links header

email icon printer icon

Unlocking Microbial Diversity with Metagenomics


Dana Willner, PhD
The University of Queensland
St. Lucia, QLD, AUSTRALIA


d.willner@uq.edu.au


dana willnerThe human body houses a vast consortium of micro-organisms, with microbial cells estimated to outnumber human cells ten to one.1 Traditional microbiology has relied on cultivation for characterization of the human microbiome and pathogen discovery. The advent of culture-independent techniques coupled with advanced DNA sequencing technologies and bioinformatics have allowed unprecedented insight into the structure of microbial communities, demonstrating that the full extent of microbial diversity in any given ecosystem is rarely described by culturing alone.

Two culture-independent techniques are widely used to describe microbial populations: shotgun metagenomics and community profiling. Shotgun metagenomics involves the wholesale extraction and sequencing of total microbial DNA in an environment, while community profiling relies on the amplification and sequencing of the highly conserved 16S ribosomal RNA (rRNA) gene.2 Shotgun studies provide insight into microbial taxonomy (who is there?), diversity (how many are there?), and function (what are they doing?), but can be difficult to conduct on a large scale due to informatic and sequencing limitations. Community profiling describes taxonomy and diversity only, but can be easily applied to hundreds of samples, enabling clinical studies of sizeable patient cohorts.

Using a community profiling approach based on 454 pyrosequencing of the 16S rRNA gene, we investigated the microbial communities of bronchoalveolar lavage (BAL) samples from lung transplant patients with and without bronchiolitis obliterans syndrome (BOS) as well as healthy controls.3 A total of 132 samples from 57 transplant patients (27 female, 29 cystic fibrosis, 22 BOS) and 8 controls were sequenced including longitudinal samples from 16 cystic fibrosis (CF) patients at up to 6 time points ranging from 2 months to over 3 years post-transplant. Microbial community composition was compared between samples using the weighted Unifrac distance which accounts for both community membership and relative abundance.4 Principal components analysis based on weighted Unifrac distance demonstrated two major microbial community types: one dominated by Pseudomonas and one dominated by Streptococcus and Veillonella. Most cystic fibrosis patients were colonized with Pseudomonas and an inverse relationship was noted between the relative abundance of Pseudomonas and Streptococcus. Nearly all CF patients with BOS had microbial communities dominated by organisms not considered typical pathogens in CF adults including Streptococcus, Lactobacillus, Enterococcus, Neisseria, and Haemophilus, while those with high abundances of Pseudomonas, Burkholderia, and Staphylococcus were more likely to be BOS-free. A stratified exact logistic regression analysis controlling for time post-transplant demonstrated that in CF individuals, the odds of BOS were reduced by more than half when Pseudomonas was the dominant organism in the microbial community (OR= 0.37 (0.11,0.83). p=0.009). This observed effect was not due to confounding by history of rejection (p=0.94), or antibiotic treatment (p=0.32). Examination of temporal changes in individual patients demonstrated that stable, high-abundance colonization with Pseudomonas was associated with an absence of BOS. These data suggest that in CF patients, BOS is accompanied by pronounced and sometimes predictable changes in the transplant lung microbiome that could inform therapeutic regimens.

Previous studies have implicated Pseudomonas colonization in the development of BOS in CF individuals;5 however, this was based exclusively on cultivation. Our results suggest that it is not the presence of Pseudomonas, but its relative abundance in the context of the microbial community as a whole that may be correlated with transplant outcomes. All but one of the CF patients in this study cultured Pseudomonas at all-time points, but its representation in the microbiome was highly variable, ranging from less than 5% to over 95%, with lower relative abundances occurring in individuals with BOS. As opposed to culture-based studies which provide semi-quantitative or qualitative information on a subset of microbial populations, community profiling allows for characterization of all microbes and their relative abundances simultaneously, adding a previously unexplored dimension to the characterization of the respiratory microbiome.


Disclosure Statement: The author has no conflicts of interest to disclose.

References:

  1. Savage DC (1977) Microbial ecology of the gastrointestinal tract. Annu Rev Microbiol 31: 107-133. doi:10.1146/annurev.mi.31.100177.000543.
  2. Hugenholtz P, Tyson GW (2008) Microbiology: Metagenomics. Nature 455: 481-483. doi:10.1038/455481a.
  3. Willner D, Hugenholtz P, Tan ME, et al. (2012) 68 Distinct Microbial Signatures of Healthy and Failing Lung Allografts. The Journal of Heart and Lung Transplantation 31: S32. doi:10.1016/j.healun.2012.01.072.
  4. Lozupone C, Lladser ME, Knights D, et al. (2011) UniFrac: an effective distance metric for microbial community comparison. ISME J 5: 169-172.
  5. Vos R, Vanaudenaerde BM, Geudens L, et al. (2008) Pseudomonal airway colonisation: risk factor for bronchiolitis obliterans syndrome after lung transplantation? Eur Respir J 31: 1037-1045. doi:10.1183/09031936.00128607.