02 March 2015
By John Lees
A PacBio sequencer, which performs SMRT (single-molecule real time) sequencing.
Credit: Genome Research Limited
For many years, short-read sequencing has been the bread and butter of bacterial studies at the Wellcome Trust Sanger Institute. Now, long-read sequencing is helping us to interpret bacterial genomes and understand how they evolve.
Short-read sequencing works by splitting the DNA of various microscopic organisms into small pieces that can then be sequenced. Researchers can then rebuild these small pieces into the whole genome of the bacteria.
This type of sequencing, sometimes referred to as shotgun sequencing, has seen huge drops in price over the past few years. Whereas we used to sequence the genome of one bacterium at a time, we can now sequence many thousands of disease-causing bugs, thus forming a picture of the entire population of each species.
We typically sequence billions of the small DNA pieces, giving reads of sequence around 100 base-pairs in length. This works very well for finding small, single-base, differences between samples. These variants are ubiquitous in human DNA, and researchers working in human genetics have carried out studies which have been able to show how these relate to disease susceptibility as well as traits such as height.
In bacteria, however, we often have much larger and more complex variations in sequence that can be unique to a single strain. In these cases, longer reads are necessary to reconstruct the entire sequence.
The recent development of SMRT (single-molecule real time) sequencing has helped us start to tackle this issue. While short-read sequencing makes lots of copies of small DNA segments then reads them all at once, SMRT reads single DNA molecules one base at a time producing much longer reads. The results are slower, but the genomes are easier to reconstruct.
SMRT sequencing can be used to analyse function of restriction modification systems in Streptococcus pneumoniae.
Credit: Public Health Image Library, CDC
Not only do we get reads with mean lengths of many thousands of base pairs, we can also report methylation at each base. Methyl markers act as on and off switches on the DNA that can affect how other machinery in the cell interprets the coding sequence.
In a recent news and analysis article 'R–M systems go on the offensive', Rebecca Gladstone and I look at some of the uses of SMRT sequencing for analysing function of restriction modification systems in the bacteria Streptococcus pneumoniae (a common cause of pneumonia).
Restriction modification systems are a bacterial equivalent of an immune system, cutting invading DNA into pieces while protecting the bacteria's DNA through methylation. Under constant threat from rapidly evolving viruses, the bacterial population must be able to rapidly switch which sequence pattern they recognise, otherwise the viruses would be able to systematically avoid their defences.
A recent study used SMRT sequencing to find the different possible DNA arrangements of the restriction modification system, which allows the bacteria to survive this viral onslaught. In a concurrent study, bacteria with different DNA arrangements were grown in the lab, and SMRT technology was used to find the different methylation patterns these rearrangements cause, and their knock-on effect on virulence.
Taken together, these studies suggest that a mechanism that exists to defend against viruses also has an effect on whether or not the bacteria cause disease in humans – an important finding that may help us better understand why only some of these bacteria lead to illnesses such as pneumonia and meningitis.
Restriction modification systems are proving to be a very important part of bacterial evolution, and new technologies such as SMRT sequencing will continue to advance our ability to understand them.
John Lees a second year PhD student at the Wellcome Trust Sanger Institute. He works with Stephen Bentley and Julian Parkhill in the Pathogen Genomics group, and Jeff Barrett in the Medical Genomics group. John's research involves combining human and pathogen sequencing data derived from cases of bacterial meningitis in the Netherlands. He's currently interested in developing tools for association analysis, and applying them in this context.
- Lees J and Gladstone R (2015). R–M systems go on the offensive. Nature Reviews Microbiology. DOI:10.1038/nrmicro3435
- Croucher NJ, et al (2014). Diversification of bacterial genome content through distinct mechanisms over different timescales. Nature Communications. DOI:10.1038/ncomms6471
- Manso AS, et al (2014). A random six-phase switch regulates pneumococcal virulence via global epigenetic changes. Nature Communications. DOI:10.1038/ncomms6055