Is PB2 E627K in Sumatra H5N1 Sequences Significant?
Recombinomics Commentary
May 31, 2006
What caused the suspected human-to-human transmission at Kubu Sembelang is still a mystery. Nature has learned that the cases differed from past Indonesia cases, in that they had much higher viral loads in the throat and nose. Human-to-human transmission is more likely through droplets coughed from the nose and throat than from infections further down the respiratory tract.
Mutations in cases in Turkey earlier this year showed a substitution of glutamic acid with lycine at position 627 in the PB2 component of the polymerase gene. The mutation is thought to allow the virus to survive in the cooler nasal regions. This mutation has not been publicly reported in Indonesia previously, but Nature has learned that it occurred in at least one case in August 2005.
The above comments raise additional questions about the acquisition of PB2 E627K in the Sumatra cluster. The only confirmed case in Indonesia in August 2005 was the second confirmed case in Indonesia. This case had a wild type H5N1 cleavage site, RERRRKKR, and was reportedly linked to fertilizer. However, the only bird H5N1 sequences with PB2 E627K is the Qinghai strain and the description of the isolate from the second confirmed case indicated it was a clade 2 sequence, like poultry sequences in Indonesia. Therefore, none of the human sequences match a published bird sequence in the area, highlighting the need for more H5N1 sequences from West Java, and the need to release the existing human sequences.
The acquisition of PB2 E627K could have been due to random mutation within the patient. However, E627K could also have been acquired via recombination, since all mammalian sero-types isolated from humans have E627K. Similarly, all Qinghai strains of H5N1 also have E627K. so recombination could have happened between two avian isolates. Although E627K could have also been acquired via reassortment involving avian genes, such reassortment would have probably mentioned in the WHO update.
In the Qinghai strain, the wild type cleavage site is GERRRKKR, which may not be as virulent as RERRRKKR. Earlier human isolates from Hong Kong, Vietnam, and Thailand that combined the wild type H5N1 cleavage site with PB2 E627K were almost universally fatal. Similarly the combination in mammals, such as the tigers in the zoo in Thailand, a domestic cat or domestic dog also led to fatalities. E627K also increased virulence in laboratory mice.
The infections in the tigers in Thailand also lead to efficient tiger to tiger transmission. All 147 of the exposed tigers either died or were euthanized, even though culling and Tamiflu was used to try to control the outbreak.
Thus, the combination of a wild type HA cleavage site, amantadine resistance, and PB2 E627K would be a combination of changes that would be cause for concern. Since none of the above would necessarily be considered "significant mutations", they would not be covered in the WHO description of the sequences from the Sumantra cluster. Moreover, Hong Kong's decline to comment on specific questions about these mutations increases the need to make these sequences public immediately.
Media Link
Map