Continued Hoarding Of H5N1 Sequences
Recombinomics Commentary
November 2, 2006
"There has no marked change in the biological characteristics of the organisms."
China, home to the world's biggest poultry population, has been at the centre of the fight against bird flu, which scientists fear could mutate into a form that can pass easily between people, potentially leading to a pandemic.
It has battled dozens of outbreaks in birds, and seen at least 21 human cases, including 14 deaths.
But the World Health Organisation says its understanding of the virus and how it might be changing is being hampered by the fact China has not shared animal virus samples since 2004.
"It's our understanding that there have been certainly changes in the virus and continual evolution in the virus since 2004 and the viruses that we requested from 2005 have still not been shared with WHO," said Julie Hall, the WHO's bird flu coordinator in Beijing.
The point was not whether there were major differences in virus strains, she said: it was about being able to keep abreast of changes, however minute, to better understand how the virus is developing.
"Whether it be dramatic or significant, it's about that regular understanding so that we can see that we're keeping on top of this," Hall said.
The above acknowledgment by WHO of the potential importance of minute changes in H5N1 is a welcome change. In the past, WHO updates have focused on the lack of major changes, such as the acquisition of human genes via reassortment. However, there has been no data to support such an acquisition by H5N1 and the H1N1 1918 pandemic also did not involve reassortment. Although the basic tenets of influenza genetics cite reassortment as the source of genetic shift and random mutations as the source of genetic drift, the H5N1 clearly demonstrates evolution by recombination.
Researchers in China have acknowledged the recombination in H5N1 in their earlier submissions to Genbank entitled, “A cohort of AIV H5N1 subtypes isolated from wild aquatic birds and domestic poultry revealed rapid transmission, frequent reassortment, and identifiable recombination events”. Sequences released last month confirmed the recombination, but researchers from St Jude and Hong Kong University have failed to acknowledge such changes, even though recombination was obvious in sequences from Hong Kong in 2002 in PB2, PB1, PA, and NP. The sequences released this week also had evidence of recombination with the Qinghai strain, as well as obvious recombination in genes that were closely related to sequences released by researchers at the Beijing Genome Institute.
The recombination provides a roadmap of prior and future interactions resulting from dual infections and the exchange of genetic information, which is the main driving force behind H5N1 evolution. Thus, a full and current database is essential for monitoring the evolution of H5N1 as well as the development of vaccines against emerging genomes.
The complaints against China are justified. Although the full sequences released last month provide a detailed history of H5N1 evolution between 1997 and 2004, the recent releases do not contain any Fujian sequences or any sequences from 2005 or 2006. Human sequences from fatal cases in 2005 and 2006 have been released, but most only contain the HA sequence and there have been more releases for the more recent cases, although a Ministry of Health report from the beginning of this year describes additional cases in 2006, which are clearly also due to infections by the Fujian strain.
In addition, presentations on H5N1 in Qinghai Province indicate many additional Qinghai isolates have been sequenced, but not released. The MOH report also suggests additional Qinghai isolates have been obtained for eastern China.
The PNAS report published this week also had the sequences of a Qinghai isolate from Shantou. This is the first public sequence of the Qinghai strain from eastern China. The sequence had the characteristic Qinghai HA cleavage site as well as PB2 E627K and was closely related to Qinghai isolates from China, Russia, Mongolia, India, Afghanistan, Europe, the Middle East, and Africa, but unlike the isolates from outside of China, the unique polymorphisms were found in H5N1 isolates in China, again highlighting the role of recombination in H5N1 evolution via acquisition of single nucleotide polymorphisms.
The PNAS report included HA sequences from 404 H5N1 isolates from China, which greatly expanded the HA database, which included 1032 sequences at Los Alamos prior to the PNAS report. Moreover, several hundred of the 1032 sequences were added in the preceding weeks from sequences released by Indonesia, Qinghai isolates sequenced under the NIAID flu blueprint program, and the sequence released by the Beijing Genome Institute. However, many of the 404 HA sequences were partial sequences, and only 156 PB2 sequences were released. Moreover, no sequences from the other 6 gene segments were released, even though the MP gene sequences were discussed in the paper. In addition, a high proportion of earlier H5N1 sequences, are only partial sequences, including the larger PB2, PB1, PA, and NP genes, which have obvious examples of recombination, which was present in the 156 sequence that were released.
The other major hoarder of sequences is the WHO affiliated Weybridge labs. Phylogenetic trees of approximately 80 Qinghai sequences from Europe were presented months ago from samples collected in late 2005 / 2006. These sequences have also been withheld, in spite of comments indicating that the sequences will be released. Thus far only one bird sequence (the 8 gene segments from a turkey in Turkey collected over a year ago) and five human sequences (four from Turkey and one from Azerbaijan) have been released.
Moreover, only 4 gene segments of one H5N2 isolate in Canada has been released, although H5 was found throughout Canada over a year ago in August 2005 isolates and no 2006 sequences from H5N1 in Canada or the United States has been released.
The hoarding of these sequences by the major sequencers worldwide is cause for concern. The sequences provide critical information on the interactions and transport of H5N1 as well as the genetic diversity, which drives the evolution. The release of full sequences from current and past H5N1 isolates, which can be done at no charge via the NIAID sequencing project, is long overdue.
Media Sources
Phylogenetic Trees