Paradigm Shift Intervention Monitoring
twitter
Commentary
Silent Spread of H1N1 RBD Change L194I
Recombinomics Commentary 23:45
December 30, 2013
Media attention on eastern Texas began with a report that a flu-like illness in Conroe, Texas had killed four of eight cases, while the other four were hospitalized and two were near death. The cases were designated as a mystery disease cluster because initial testing was negative for influenza. However, retesting of the four surviving identified H1N1 in two and the retest results on the other six has not been released. However, six H1N1 deaths were confirmed in the Houston area and now media reports have increased the death toll to 13.
Moreover, severe and fatal cases were reported in Austin, Texas, where two pregnant cases had died and five additional adults were on ECMO machines. The maxing out of ECMO machines also signaled more severe cases, and media reports cite five more deaths in the Dallas area. Almost all of these cases have been young or middle age adults. A pediatric death was also reported in Houston, while at least one child was on an ECMO machine in Austin.
These severe and fatal cases had much in common with an outbreak in Ukraine in the fall of 2009. The high death toll in young and middle aged adults led to an investigation by WHO (primarily through Mill Hill and CDC regional centers). Sequence data was delayed and changes in the receptor binding domain were predicted (D225G and D225N). These two changes had been seen in severe and fatal cases in Brazil and China, and D225G was also identified in the lungs of fatal H1N1 cases during the 1918 pandemic. Moreover, it was known that D225G target 2,3 gal receptors which were in avian influenza (H5N1 and H7N9) as well as human lung (and Ukraine cases were described as suffering from total destruction of lung).
Although, D225G and D225N was identified in the fatal cases, WHO claimed that no ‘significant” genetic changes were found, which was not supported by the initial sequences (where D225G and D225N were limited to fatal cases). Others also reported a concentration of D225G in severe and fatal cases, but WHO published a report at the end of 2009 claiming that D225G could be found in milder cases and testing of original samples lacked D225G, suggesting D225G was selected during lab isolation of virus from the severe and fatal cases, especially if isolated in eggs, which had 2,3 gal receptors and would selected for D225G formed by “random mutations.”
Days after the WHO report, the selection of random mutations was thoroughly refuted by the detection of D225G and D225N by direct sequencing of autopsy lungs from Ukraine. The samples had D225G, D225N, or both and most samples also have wild type, raising concerns that isolation of virus on mammalian cells would select against D225G and only wild type would be detected.
These concerns were confirmed when the CDC published sequences generated by the same sample was used to isolate H1N1 using mammalian cells or eggs. Wild type H1N1 was found using mammalian cells, while D225G was found in virus grown in eggs. The CDC has subsequently published two sets of sequences from the same patients for a small subset of samples, and these data raise concerns that changes that target 2.3 gal, which are found in human lungs, are significantly under-represented in the sequence database.
Analysis of recent sequences from cases in the United States has identified another receptor binding domain change that is present in egg isolates. This change, L194I, is known to increase affinity for 2,3 gal receptors, which are abundant in human lung. Recent sequences from Florida and Massachusetts cases (A/Florida/43/2013 from a 42M collected October 27 and A/Massachusetts/15/2013 from a 34F collected November 13) were wild type in FL or MA isolates from mammalian cells, but had L104I in FL or MA isolates from growth in eggs. Similar results were seen in cases in Connecticut (Connecticut/02/2013) and Nevada (A/Nevada/06/2013) cases identified in July. CT and NV mammalian isolates were wild type, while CT and NV egg isolates had L194I.
Most sequences however, are from mammalian isolates, and the 13 mammalian isolates listed below have a H1 sequence (1701 BP) which exactly matches A/Massachusetts/15/2013.
A/Alabama/08/2013
A/Florida/38/2013
A/Florida/42/2013
A/Louisiana/08/2013
A/Louisiana/10/2013
A/Louisiana/11/2013
A/Louisiana/12/2013
A/Louisiana/13/2013
A/Maryland/08/2013
A/Michigan/08/2013
A/Mississippi/09/2013
A/Mississippi/11/2013
A/Norway/2609/2013
Recombinomics
Presentations
Recombinomics
Publications
Recombinomics
Paper
at Nature Precedings