The Enteric Microbiota

Microbial Culture

Early studies dissecting the microbiota composition were technologically limited by culture-based techniques, which in turn relied on specialized growth media under varying conditions to identify specific microbes. This restricted our ability to identify only a small subset of organisms for which established culture conditions had been described and which accounted for 5%-15% of the intestinal bacteria we know to constitute the microbiome today. As a result, locations with limited diversity were often considered sterile given the inability to culture their resident bacteria. Today, however, nearly all locations in the body have been described to have characteristic resident microbes as a result of culture-independent sequence-based identification methodologies. Sequencing-based data have also improved our ability to culture bacteria previously considered to be unculturable. We are now able to culture a significant proportion of an individual’s fecal microbiota, using various culturing conditions, which has allowed us to determine the relevance of microbial compositional changes and the interactions and impact of individual or groups of bacteria on host phenotypes using models such as germ-free mice.

Microscopy

Early methods included scanning and transmission electron microscopy of intestinal tissue, which provided estimates of diversity based on morphology and high-resolution images of individual bacteria but did not allow bacterial identification. The use of general stains such as the Gram stain provides resolution beyond morphology, but also is insufficient for identification. Fluorescence microscopy provides the opportunity to identify bacteria by fluorescence in situ hybridization to microbe specific 16s rRNA. The increased availability of sequencing data has allowed development of more precise fluorescence in situ hybridization probes and carries the advantage of not requiring culture. The fixation methods are compatible with preserving mucus and the use of multiple probes simultaneously allows detection of several bacteria within a sample. In addition, it is 1 of the primary tools to define the biogeography of microbes within the intestine and the interaction of bacteria with the host at the mucosal surface. The advances in fluorophores, imaging, and computational tools have significantly improved our ability to visualize microbes both in vivo and in vitro. Conventional fluorescent probes require oxygen limiting their utility in vivo, but new tools using “click” chemistry allow tagging of bacteria with oxygen-independent fluorescent tags for in vivo tracking.

Next-Generation Sequencing

The early culture-independent compositional tools used denaturing gradient gel electrophoresis to separate different-sized bands that represented distinct taxonomic groups. However, with the advent of next-generation sequencing technologies (e.g., Illumina, 454, Ion Torrent, SOLiD, etc.), marker-based (16S rRNA gene) and shotgun sequencing of all genes within a community have superseded denaturing gradient gel electrophoresis, especially given the declining cost of sequencing. The marker-based approach takes advantage of the conservation of DNA sequence in the gene encoding the 16S rRNA subunit that is found in all microbes. Interceding variable regions are targeted for amplification by polymerase chain reaction, allowing simultaneous identification of different taxa within a sample. However, marker-based sequencing is limited in its ability to identify taxa beyond the genus level given the small amplicon sizes. Third-generation sequencing technologies, such as single-molecule real-time sequencing have emerged, which will likely supersede the current methodologies, given their potential to generate read lengths (continuous sequence from a single piece of DNA) of 10 kilobases.

Metagenomics

Often referred to as whole genome sequencing, or shotgun sequencing, metagenomics allows characterization of all genes in a microbial community and provides the broad functional potential of a community. It cannot, however, provide the specific functionality under a given set of conditions.

Metatranscriptomics

Transcriptomic approaches like RNAseq provide a snapshot of gene-expression profiles of microbial communities under a given condition. This data can be used to further infer differential expression of metabolic pathways using analysis tools like HUMAnN2.

Metaproteomics and Metabolomics

Metaproteomics provides comprehensive characterization of proteins, whereas metabolomics provides comprehensive characterization of small molecules and metabolites, each from microbial communities. Both approaches allow characterization of the overall metabolic state of complex communities resulting from differential gene expression among communities or the same community under different conditions. For proteomics, proteins can either be directly separated based on hydrophobicity, charge, or both, using liquid chromatography (LC) or digested to peptides via proteases such as trypsin prior to chromatographic separation followed by mass spectrometry (MS) for the parent peptide and tandem MS-MS for fragmentation information. The biggest challenge currently is the downstream bioinformatics processing because a predicted protein database needs to be constructed from metagenomic information to assign the obtained peptide sequence information to the proteins. Alternatively, the vast diversity of small molecules, and differences in properties and concentrations, require that multiple methods be used to cover the vast array of metabolites; these include separation using LC, gas chromatography, high-pressure LC, ultra-pressure LC, coupled to MS, and proton nuclear magnetic resonance spectroscopy ( ¹ H-NMR). Metabolomics can be done in a targeted or non-targeted manner and downstream processing using statistical methods allows identification of discriminative features. One of the challenges that remains is the accurate identification of metabolites in MS spectra, though there has been significant progress with multiple spectral databases such as HMDB, METLIN, and ChemSpider, all of which are being constantly updated.

Modeling Microbes In Vitro and In Vivo