Progress 01/01/16 to 12/31/18
Outputs
Target Audience:During the reporting year our target audiences have included research scientists, specifically microbial ecologists and animal scientists. Research findings from this project were shared at conferences, including the American Society for Animal Science, intra-university meetings at the University of Nebraska, and at invited talks at other campuses. Changes/Problems:The first objective of the research in this fellowship was to utilize targeted single cell and metagenomic approaches to glean more information on the genetic diversity of rumen methanogens. Targeted approaches were proposed because of the low abundance of methanogens, typically accounting for ~1-3% of microbial cells in the rumen ecosystem. However, the proposed targeted approaches failed to yield quality data or data demonstrating enrichment for methanogen DNA. Below, I summarize those efforts and the pitfalls that were associated with the targeted approaches. Lastly, I touch on the approach we implemented instead that generated 37 methanogens genomes, the majority of which are novel and represent the first genomic references of several methanogen species. First, as part of the proposed protocol in experiment 1 of objective 1, we attempted to use fluorescent in situ hybridization (FISH) to label methanogen cells and sort the labeled cells into isolated wells for amplification and ultimately whole genome sequencing. However, we could not reliably and consistently hybridize population specific probes to methanogens to generate sufficient fluorescent intensity for FACS sorting. We feel the downfall of the approach was related to the inability to partially breakdown the cell walls of methanogens while still maintaining their structural integrity. Most protocols for FISH are applicable to bacteria. Methanogens have unique cell wall characteristics though and require modifications to standard protocols. Issues relating to adapting protocols for the efficient break down of methanogen pseudopeptidoglycan were a reoccurring issue with the proposed work in this fellowship. Ultimately, since we could not reliably adhere probes and sort pure methanogen populations, and the potential downstream issues related to genomic amplification in microliter volumes, we attempted to design a microfluidic single-cell genomics pipeline. The lab mentor purchased a microfluidic device in an attempt to alleviate the above issues and increase the throughput of the genomics application. Through this approach, we were able to isolate single cells within droplets along with lysis reagents and agarose. After lysis, the agarose was cooled resulting in ~50-micron agarose droplets with volumes on the picoliter scale. A significant amount of time was spent trying to optimize the lysis reaction in order to produce efficient amplification of the cell's DNA. However, lysing the methanogen cell wall was more difficult than anticipated and we believe this is related to the difficulty in lysing methanogen cells through mechanical processes that are complimentary with maintaining the integrity of the droplet. Further, no available enzymes target the cell walls of methanogens and products such as lysozyme are ineffective due to the lack of peptidoglycan in the methanogen cell wall. In objective 1 we also proposed to implement stable isotope probing (SIP) experiments to target metagenomic sequencing of methanogens linked to anaerobic methane oxidation (methanotrophs) and methylamine utilization (methylotrophs) with 13-carbon labeled methane and methylamine, respectively. We could not identify 13-carbon labeled DNA after incubation of rumen samples with 13-carbon labeled methane. This indicates methane metabolizing organisms are either absent in the rumen or represent a rare minority that we are unable to detect through SIP techniques. While we were able to isolate and sequence 13-carbon DNA after incubation with 13-carbon labeled methylamine, the isolated DNA displayed little enrichment for the population of interest. We remain unsure why poor enrichment occurred, but we suspect the in vitro enrichments resulted in pH and hydrogen concentration changes that favored the utilization of the substrate by other microbial populations. As a result of the issues in objective 1, we assembled rumen microbial genomes through the application of metagenomic binning across a large collection of publicly available and novel rumen metagenomic datasets. By leveraging ~3.3 tera base pairs of data from 449 rumen metagenomes, we produced 2,150 high quality and near-complete genomes from the rumen ecosystem. These genomes will be made available to the research community and will serve as valuable references for future studies. Our findings suggest that with these new genomes, upwards of 50-80% of rumen metagenome data can now be mapped to a reference genome. Previously, only 2-10% of data could be associated with a rumen species. In the second objective of the proposed research, we set out to identify ecological patterns of rumen viral-methanogen interactions. Methanogens, and thus their associated viral populations, are of low abundance in the ecosystem, and therefore, sequencing the rumen viral fraction would elicit few methanogen viruses and make it difficult to link the identified viruses to their microbial host. Our recent publication on the rumen virome supports this notion. Since we had little success in developing approaches to enrich for methanogens in objective 1, we focused our analysis on the viruses found as prophages within the metagenome-assembled methanogen genomes. In support of this, the majority of viruses in gut ecosystems, including the rumen, are thought to typically operate via the lysogenic cycle. This approach yielded the identification of 20 lysogenic viruses from the 37 methanogen genomes. This represents a substantial improvement in our knowledge of viruses infecting rumen methanogens. Further, by mapping metagenomic reads from previous experiments to the prophage regions, we elicited the population dynamics and preferred replication cycle of the recovered viruses and their hosts during the feeding of contrasting diets representative of high and low methane production conditions. Despite changes to the proposed research, the outputs of this work will be of tremendous value to the rumen microbiology field. The assembled genomes allow for improved annotation of rumen metagenomic datasets and will thus improve our understanding of microbial population dynamics in the rumen ecosystem. Further, we accomplished the overarching goal set forth in the research proposal to improve the genomic representation of rumen methanogens. Finally, we successfully linked thousands of viruses to their host microbes. Information on viral-host interactions will be important to build a more thorough and predictive understanding of rumen metabolism. What opportunities for training and professional development has the project provided?The USDA-NIFA fellowship provided professional development and training opportunities for the PI in rumen microbiology, ruminant nutrition, and data science. The PI attended conferences in the fields of microbiology and animal science and presented results tied to this project. These opportunities allowed for the development of the PI's professional network and bridging expertise in microbiology and animal science by discussing the associated work of the project with members of the microbial ecology and ruminant nutrition research fields. Further, the fellowship allowed the PI to acquire training from the Data Carpentry organization to better prepare the PI to teach fundamental data science skills. The PI has begun to utilize this training to teach workshops on introductory computational skills, data management, and computational biology to undergraduate and graduate students. How have the results been disseminated to communities of interest?Results stemming from this project have recently been submitted for publication in the journal of Nature Microbiology. Additionally, work from this fellowship has been shared with the appropriate research communities by attending conferences and giving invited research talks at universities. The targeted audiences of these events ranged from largely undergraduate students to more senior researchers in the field of microbial ecology. What do you plan to do during the next reporting period to accomplish the goals?
Nothing Reported
Impacts
What was accomplished under these goals?
In regard to the first objective, we attempted several approaches to isolate single methanogen cells or enrich for methanogens through single cell sorting and stable isotope probing. The original outlined approach was to utilize fluorescent activated cell sorting (FACS) to partition single methanogen cells for sequencing. However, hybridizing a probe to sort individual methanogen cells while maintaining DNA quality for sequencing proved not feasible in our hands. In a second effort, we attempted to isolate single methanogen cells through microfluidic assisted encapsulation in micron-sized agarose droplets. While we were successful in partitioning single cells into droplets, we were unable to reliably lyse methanogen cells while maintaining the structure of the agarose matrix. Additionally, in the second experiment tied to the first objective, we outlined an approach to selectively sequence methanogen DNA through stable isotope probing with 13-carbon labeled carbon dioxide and methylamine. However, the recovered 13-carbon DNA did not demonstrate enrichment for methanogens. The above research challenges are elaborated upon in the "Changes/Problems" section of this report. Ultimately, to address the challenges in the two experiments of the first objective, we turned to metagenomic binning as an approach to assemble methanogen genomes from rumen metagenomic data. We recovered 2,150 metagenome-assembled genomes by leveraging the genomic properties and co-abundance patterns of assembled contigs derived from 449 publicly available and new rumen metagenome datasets. The metagenomes were mostly from cattle and sheep, but also included data collected from moose, deer, and bison. The resulting genomes met the following minimal quality criteria: >75% complete, <10% contamination, N50 >10,000 base pairs, and <500 contigs. The median completeness and contamination of the genomes was 88% and 0.5%, respectively. Thus, genomes of the resulting dataset should be considered near-complete with minimal contamination. Due to the low abundance of rumen methanogens and the complexity of rumen of the rumen microbial ecosystem, methanogens are difficult to assemble from metagenomes and remain fragmented. By leveraging co-abundance patterns across a large number of samples we were able to construct 37 archaeal genomes, all of which are methanogens from the Euryarchaeota phylum. Of the 37 methanogen genomes, only 10 were found to have a conspecific strain in standard reference databases. Thus, 27 of these genomes represent novel archaeal species that have not been previously described. A substantial fraction of the methanogen genomes were of the Methanobrevibacter and Methanosphaera genera, but two methylotrophic methanogen genomes of Methanomethylophilaceae were also recovered through the metagenomic binning process. While the resulting dataset expands upon our knowledge of the genomic diversity of rumen methanogens, it also more broadly advances our understanding of the structure-function relationship of the rumen microbial ecosystem. Nearly half (49.1%) of the recovered microbial genomes represent undescribed species, providing the first genomic representative for many rumen microbes. The resulting genomes will serve as a scaffold for future studies of the rumen ecosystem by improving the classification of metagenomic and metatranscriptomic data as well as allowing more genome-centric analyses. Improved classification rates and genome-centric analyses may aid in better identifying actionable insights into the roles of rumen microbes in shaping host traits. Genome-centric approaches can be leveraged to better link rumen microbial populations to metabolic processes. To demonstrate, we linked microbial populations to the encoding and expression of secondary metabolites. This analysis resulted in the identification of over 3,000 biosynthetic gene clusters, including 670 putative bacteriocin, sactipeptide, lanthipeptide, and lassopeptide clusters, providing evidence that the rumen is a rich resource of alternative peptides and metabolites for human and animal therapeutics. In objective 2 of the proposed research, we set out to identify putative viral-methanogen interactions. The original intention was to sequence viral enrichments and identify lysogenic and lytic viral signatures that could be bioinformatically linked to methanogen genomes. However, due to the low abundance of methanogens, and thus, a low abundance of viruses infecting methanogens, eliciting viral interactions from viral enrichments alone was difficult. Yet, in the methanogen genomes described above, we were successful in identifying prophage sequences in 20 of the 37 methanogen genomes, indicating lysogeny is prevalent within rumen methanogens. The viral prophage regions served as a reference to map metagenomic data and examine the ecological dynamics of methanogens and their associated viruses. Using data from previously sequenced metagenomes, we did not note a significant change in the abundance of methanogens or methanogen viruses in high and low methane emitting cattle. Additionally, we did not identify a dietary influence on the mode of infection for these viruses. Rather, across most metagenomes, we identified the virus in its prophage form and rarely found evidence for the viruses replicating via the lytic cycle. This finding may have consequences for the limits of phage therapy to control methanogen abundances. Viruses predate on microbial hosts but may also drive adaption of microbes. We did not identify auxiliary metabolism genes within the genomes of viruses infecting methanogens, indicating viruses may not encode adaptive metabolic functions for their host. This finding is in contrast to previous research we have conducted on phages infecting bacteria in the rumen. While data analysis for this fellowship has focused on methanogens, viral-infection networks for the 2,150 rumen microbial genomes recovered in this work were also completed. This will be a valuable resource for the rumen microbiology community to better elicit the roles of viruses in shaping rumen microbiome composition and function. Future research may focus on culturing the methanogen microbes described above. The metabolic pathways encoded within the genome sequences may provide a blueprint for the specific growth requirements needed to isolate more rumen methanogen populations. Culturing rumen methanogens will also allow for the isolation of viruses infecting these populations as well. Once viruses are isolated, they could be tested in vitro and in vivo for the potential to decrease the abundance of specific methanogen populations, thus potentially decreasing methane emissions. However, phages can have numerous indirect effects on microbial communities and may break down mutualisms, such as those that exist between methanogens and hydrogen producers. The viruses identified in this work may also be mined for novel virally encoded genes with anti-methanogen activity. Additional future work may focus on eliciting the ecological conditions (i.e. dietary) that may favor phage predation via the lytic cycle to suppress the abundance of methanogen populations. Overall, much deeper sequencing will likely be needed in order to better link the ecological and evolutionary dynamics of rumen methanogens and their associated viral populations. The viruses identified in this work improve the knowledge of lysogenic viruses interacting with rumen methanogens and provides valuable viral references that will underpin analyses and guide the design of future experiments. Overall, the work in this fellowship has provided an improved genomic framework to understand the composition and function of the rumen microbial ecosystem.
Publications
- Type:
Journal Articles
Status:
Submitted
Year Published:
2019
Citation:
Christopher L. Anderson and Samodha C. Fernando. Insights into rumen microbial biosynthetic gene cluster diversity through genome-resolved metagenomics. Nature Microbiology, 2019.
- Type:
Theses/Dissertations
Status:
Awaiting Publication
Year Published:
2019
Citation:
Christopher L. Anderson. Diversity, dynamics, and drivers of the rumen microbial ecosystem. University of Nebraska-Lincoln, 2019.
|
Progress 01/01/17 to 12/31/17
Outputs
Target Audience:During the reporting year our target audiences have included research scientists, specifically microbial ecologists. Research findings from this project were shared at conferences, including the American Society for Animal Science (Midwest Meeting), and intra-university meetings at the University of Nebraska. We expect the target audiences to expand to more diverse audiences in the second year of the project as issues with protocols are overcome (see Changes/Problems section for details). Changes/Problems:We have had problems implementing the proposed protocol in experiment 1 of objective 1. We planned to use fluorescent in situ hybridization (FISH) to label methanogen cells and sort the labeled cells into isolated wells for amplification and ultimately whole genome sequencing. However, we wanted to move away from doing amplification in microliter volumes due to known biases with multiple displacement amplification in such volumes. Other studies have suggested doing multiple displacement amplification in smaller volumes reduces such biases. Further, we had issues labeling the mcrA gene of methanogens for sorting as the resulting hybridization did not produce enough fluorescent intensity for FACS sorting. As a result, the lab mentor purchased a microfluidic device in an attempt to alleviate the above issues and increase the throughput of the application. Using the equipment we have been able to isolate single cells within picoliter droplets along with lysis reagents and agarose. After lysis, the agarose is cooled resulting in ~50 micron agarose droplets with volumes on the picoliter scale. The current year has been spent trying to optimize this reaction to produce efficient amplification of the cell's DNA. However, this has proven more difficult than anticipated and we believe this is related to the difficulty in lysing methanogen cells through mechanical processes that are complimentary with microfluidics. Further, no available enzymes target the cell walls of methanogens and products such as lysozyme are ineffective due to the lack of peptidoglycan in the methanogen cell wall. We are still committed to solving the above problems, as this approach would be broadly applicable to other fields of microbial ecology. The use of microfluidics combined with probes and flow cytometery would allow a researcher to target any cell population of interest and generate population-specific metagenomes or sorted single cell amplifications. Further, in experiment 2 of objective 1, we failed to detect enrichments of cell populations of interest after sequencing the 13-carbonDNA fractions following incubation of a rumen sample with the 13-carbonmethylamine substrate. It seems either the incubation time was not long enough to produce adequate labeling or the conditions of the incubation resulted in inefficient labeling. As a result, we are hoping to identify the populations we sought through experiment 2 by using the before-mentioned single cell microfluidic approach or through the genome binning of rumen metagenomes. In case we are not able to overcome this lysis and amplification issues related to the single-cell genomics approach above, we have also implemented a genome-binning approach using publically available rumen metagenomes. In this approach, we assemble data from these previous studies and use modern genome-binning algorithms that bin assembled contigs into groups that are representative of approximate species-level genomes. These binning algorithms rely heavily on the fact that fragments of the same genome should have a similar co-abundance pattern across multiple samples. By leveraging 3.3 tera base pairs of data from 384 rumen metagenomes, we have produced ~1,500 high quality and near-complete genomes from the rumen ecosystem. These genomes could serve as a scaffold to future studies, allowing upwards of 50-80% of rumen metagenome data to be mapped to a species with a reference genome. Previously,only 2-10% of data could be associated with a rumen species. Due to the low abundance of rumen methanogens and the complexity of rumen metagenomes, typically, methanogens are difficult to assemble from metagenomes and remain fragmented. By collecting such a large data set and leveraging co-abundance patterns across 384 samples, we have been able to construct near-complete genomes of methanogens. We are currently in the process of describing the diversity of these assembled genomes in the context of other rumen methanogens previously identified in the rumen and other ecosystems. Ultimately, we do not believe these changes will have dramatic alterations on the proposed outcomes of the project. We aim to submit 2 manuscripts in regards to the project in the next period on the novel diversity of methanogens we uncover and will further disseminate the results via conferences in 2018. What opportunities for training and professional development has the project provided?The project has provided professional training opportunities for one graduate student (PI) in ruminant nutrition, rumen microbiology, microbial ecology, and bioinformatics. The PI had the opportunity to attend and communicate with others in the field at the American Society for Animal Science meeting in Omaha, Nebraska. These opportunities allowed the PI to discuss outcomes and details surrounding the current project while networking with others in field. At this meeting, the PI receivedtheYoung Scholar award for the Midwest Section of the American Society for Animal Science. This award was based partly on the results of the current project. How have the results been disseminated to communities of interest?Results related specifically to this project have not been formally disseminated as publications in the current period due to the progress discussed above and in the Changes/Problems section. However, the PI did attendaAmerican Society for Animal Science meetingand presented work related to the current project.At the conference the PI had the opportunity to discuss the current project informally with members of the microbial ecology and ruminant nutrition research fields. This allowed the PI to gauge interest in the project and discuss challenges and potential solutions to issues with the current project. Research on the diversity of methanogens in other environments was the topic of a session at the conference, and indicated the research outcomes of the project still fill a need in the community. What do you plan to do during the next reporting period to accomplish the goals?We will continue to work on a protocol to generate single cell methanogen genomes in a high throughput fashion using microfluidics (Objective 1). This will still provide the most value to ruminant microbiology as a tool for others to use and invaluable data to guide future experiments. In addition, in case we continue to be presented with problems in generating single cell genomes, we will continue to analyze the genomes assembled through publically available rumen metagenomes. These genomes will also serve as a valuable resource for rumen microbiology moving forward. Specifically, these genomes provide the first reference genome for many rumen bacteria and archaea populations. Further, viral-bacterial interactions can be mined from these assembled genomes, providing needed insight into the drivers of viral-bacterial interactions in the rumen ecosystem (Objective 2). Work regarding the targeted approaches to identify novel methanogens (whether it be single cells or genome from metagenomes) will be disseminated throughpublications and at a conference in 2018. In regards to community outreach, in collaboration with my primary mentor, Dr. Samodha Fernando, we hope to restart our annual summer camp for children at the Lincoln Children's Museum. Further, we continue to visit a local middle school to teach hands-on microbiology experiments. Specifically, we try to focus on the fact that microbes are not always bad, but instead are largely a benefit to our life. Additionally, the PI has applied to become a Data Carpentry instructor and teachworkshops through the organization (http://www.datacarpentry.org). Data Carpentry aims to provide workshops with the goalof teaching research scientistsfundamental data skills. This would allow the PI to share and teach the skills acquired during the project with other students.
Impacts
What was accomplished under these goals?
In regards to the first experiment of objective 1, the developed plan was to use a fluorescent in situ hybridization (FISH) protocol to label the mcrA gene of methanogens in a rumen sample and subsequently sort single cells based on the fluorescent signal with fluorescence-activated cell sorting (FACS). However, we experienced issues with the proposed protocol. In brief, the fluorescent signal after hybridization of the probe to the single copy mcrA gene was not intense enough to be used for cell sorting. Further, research published in the previous yeardescribed novel methanogens with diverse mcrA sequences that would not hybridize to current mcrA primer sets. Consequently, moving into the second year of the project we elected to move to a different protocol to overcome these limitations. Using microfluidics we have been able to isolate single cells within picoliter droplets along with lysis reagents and agarose. After lysis, the agarose is cooled resulting in ~50 micron agarose droplets with volumes on the picoliter scale. However, we have had troubles either reliably lysing methanogen cells within the droplets or amplifying DNA to detectable limits. We still believe this protocol can provide the best way forward to study rumen methanogen diversity and holds promise in other avenues of microbial ecology as well. Consequently, we are continuing to explore the cause of the above issues and how to alleviate them. In particular, we are beginning to implement new techniques to lyse isolated cells in the hopes that these new approaches will better break down the cell wall of methanogens, which are notoriously difficult to lyse. However, in the event that we are unable to reliably lyse and amplify DNA from the target cell population, we have alsobegan genome-binning approaches on publically available rumen metagenomes to assemble rumen methanogen genomes (see Changes/Problems for more details). With this approach, we have downloaded all large rumen metagenomes that have been deposited online, assembled the data, and used modern genome binning algorithms to assemble high-quality, near-complete genomes (greater than 80% complete and less than 5% contamination based on conserved single copy gene analysis). Using this bioinformatic approach, we have assembled ~1,500 bacteria and archaea genomes from the rumen ecosystem. We have begun to pull out the genomes of interest as it relates to the current project and are in the process of detailing the diversity of rumen methanogens we were successful in assembling. Beyond the current project, this data will prove useful to rumen microbiology and microbial ecology at large. It will aid rumen microbiology through 1) serving as a reference to map data from future rumen microbial metagenomic studies (previously <20% of rumen metagenomic data had a known genome reference); 2) expand the diversity of rumen genomes with a known reference; 3) serve as a source of novel carbohydrate metabolism genes and thereforeproviding key information regarding the relationship between carbohydrate function and phylogeny; 4) serve as a source of rumen genomes that can be mined for viral signals and other signals of viral-bacterial interactions (Objective 2 of the current study). In the second experiment of objective 1, we proposed to use stable isotope probing (SIP) to identify rumen microbes involved in anaerobic methane oxidation (methanotrophs) and methylamine utilization (methylotrophs) with 13-carbon labeled methane and methylamine, respectively. We could not identify 13-carbon labeled DNA after incubation of rumen samples with 13-carbon labeled methane. This indicates methane metabolizing organisms are either absent in the rumen or represent a rare minority that we are unable to detect through SIP techniques. While we were able to isolate and sequence 13-carbon labeled DNA after incubation with 13-carbon methylamine, the isolated DNA displayed little enrichment in the population of interest. At this time, we remain unsure why such little enrichment was found, but we suspect the incubation of the rumen sample with the 13-carbonsubstrate was not of sufficient time or produced conditions not suitable for uptake of the substrate needed for labeling. However, using the publically available rumen metagenomes and associated genome-binning analysis described above, we were able to assemble a few near-complete genomes containing genes related to the methylamine utilization. We are in the process of comparing these genomes to others with similar metabolism to provide better insight into the physiological characteristics of these genomes.
Publications
|
Progress 01/01/16 to 12/31/16
Outputs
Target Audience:During the reporting year our target audiences have included research scientists, specifically microbial ecologists. At the International Society for Microbial Ecology (ISME) meeting the PI had the opportunity to discuss the project, its benefits to the field, and current issues surrounding the project with others in the field. We expect the target audiences to expand to more diverse audiences in the second year of the project as issues with protocols are overcome (see Changes/Problems section for details). Changes/Problems:We have had problems implementing the proposed protocol in experiment 1 of objective 1. We planned to use fluorescent in situ hybridization (FISH) to label methanogen cells and sort the labeled cells into isolated wells for amplification and ultimately whole genome sequencing. However, we wanted to move away from doing amplification in microliter volumes due to known biases with multiple displacement amplification in such volumes. Other studies have suggested doing multiple displacement amplification in smaller volumes reduces such biases. Further, we had issues labeling the mcrA gene of methanogens for sorting as the resulting hybridization did not produce enough fluorescent intensity for FACS sorting. As a result, the lab mentor purchased a microfluidic device in an attempt to alleviate the above issues and increase the throughput of the application. Using the equipment we have been able to isolate single cells within picoliter droplets along with lysis reagents and agarose. After lysis, the agarose is cooled resulting in ~50 micron agarose droplets with volumes on the picoliter scale. Currently, we are trying to optimize the amplification of the single cells within the agarose matrix. After amplification, we can sort the agarose droplets with FACS as was proposed in the initial protocol and conduct whole genome sequencing on the resulting sorted methanogen cells. Hybridizing a probe to the mcrA gene will result in increased fluorescent intensity when done after amplification of the genome due to the increase in number of target genes. Our revised protocol using microfluidics should alleviate concerns with multiple displacement amplification in large volumes by conducting the reaction in picoliter volumes and increase the throughput of the protocol. Microfluidics allows for the extension of the protocol to millions of droplets. Increasing the throughput of our application to millions of droplets will allow us to access more methanogen cells and is necessary when investigating the diversity of a low abundance population (methanogens compromise ~1-3% of the rumen microbiota). If for some reason we are still unable to move ahead with the proposed alterations above, we will conduct mini-metagenomes in which a small pool of cells are FACS sorted into wells, amplified, and sequenced. Next, we can conduct PCR screening to identify pools containing methanogens of interest and sequence the mini-metagenomes. This will increase access to methanogens in comparison to "whole" metagenomics in which the entire rumen microbiota is sequenced in one collective sample. Typically, assemblies from rumen metagenomes contain very few fragments from methanogens due to issues caused by the complexity of the rumen microbiota. Ultimately, we do not believe these changes will have dramatic alterations on the proposed outcomes of the project. We aim to submit 2 manuscripts in regards to the project in the next period regarding the novel diversity of methanogens we uncover and will further disseminate the results via conferences as well as the USDA directors meeting in 2017. What opportunities for training and professional development has the project provided?The project has provided professional training opportunities for one graduate student (PI) in ruminant nutrition, rumen microbiology, microbial ecology, and bioinformatics. The PI had the opportunity to attend and communicate with others in the field at the Beef Methane Conferences (May11-12, Lincoln, NE) and the International Society for Microbial Ecology conference (August 21-26, Montreal). These opportunities allowed the PI to discuss outcomes and details surrounding the current project while networking with others in field. How have the results been disseminated to communities of interest?Results related specifically to this project have not been formally disseminated in the current period due to the progress discussed above and in the Changes/Problems section. However, the PIdid attened the 2016 International Society for Microbial Ecology (ISME) conference and presented a poster on research that was largely accomplished prior to the current project. At the conference the PIhad the opportunity to discuss the current project infromally with members of the microbial ecology research field. This allowed the PI to gauge interest in the project and discuss challenges and potential solutions to issueswith the currrent project. Research on the diversity of methanogens in other environments was the topic of a session at the conference, and indicated the research outcomes of the project still fill a need in the community. What do you plan to do during the next reporting period to accomplish the goals?We will continue to work on a protocol to generate single cell methanogen genomes in a high throughput fashion using microfluidics (Objective 1). This will still provide the most value to ruminant microbiology as a tool for others to use and invaluable data to guide future experiments. However, if we continue to be presented with problems in generating single cell genomes we will move to a different approach, mini-metagenomics, to target methanogens for genome sequencing. Mini-metagenomics includes sorting small groups of cells (~5-10 cells) into individual wells, amplifying the genetic content of these cells, and sequencing the smaller population. By using PCR we can identify which collections of cells include methanogens and target them for sequencing. Post sequencing, contig binning methods can leverage the fact that cells are present in multiple metagenomes to better improve assembly of methanogens in the targeted mini-metagenomes. Mini-metagenomics is an up-and-coming technique in microbial ecology and presents fewer technical difficulties than single cell genomics does.Further, it is well within the realms of our technical expertise and equipment on campus. Viruses have been enriched from the sample these targeted methanogens are being sequenced from. The viral fraction of the sample will be sequenced in early 2017 alongside the methanogens (Objective 2). This will allow us to identify viral signatures in the sequenced methanogens, from which we can infer lytic and lysogenic viral-methanogen interactions. Work regarding the targeted approaches to identify novel methanogens (whether it be single cells or mini-metagenomes) will be disseminated through a publicationand at a conference in 2017. In regards to the stable isotope probing (SIP) aspects of the project, we will begin to annotate the metagenomic fragments sequenced from the 13-carbon methylamine incubation (Objective 1). We will summarize both the taxonomic and functional annotations to detail the novelty of the species involved in the metabolism of methylamine and the pathways of methane production. We expect the labeled organisms to include Thermoplasmata methanogen species, which have been proposed to be targets for methane control in previous studies. Genetic descriptions of this population are largely absent in current databases though. These results will bedisseminated through a second publication in late 2017. The abundance of methanogen populations identified via both single cell and SIP methods can then be tracked across previously sequenced rumen metagenomes from the lab. These metagenomes were sequenced from differentgrowing and finishing diets that resulted in contrastingmethane production.Mapping reads from these metagenomes to the newly identified populations will allow us to identify methanogens unique to low and high methane production. In regards to community outreach,in collaboration with my primary mentor, Dr. Samodha Fernando, we hope to restart our annual summer camp for children at the Lincoln Children's Museum. This activity had been held in previous years, but due to unforeseen circumstances regarding staffing at the museumwe were unable to hold the event in 2016. Activities at the camp are geared to teach children about microbial life, specifically that microbes are not always bad, but instead are largely a benefit to our life. Additionally, in late 2017 the PI plans to apply to become a Data Carpentry instructor and teach a workshop(http://www.datacarpentry.org). Data Carpentry aims to provide workshops regardingthe fundamental data skills to empower researchers in today's data driven researchecosystem. This would allow the PI to share and teach the skills acquired during the project with other students.
Impacts
What was accomplished under these goals?
In regards to the first experiment of objective 1, we planned to use a fluorescent in situ hybridization (FISH) protocol to label the mcrA gene of methanogens in a rumen sample and subsequently sort single cells based on the fluorescent signal with fluorescence-activated cell sorting (FACS). However, we experienced issues with the proposed protocol (see Changes/Problems section for more detail). In brief, the fluorescent signal after hybridization of the probe to the single copy mcrA gene was not intense enough to be used for cell sorting. Further, research published this year described novel methanogens with diverse mcrA sequences that would not hybridize to current mcrA primer sets. Consequently, we elected to move to a different protocol to overcome these limitations. The mentor of the lab purchased a microfluidic system with which we can encapsulate rumen microbial cells in picoliter sized agarose droplets. The goal was then to lyse and amplify the single cells within the agarose followed by probe hybridizations to target cells of interest and sorting of methanogen cells for genomic sequencing. We have been able to isolate single cells within agarose droplets and efficiently lyse the captured cells. Further, the cells are distributed across the droplets in a Poisson distribution to ensure droplets with 2 or more cells are minimized to a rare minority of generated droplets. Currently, our efforts are focused on amplifying the genomes within the agarose matrix while minimizing DNA contamination between droplets. We believe this new protocol will allow use to isolate and sequence methanogen genomes in a high throughout fashion and maintain this is the best way to push forward the study of rumen methanogen diversity. If this proposed protocol does not work in the coming weeks, we have outlined a less novel approach below in the section detailing our plans for the next reporting period. In the second experiment of objective 1, we proposed to use stable isotope probing (SIP) to identify rumen microbes involved in anaerobic methane oxidation (methanotrophs) and methylamine utilization (methylotrophs) using 13-carbon labeled methane and methylamine, respectively. We could not identify 13-carbon labeled DNA after incubation of rumen samples with 13-carbon labeled methane. This indicates methane metabolizing organisms are either absent in the rumen or represent a rare minority that we are unable to detect through SIP techniques. We were able to isolate 13-carbon labeled DNA after incubation with 13-carbon methylamine along with the unlabeled 12-carbon DNA. We subsequently sequenced the DNA from each fraction and are in the process of analyzing these results to identify the microbes involved in methylamine metabolism, their functional content, and implications for rumen methane production. In regards to objective 2, we proposed to sequence viral metagenomes from the same samples analyzed for SIP and single cell sequencing. Currently, we have enriched viral particles from the rumen sample being used to generate single cell methanogen genomes. A viral metagenome from this enrichment will be sequenced alongside the methanogen genomes upon completing the amplification and sorting of methanogens. Afterwards, we can use the viral metagenome to identify putative viral signals within the methanogen genomes.
Publications
- Type:
Journal Articles
Status:
Published
Year Published:
2016
Citation:
Xie, Fang; Anderson, Christopher L; Timme, Kelsey R; Kurz, Scott G; Fernando, Samodha C; Wood, Jennifer R. Obesity-Dependent Increases in Oocyte mRNAs Are Associated With Increases in Proinflammatory Signaling and Gut Microbial Abundance of Lachnospiraceae in Female Mice. Endocrinology, 157, 4, 1630-1643,2016.
- Type:
Journal Articles
Status:
Published
Year Published:
2016
Citation:
Paz, Henry A; Anderson, Christopher L; Muller, Makala J; Kononoff, Paul J; Fernando, Samodha C. Rumen bacterial community composition in Holstein and Jersey cows is different under same dietary condition and is not affected by sampling method. Frontiers in Microbiology, 7, 2016.
- Type:
Journal Articles
Status:
Accepted
Year Published:
2016
Citation:
Walter, Mary E; Ortiz, Alicia; Sondgeroth, Casey; Sindt, Nathan M; Duszenko, Nikolas; Catlett, Jennie L; Zhou, You; Valloppilly, Shah; Anderson, Christopher; Fernando, Samodha; Buan, Nicole. High-throughput mutation, selection, and phenotype screening of mutant methanogenic archaea. Journal of Microbiological Methods, 2016.
- Type:
Journal Articles
Status:
Published
Year Published:
2016
Citation:
Ramirez-Ramirez, HA; Lopez, E Castillo; Jenkins, CJR; Aluthge, ND; Anderson, C; Fernando, Samodha C; Harvatine, KJ; Kononoff, PJ. Reduced-fat dried distillers grains with solubles reduces the risk for milk fat depression and supports milk production and ruminal fermentation in dairy cows. Journal of dairy science, 99, 3, 1912-1928, 2016.
- Type:
Journal Articles
Status:
Submitted
Year Published:
2016
Citation:
Anderson, Christopher L; Sullivan, Matthew B.; Fernando, Samodha C. Dietary energy drives the dynamic response of bovine rumen viral communities to dietary pertubation. Microbiome Journal, 2016.
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2106
Citation:
Christopher L. Anderson and Samodha C. Fernando. 2016. Dietary energy drives the dynamic response of bovine rumen viral communities (Viral Ecology Session, Poster 239B). International Society of Microbial Ecology, Montreal, Canada. August 21-26.
|
|