Performing Department
(N/A)
Non Technical Summary
Most (>90%) bacteria in the gut microbiome are uncultured and known only from ribosomal DNA sequences. These sequences do not reveal the metabolic activities of a bacterium, representing a challenge for identifying the role of uncultured bacteria in animal nutrition and disease. For example, sequencing can link the abundance of an uncultured bacterium with improved animal feed efficiency, but it cannot reveal its metabolic activities and exact role in improved efficiency. A new tool, relying on fluorescent substrate analogs, could reveal an uncultured bacterium's most fundamental metabolic characteristic--what substrates it uses. Our central objective is to develop a tool that involves 1) incubating rumen bacteria in fluroscent glucose analogs, 2) separating out cells that take up analogs with fluorescence-activated cell sorting, and 3) 16S rDNA sequencing of those cells, revealing bacterial species (OTUs) that utilize glucose. Based on preliminary data, we estimate that our tool can identify 2,250 rumen species as glucose utilizers--several fold more than all species discovered in this history of culturing this community. The tool could be adapted to other fluorescent substrate analogs and agriculturally-important communities (soil, natural waters, plants); thus, it could cause a paradigm shift in identifying metabolic activities of uncultured bacteria. It would help determine the role of specific gut bacteria in animal nutrient efficiency and disease, for example, and has relevance to many projects addressing the microbiome.
Animal Health Component
0%
Research Effort Categories
Basic
50%
Applied
0%
Developmental
50%
Goals / Objectives
Objective 1. Identify a set of fluorescent glucose analogs taken up by pure cultures of bacteria and their glucose transportersObjective 2. Using the set of glucose analogs above, a) perform sorting of mixed rumen bacteria that take up the analogs and b) sequence those bacteria in order to identify taxa that use glucose
Project Methods
Objective 1. Identify a set of fluorescent glucose analogs taken up by pure cultures of bacteria and their glucose transportersIntroduction. Most bacteria from the rumen and other gut habitats are uncultured, and the substrates they use are unknown, but a tool relying on fluorescent glucose analogs and cell sorting could identify bacteria that use glucose. With pure cultures of rumen bacteria and E. coli, we have demonstrated one glucose analog (2-NBDG) targets the most common transporter in gut bacteria. Following this work, we will use three related analogs (1-NBDG, 3-NBDG, 4-NBDG) and determine which of the remaining glucose transporters they individually target.Wild-type rumen bacteria. The three analogs (1-NBDG, 3-NBDG, 4-NBDG) will be synthesized commercially from NBD-Cl and the appropriate glucosamine positional isomers. No glucosamine precursor is available to synthesize 5-NBDG. 6-NBDG has been synthesized, but as a 6-C glucose derivative, it cannot be phosphorylated intracellularly.To determine which transporters each analog targets, we will test transport with pure cultures of rumen bacteria (n = 14) in our collection.These species collectively possess the eight predominant glucose transporters in cultured rumen species.Following our method for 2-NBDG, we will incubate bacteria in an analog (0.05 to 100 μM for 0 to 5 min), collect and wash cells by filtration, and measure fluorescence with a fluorimeter. Each experiment will test all analogs with one bacterium and will be replicated on two separate days.Kinetics of transport (Vmax, Km) will be determined using nonlinear regression with the R statistical program, and transport will be considered positive if Vmax > 0. We will infer which analogs target which transporters from the pattern of transport across rumen species.For example, we would infer that 1-NBDG targets the Mgl ABC transporter, as predicted, if bacteria that transport it possess that transporter in common.E. coli mutants. To corroborate findings with rumen bacteria, we will test transport with mutants of E. coli as we have previously. We will order gene deletion mutants from the Keio collection and, when needed, move multiple deletions into a single strain using P1 transduction. For the four transporters not found in E. coli, we will heterologously express the transporter. Transport will be tested as above and on two separate days. Loss of transport with deletion of a transporter--or gain of transport with heterologous expression of a transporter--will indicate a transporter has activity towards an analog.Objective 2. Using the set of glucose analogs above, a) perform sorting of mixed rumen bacteria that take up the analogs and b) sequence those bacteria in order to identify taxa that use glucoseIntroduction. Objective 2 will use analogs identified in Objective 1 to sort, sequence, and identify rumen bacteria that use glucose. Mixed rumen bacteria will be prepared from cows and then incubated in a mixture of analogs. Positive cells will be sorted and be subjected to 16S rDNA sequencing and taxonomic assignment, revealing the sequences of thousands of bacteria that use glucose. We will optimize incubation and sorting conditions to increase the number of positive cells found in preliminary experiments. We will evaluate our tool with a mock community and dense culture of bacteria grown on glucose-containing medSorting, sequencing, and OTU assignment of mixed bacteria. Mixed rumen bacteria will be prepared from rumen fluid following our procedure of differential centrifugation. Rumen fluid will be collected from one of eight cows given a 50:50 grain:forage diet (2 h after feeding).Bacteria will be incubated in a mixture of analogs.We plan to use all four analogs (1-NBDG, 2-NBDG, 3-NBDG, 4-NBDG), but we will exclude an analog if it does not target glucose transporters as expected in Objective 1. Based on preliminary experiments with 2-NBDG, we will use 100 μM of each and incubation length of 5 min.After incubation, cells will be re-harvested, washed anaerobically by centrifugation, and then taken for sorting.Control experiments will be performed with unstained cells (fixed with 10% formalin before incubation in analogs).Cells will be sorted using a Sony SH800 flow cytometer with a 488-nm excitation laser.Based on preliminary results with 2-NBDG, we plan to sort out positive cells by using gates for 1) analog fluorescence (fluorescein channel) and 2) cell autofluorescence (APC-A channel).Additionally, we will use forward scatter (~0.5% threshold) to discriminate cells from electronic noise. Approximately 1,500 events/s will be achieved by adjusting cell concentration and sample pressure.With 2-NBDG, these conditions gave 94% purity during post-sort analysis.We will sort a large number (~106) of positive cells per experiment in order to yield ≥10 ng DNA. We will collect sorted cells on filters and extract DNA by bead-beating. Following our standard protocol, we will PCR amplify the V1 to V3 region of 16S rRNA gene; generate 300 bp paired-end reads with the Illumina MiSeq v. 3 platform; and assign OTUs at 97% sequence similarity with QIIME. Sequence data will be made available on the NCBI Sequence Read Archive.Incubation and sorting conditions above are tentative, and we will optimize them in early experiments. We will adjust incubation conditions (e.g., analog concentrations) using flow cytometry histograms in FlowJo and maximize separation between positive and negative cell populations.We will adjust sorting conditions (e.g., sorting rates) to maximize purity of cells during post-sort analysis. As a further check on purity, we will image sorted cells with epifluorescence microscopy and analyze images with ImageJ. Some early experiments (those prior to cell sorting) will be done on the BD Accuri C6 flow cytometer under our control.Early optimization experiments will be done with one or two cows. Final experiments will be done with eight cows sampled once (one cow per day), ensuring high coverage of OTUs.With the same sampling effort, Jami et al. detected ~4,650 OTUs (93% coverage).Dense culture of glucose-utilizing species. To evaluate our tool, we will culture rumen bacteria using selective media including glucose as the only carbohydrate. We will grow colonies to high density (~5000/plate), scape off colonies en masse, sequence, and assign OTUs as above. We will determine if OTUs identified in this dense culture had been previously identified by cell sorting of mixed bacteria (see above). Inoculum will be the same mixed bacteria used in sorting experiments above (from 8 cows sampled once).Mock community. To evaluate our tool further, we will prepare a mock community of ten pure cultures from our collection. The community will include five rumen species that take up glucose and five that do not (representing the ratio we expect for mixed rumen bacteria). Cultures will be grown to exponential phase and then combined in equal cell numbers to form the community. We will incubate cells in analog, sort, sequence, and assign OTUs as for rumen samples. We will determine if OTUs identified in our mock community correspond to the five species known to use glucose. Experiments will be replicated on two different days.