New influenza hybrids can be made in many different ways in the laboratory. I think most people don’t realise just how easy it is to mix and match viral components with nothing to start with but a GenBank sequence.
A recent paper describes the use of gene synthesis to test the infectiousness of the hemagglutinin (HA) and neuraminidase (NA) genomic segments from several different influenza viruses. From Zhang et al. (2010):
cDNA fragments encoding the full-length HA of A/South Carolina/1/1918 (GenBank Acc. No. AF117241) and A/California/05/2009 (GenBank Acc. No. FJ966952) and the full-length NA of A/Brevig_Mission/1/1918 (GenBank Acc. No. AF250356) and A/Ohio/07/2009 (GenBank Acc. No. FJ969534) were synthesized and inserted into the vector pVRC (Liuhetong Inc., Beijing, China) to produce the constructs 1918H1, 09H1, 1918N1, and 09N1, respectively , .
Similarly, cDNA fragments encoding the full-length HA and NA from H5N1 (A/Anhui/1/2005, GenBank Acc. No. DQ371928) were amplified and cloned into the vector pVRC; they are referred to as AH H5 and AH N1, respectively.
For the genomic fragments from the 1918 pandemic virus and the current H1N1 pandemic virus, the experimenters simply looked up the sequences they wanted from GenBank and ordered them from a company which synthesised the relevant genes and put them in a plasmid vector. These plasmids were then introduced into cells in culture along with elements from another virus. This resulted in creation of artificial products called pseudotyped particles (pps). This system was used to mix and match HA and NA genes from different flu viruses. The hybrids were tested for their ability to infect cells in culture.
As previous reassortment experiments have shown, hybrid viruses with genes from different influenza origins can express novel properties. From the paper:
Of the pps containing 09N1, the combination of 09N1+1918H1 had greater infectivity than the ancestor 09N1+09H1 (p<0.05). The pps containing 1918N1, 1918N1+AH H5, and 1918N1+09H1 also showed greater infectivity than their ancestor, 1918N1+1918H1. 09N1+1918H1 also showed greater infectivity than another ancestor, 1918N1+1918H1 (p<0.05). Take together, these data suggest that valid reassortment may occur not only between viruses belonging to the same subtype (H1N1), but also between viruses belonging to different subtypes (H1N1 and H5N1); furthermore, some viruses, the products of reassortment, may have greater infectivity than their ancestors.
One thing this paper demonstrates is that by simply scanning GenBank entries and ordering synthetic genes from a company, you can discover how to create flu viruses with novel properties.
The sample space described in this and the previous experiments I have discussed in this series represents only a tiny fraction of all possible influenza strains and combinations of genes dervied therefrom. We know only a small amount of what is possible.
And what we know is not reassuring.
Zhang et al. (2010) Neuraminidase and Hemagglutinin Matching Patterns of a Highly Pathogenic Avian and Two Pandemic H1N1 Influenza A Viruses. PLoS One. 5: e9167.