|Home||Founder||What's New||In The News||Contact Us|
|Paradigm Shift Intervention Monitoring||Commentary
H5N1 Acquisitions in Egypt Challenge Influenza Dogma
April 23, 2007
The geographical expansion of the Qinghai strain (clade 2.2) of H5N1 offers a unique opportunity to monitor the genetic evolution of the strain. Clade 2.2 was first reported at Qinghai Lake in May, 2005. Initially, 178 bar headed geese were found dead at the nature reserve in central China. The initial OIE report described 519 dead waterfowl. Five species were identified, and the presence of H5N1 in the long range migratory birds was cause for concern. Sequences of the H5N1 indicated the clade was readily distinguished from earlier H5N1 isolates in Asia. In addition to a novel HA cleavage site, GERRRKKR, the isolates also had PB2 E627K, a change that was linked to human seasonal flu. All H1N1, H2N2, and H3N2 human isolates dating back to 1918 had E627K. In contrast, no prior H5N1 from birds had E627K.
In July and August, H5N1 outbreaks were reported in Russia, Kazakhstan, and Mongolia. Like the outbreak at Qinghai Lake, migratory waterfowl at Chany Lake in Russia and Erhel Lake in Mongolia tested positive for H5N1, and the sequences indicated the H5N1 which was reported for the first time in the three northern countries, were the Qinghai strain.
The presence of H5N1 in the long range waterfowl, which summer in the northern nature reserves signaled subsequent migration to the east in Asia, as well as west to Europe, the Middle East, and Africa. This long range migration offered a unique opportunity to monitor H5N1 evolution, as H5N1 expanded its geographical reach. In 2006, H5N1 spread into over forty countries which had not previously reported H5N1. In all cases, the H5N1 reported was exclusively the Qinghai strain. Sequence analysis indicated the emerging H5N1 had a series of regional markers, which distinguished various sub-groups within the 2.2 sub-clade.
Sequences from Egypt provided a large database of sequences from birds and patients in the 2006 season, as well as the current 2006/2007 season. The more recent isolates were more genetically complex and included a series of polymorphisms which had been previously identified in recent H5N1 isolates in Asia. The Egyptian markers in the isolates from early 2006 were in the isolates from the current season, and these additional polymorphisms were overlaid into the Qinghai Egyptian genetic background. These acquisitions were consistent with acquisition via recombination.
The acquisitions were not easily explained by influenza genetics dogma, which maintains that sequences of influenza genes change via random mutations which are selected during escape from immunological recognition, or adaptation to a new host or new region. Dogma also maintained that the polymerases encoded by the influenza genome lacked a proof reading function, which led to rapid evolution.
However, the acquisitions of previously reported polymorphisms suggested that the annual changes were due to recombination, and not de novo “mutations”. The acquisitions in Egypt were found in H5N1 isolates that were easily linked via migratory bird flyways.
Recent simultaneous acquisition of the same polymorphism, NA G743A, created additional challenges for explanations based on random mutations. The acquisitions were on multiple genetic backgrounds in Egypt and Moscow. Earlier sequences had changes that created a number of new branches on the phylogenetic trees, and G743A was acquired simultaneously on the more recent isolates.
These changes can be seen on the NA phylogram. The top portion of the phylogram is on the left side of the figure. At the bottom are a series of isolates from Germany. These isolates formed a distinct branch on the tree and all isolates had G743A. However, this change was not found in other Qinghai 2006 isolates, including two other branches, which were largely populated by isolates from Germany. Thus, the isolates from Germany formed three branches, but G743A was only found in one.
In Egypt, H5N1 in patients and poultry was reported between February and April of 2006. Reported H5N1 infections marked the re-emergence of H5N1 in Egypt in the fall. These isolates had the regional markers that were present in the isolates from early 2006, but the more recent isolates were more genetically complex. This complexity was due to a series of polymorphisms that were previously reported in H5N1 isolates in Asia. The sequences generated by US NAMRU-3 provided a monitor of H5N1 evolution in Egypt.
One example of this added complexity was in isolates from the Gharbiya cluster. The figure has four isolates from the cluster. These isolates were collected 48 hours apart from two of the three cluster members. Although the sequences from the two patients were distinguishable, the sequences were similar and had changes in or near the HA receptor binding domain (V223I and M230I) and had the oseltamivir (Tamiflu) resistance marker, N294S. This marker was present in all four isolates, even though two were collected prior to treatment. Like all of the sequences from Egypt in late 2006 and early 2007, G743A was not detected.
G743A was first detected in three chickens from Gharbiya, collected on February 15, 2007. One of the isolates, A/chicken/Egypt/1892N3-HK49/2007 (see red arrow at the top of tree on right side of figure), had the receptor binding domain changes, as well as a number of additional polymorphisms found in the Gharbiya cluster. However, G743A was appended onto this genetic background.
The two other chicken isolates, (A/chicken/Egypt/1890N3-HK45/2007 and A/chicken/Egypt/1891N2-CLEVB/2007, were on another branch with earlier bird isolates from the Nile Delta. Although these earlier isolates also did not have G743A, the two February 15 isolates had acquired G743A.
At this time, H5N1 outbreaks were reported in the suburbs of Moscow. As seen in the lower portion of the right hand side of the figure, the 2007 Moscow isolate, A/chicken/Moscow/2/2007, was similar to Azerbaijan isolates from the prior year and the chicken isolate also had G743A (as did a second closely related isolate from Moscow, A/chicken/Russia_Moscow oblast_Odintsovo/1/2007, which is not represented in the figure).
In addition, 2007 isolates from central Egypt formed another branch on the tree. These sequences had a 3 BP deletion that was identical to chicken isolates from Hunan Province in China. Although these sequences were distinct from clade 2.2, the 3 BP deletion was of the same three nucleotides, proving additional compelling data for acquisition via recombination. The isolates from early 2007 in Egypt did not have G743A, but recent isolates from two siblings from Qena, A/Egypt/2621/2007 and A/Egypt/2629/2007, had the acquisition.
Another recent isolate, A/Egypt/2630/2007 from a patient in Sohag, is on yet another branch, and it too has G743A (see red arrows).
The grouping of these isolates onto separate branches can also be seen in the HA phylogram. The relationships described above on the NA tree are also found on the HA tree, further increasing the genetic differences between the various isolates from Egypt and Russia.
The simultaneous acquisition of G743A onto these diverse genetic backgrounds is most easily explained by recombination with a common source. The acquisition is very difficult to explain by random mutations.
However, random mutations and selection are the cornerstone of current influenza dogma on the mechanism of genetic drift.