H5N1 Evolution Outpaces Pandemic Vaccine Selection
Recombinomics Commentary 03:58
March 4, 2008
WHO has published its latest update on “Antigenic and genetic characteristics of H5N1 virus and candidate vaccine viruses developed for potential use as human vaccines." Although it is dated February, 2008, it is already clearly out of date. H5N1 is rapidly evolving as evidenced by 8 vaccine targets already approved (in red in the phylogenetic tree in Figure 1 on page 5), the four targets developed and waiting approval (in blue) and the three selected for development (in green). The tree in this year's report is more complex than the tree from a year and a half ago.
However, in the rapidly evolving clade 2.2 which was been reported in approximately 50 countries west of China, the selections are significantly lagging H5N1 evolution. The approved isolates are from 2005 and include one of the original Qinghai isolates from China in the spring of 2005, an isolate from Mongolia in the summer of 2005, and an isolate from Turkey in the fall of 2005. These three were followed by an isolate from India in the winter of 2006, which is awaiting approval. Although Europe reported little activity in late 2006, early 2007, clade 2.2 was rapidly evolving in Egypt as well as at Uvs Lake in Mongolia in the summer of 2006.
These new and improved versions of H5N1 were easily seen a year ago at the beginning of 2007. The Uvs Lake version appeared in Kuwait in February 2007. Although it was clade 2.2.3, like the Indian isolate, it had evolved markedly, as represented on the tree by isolates from Kuwait and Krasnodar. This evolved version of clade 2.2.3 has now spread throughout Europe, as seen in the recently released sequences from the Czech Republic, Germany, Romania, Saudi Arabia, and Ukraine (as well as reported but unpublished isolates from England, Poland and northeastern Germany). All of these have significantly evolved away from the Indian isolate awaiting approval.
Similarly, the evolution in Egypt has been extensive. Consequently, the tree has five 2007 isolates from Egypt, representing three major branches. One of these five has been selected for vaccine development. The selected isolate is from a mild case in Aswan that had the Mongolian cleavage site. However, the diversity in Egypt a year ago is under-represented on the tree. There are no isolates representing the Gharbiya cluster, which has two receptor binding domain changes, including M230I, which was only detected in fatal cases. Similarly, other sequences from a year ago, including otehr Gharbiya sequences are also not represented in the published tree.
Thus, the current report has not kept pace with samples collected a year ago. The recent samples from Egypt are significantly more evolved than the sequences represented in the published tree or the early 2007 sequences not represented.
The evolution of H5N1 in India and Bangladesh is not public, and may represent significant additional heterogeneity.
The phylogenetic tree in the latest report is clearly dated, and the vaccine targets are lagging further because of production and approval issues.
H5N1 obviously knows how to evolve more rapidly than the WHO vaccine selection committee.
Media Links
Recombinomics Presentations
Recombinomics Publications
Recombinomics Paper at Nature Precedings