Significant Mutations in H5N1 Bird Flu In North Sumatra
Recombinomics Commentary
June 7, 2006
Full genetic sequencing of two viruses isolated from cases in this cluster has been completed by WHO H5 reference laboratories in Hong Kong and the USA. Sequencing of all eight gene segments found no evidence of genetic reassortment with human or pig influenza viruses and no evidence of significant mutations. The viruses showed no mutations associated with resistance to the neuraminidase inhibitors, including oseltamivir (Tamiflu).
The human viruses from this cluster are genetically similar to viruses isolated from poultry in North Sumatra during a previous outbreak.
The above description of the H5N1 bird flu isolated from patients in the north Sumatra has raised more questions that it has answered. The unanswered questions have led to a good deal of confusion because the cluster is the largest and most deadly recorded for Indonesia, yet the WHO description suggests that there have been no genetic changes that are "significant". The lack of "significant" mutations has also raised questions about the increased level of H5N1 in the patient's nose and throat, which is a property associated with more efficient transmission.
WHO and consultants take a very narrow and dogmatic view of influenza evolution. This view holds that the genes change gradually via random mutation. The gradual change is called genetic drift. Alternatively, since the genetic information of influenza is on eight gene segments the genetic composition can change by swapping entire genes resulting in genetic shifts. This concern is focused on one or more avian genes being replaced with a human or swine mammalian gene
Therefore, the changes considered significant are such a swap in genes called reassortment, or random mutations that change the receptor binding domain, specifically, Q226L and G228S. These two changes allow the virus to more efficiently bind to receptors in the upper respiratory tract of humans.
Therefore, the characterization published by WHO simply indicates that all eight genes from H5N1 are avian, and that position 226 in the HA sequences still has the avian amino acid, Q, and position 228 still has the avian amino acid G. Neither position has change to the amino acid that would increase affinity for human receptors in the upper respiratory tract.
The above descript also indicates that the NA of the isolates have also remained wild type and are therefore sensitive to neuraminidase inhibitors, such as Tamiflu.
However, the report says nothing about other changes which some, including WHO in the past, place in the "significant category.
One change that could account for the higher levels in the nose and throat as well as increased transmission is the PB2 polymorphism E627K. as discussed by Declan Butler in his Nature news story and his blog. E627K is found in all human isolates and allows the H5N1 to grow more efficiently at lower temperatures, 33 C. Since the body temperature of birds is 41 C and of humans is 37C, the lower temperature is closer to human body temperature and even close to the temperature in the nose and throat, especially in the winter, which could explain the seasonality of seasonal flu in the areas away fro the equator.
E627K is also associated with increased virulence in mammals and was first identified in H5N1 in some patients in 1997 in Hong Kong. It was also found in patients in Vietnam and Thailand In 2004 and 2005. Almost all patient infected with H5N1 with E627K died. Similarly, wild and domestic cats and a domestic dog were also fatally infected with H5N1 with PB2 E627K.
The PB2 E627K had only been found in H5N1 from mammals prior to the Qinghai Lake outbreak. It become fixed in the Qinghai strain and WHO commented on the presence of PB2 E627K. However, WHO failed to indicate that the second human isolate from Indonesia had E627K when it described the isolate in its WHO update. That case was also fatal and had the wild type cleavages site, as did all of the above human and animal infections. The realtionship between the first human isolated from Indonesia, A/Indonesia/5/2005 and the second isolate, A/Indonesia/6/2005, can be seen in the phylogenetic tree released by the Chinese Minister of Health.
The increased levels in the nose and throat in patients in north Sumatra may signal the presence of E627K. The H5N1 was easily transmitted to family members and 7 of the 8 infected family members died. This mechanism of transmission may have contributed to teh large cluster in Turkey as described by Don McNeil in the New York Times. The index case has H5N1 with an altered receptor binding domain, S227N, in addition to E627K.
Similarly, the WHO update failed to indicate that the cleavage site was the wild type, RERRRKKR, associated with the high fatality rate when E627K was present. Thus, this combination could account for the increased transmission since the cleavage site determines which proteases can cleave HA, which generates tissue tropism.
Thus, WHO's comments that there were no significant mutations did not address two areas that outs would have considered significant. This selective release of information is also applied to sensitivity to anti-virals. When isolates from Vietnam and Thailand were found to be amantadine resistant, the resistance was reported. The lack of this resistance was noted in the description of the first human H5N1 isolate from Indonesia. However, the description of the isolates in north Sumatra failed to indicate that they were amantadine resistant. Only the neuraminidase sensitivity was mentioned.
Thus, the WHO update on "significant" mutations was very narrow. The update did not disclose information on the HA cleavage site, amantadine, resistance, or E627K. This withholding of information is yet another reason for immediate release of the sequences sequestered in the private WHO database.
Media Source Link
Map