdown of plant polymers in the calf rumen. PLoS ONE 7:e pone Ferrer, M., O. V. Golyshina, T. N. Chernikova, A. N. Khachane, D. Reyes-Duarte, V. A. P. M. D. Santos, C. Strompl, K. Elborough, G. Jarvis, A. Neef, M. M. Yakimov, K. N. Timmis, and P. N. Golyshin Novel hydrolase diversity retrieved from a metagenomelibrary of bovine rumen microflora. Environ. Microbiol. 7: Findley, S. D., M. R. Mormile, A. Sommer- Hurley, X. C. Zhang, P. Tipton, K. Arnett, J. H. Porter, M. Kerley, and G. Stacey Activity-based metagenomic screening and biochemical characterization of bovine ruminal protozoan glycoside hydrolases. Appl. Environ. Microbiol. 77: dx.doi.org/ /aem Flint, H. J., K. P. Scott, S. H. Duncan, P. Louis, and E. Forano Microbial degradation of complex carbohydrates in the gut. Gut Microbes 3: Fonty, G., P. Gouet, J. P. Jouany, and J. Senaud Ecological factors determining establishment of cellulolytic bacteria and protozoa in the rumen of meroxenic lambs. J. Gen. Microbiol. 129: Fonty, G., P. Gouet, J. P. Jouany, and J. Senaud Establishment of the microflora and anaerobic fungi in the rumen of lambs. J. Gen. Microbiol. 133: dx.doi.org/ / Genini, S., B. Badaoui, G. Sclep, S. C. Bishop, D. Waddington, M. H. Pinard van der Laan, C. Klopp, C. Cabau, H. M. Seyfert, W. Petzl, K. Jensen, E. J. Glass, A. De Greeff, H. E. Smith, M. A. Smits, I. Olsaker, G. M. Boman, G. Pisoni, P. Moroni, B. Castiglioni, P. Cremonesi, M. Del Corvo, E. Foulon, G. Foucras, R. Rupp, and E. Giuffra Strengthening insights into host responses to mastitis infection in ruminants by combining heterogeneous microarray data sources. BMC Genomics 12: dx.doi.org/ / Griffith, G. W., S. Baker, K. Fliegerova, A. Liggenstoffer, M. van der Giezen, K. Voigt, and G. Beakes Anaerobic fungi: Neocallimastigomycota. IMA Fungus 1: Guo, Y. Q., J. X. Liu, Y. Lu, W. Y. Zhu, S. E. Denman, and C. S. McSweeney Effect of tea saponin on methanogenesis, microbial community structure and expression of mcra gene, in cultures of rumen microorganisms. Lett. Appl. Microbiol. 47: Hess, M., A. Sczyrba, R. Egan, T. W. Kim, H. Chokhawala, G. Schroth, S. Luo, D. S. Clark, F. Chen, T. Zhang, R. I. Mackie, L. A. Pennacchio, S. G. Tringe, A. Visel, T. Woyke, Z. Wang, and E. M. Rubin Metagenomic discovery of biomass-degrading genes and genomes from cow rumen. Science 331: science Hobson, P. N., and G. Fonty Biological models of the rumen function. Pages in The Rumen Microbial Ecosystem. P. N. Hobson and C. S. Stewart, ed. Blackie Acad. Prof., London, UK. Hong, S. H., J. S. Kim, S. Y. Lee, Y. H. In, S. S. Choi, J.-K. Rih, C. H. Kim, H. Jeong,