Wednesday, 28 August 2013

An entire H7N9 genome from a clinical specimen in one shot

Some weeks back, Ren and colleagues described their use of next generation sequencing (or "deep" sequencing; unbiased, massively redundant sequencing of all the DNA or RNA in a sample - to put it very basically) to pull out the entire genome of an influenza A(H7N9) virus from 1 clinical sample.

he new genome is called A/Jiangsu/2/2013(H7N9) and can already be found on GenBank. It's 8 segments are numbered in order from KF226105 to KF226112.

The authors found the E672K (PB2 gene segment) and I368V (PB1) mutations related to virulence and transmissibility

The caveat for this sequencing, in case you're making comparisons to the recent small sequence fragment obtained from the MERS-CoV strain detected in bat poo pellets, is that they had enough fresh human sputum (probably not freeze/thawed and left at room temperature as with bat samples) from the 54-year-old female case from Zhejiang to be able to purify, concentrate and clean up the virus before preparing the nucleic acids for NGS. That likely means a lot more virus, from a more hospitable environment for virus, and a lot less other nucleic acid and junk to interfere with the amplification and sequencing components.

Interestingly the authors found that in 22 nucleotide positions, more than a single nucleotide could be reliably identified - sequence heterogeneity that shines a light on how much viral change goes on, even within a host, as the virus keeps testing out different ways to replicate, bind and enter cells and interfere and avoid the immune response against it. Some of those changes altered the protein (amino acid) sequence.  Sometimes those changes are no good for the resultant virus and the strain with that/those changes replicate poorly, sometimes there is no change in replication or infection efficiency and sometimes the process generates a more efficient viral strain that may then grow to become the dominant strain which we cough and sneeze onto our contacts. 

This unbiased (doesn't rely on a very specific PCR, and the amplification of the most dominant sequence in a sample as do traditional Sanger sequencing methods  sequence analysis approach is a great way to find those changes in a hole-of-genome approaches - because it's likely that "favourite mutation X, Y and Z" only represents a portion of the changes that go into making more successful influenza viruses.