Growing H5N8 Genetic Diversity And Pathogenicity

[niman]

The fall H5N8 outbreak continues to grow, and recently released sequences reveal a growing complexity.  H5N8 was initially reported in India, followed by multiple outbreaks in Europe (initial cases were largely in symptomatic or dead wild birds).  Prior to these outbreaks H5N8 had been reported at Uvs Lake in Tyva, Russia.

Initial H5 and H8 sequences in Germany and Denmark matched the Uvs Lake sequence.  However, sequences from Qinghai Lake in China were subsequently released.  The large number of sequences suggested the outbreak was large (no OIE report was filed) and collection dates preceded those from Uvs Lake.  Most sequences were from dead Bar-Headed geese, followed by Brown-Headed gulls, much like the H5N1 outbreak at Qinghai Lake in the spring of 2005.

In 2005, the characterization and sequencing of the H5N1 at Qinghai Lake identified a new sub-clade (clade 2.2) which raised concerns that its presence in long range migratory birds (most of the deaths were also Bar-Headed geese) would lead to geographical expansion because of migration to southern Russia and northern Mongolia where multiple flyways overlap in summer breeding locations.  The concerns were realized when Russia, Kazakhstan, and Mongolia reported H5N1 outbreaks for the first time and all sequences were clade 2.2 (Qinghai strain).

In the fall the H5N1 dramatically increased its geographic reach. Although reported cases in Europe were limited to four countries, reports increased dramatically in early 2006, and included South Asia as well as Europe, the Middle East, and Northern and Western Africa.  The patterns were reported on all smaller scale in subsequent seasons, initially by clade 2.2 followed by Fujian clades 2.3.2 and then 2.3.4.  The 2.3.4 (clade 2.3.4.4) reasserted in China to generate a number of H5 serotpes.  H5N8 casued major outbreaks in the spring of 2014 in South Korea poultry and Japan wild birds.  In the late summer it was in Russia and then caused a limited number of outbreaks in Europe in the fall.  The same lineage appeared in Japan, where two additional H5N8 lineages were circulting.  One of these other lineages (identified in a crane) gave rise to outbreaks in Taiwan and North America that included devastation in poultry populations.

In Taiwan there were three different serotypes.  All had the H5 from the crane, but had replaced some of the remaining gene segments with genes from low path serotypes circulating in Asia.  In addition to H5N8, Taiwan reported H5N2 and H5N3.  These outbreaks decimated duck and goose farms primarily.

In North America, three different serotypes were identified.  For H5N8 all eight gene segments matched the crane from Japan, but the other two seroptypes (H5N1 and H5N2) had a low path N gene from North American wild birds as well as additional internal genes, so each novel serotype had Asian gene segments from the crane, including H5, mixed with North American sequences for some fo the internal genes.  Most of the infections in the fall in the US were wild birds and a limited number of backyard farms,  In Canada large commercial farms in British Columbia were infected with H5N2.

 

However, in the spring H5N2 cause major damage to commercial farms.  In Minnesota, the largest turkey producer in the US, the turkey farms were hit hard, while in Iowa, the largest egg producing state in the US, the layer farms were hit hard.

In spite of increased surveillance since these outbreaks, newly reported infections have been limited to a small handful of wild birds (the most recent in the summer near Fairbanks Alaska).

Similarly, Europe had little activity in the 2005/2006 season.  However, the current outbreak is the largest to date, and limited sequence data indicates the genetic variation is very volatile.  Initial H5 and N8 sequences (from Germany and Denmark) matched the lineage reported for the Qinghai and Uvs Lake outbreaks.  However, the first full 2016 European sequence was generated in Poland, and three of the internal genes (PB2, PA, NP) were from low path wild birds.  This constellation was seen in subsequent sequences from Sweden and England.  However, sequences from Croatia had a different constellation, which did not include a change in the PA.  Moreover, the PB2 and NP genes were from different lineages and the genes that matched the Qinghai/Uvs Lake lineage were slight different from the other constellation.

Similarly, two full sequences from India were released, and they had different constellations relative to each other and the European sequences.  Both sequences from India had a wild bird NP and one of the sequences also had a wild bird PA.  However, as was seen in the European sequences, the wild bird sequences were from different lineages, and the segments that matched Qinghai/Uvs Lake had drifted and could be easily distinguished from the sequences in the other reassortants.

Recently released sequences from France and Poland have provided additional clarity and suggest significant genetic variation.  France released two full sequences which matched the pror two constellations.  The northern constellation match the prior sequences from Poland, Sweden, and England, and provided additional support for lineage differences between the two constellation indicating prior H5 and N8 sequences from Germany and Denmark were from this dominant constellation.  The sequence from southwest France matched the Croatia sequence, which raised the possibility that this constellation was more common in southern Europe.  Similarly, the relatedness of the H5 adN8 sequences indicated two other sets of H5 and N8 sequences from Croatia were from the same sub-clade.

Poland released 7 additional sets of H5 and N8 sequences, none of which matched the dominant constellation in Europe.  The three most southern sets were closely related to the Croatia sequences, further supporting the higher prevalence of this sub-clade in southern Europe.  These results also raised concerns that this constellation was particularly damaging to poultry.  France is intensifying culling as the outbreak at predominantly duck farms in southwest France continues to grow at an accelerating rate.  Similarly the sequences in Croatia and Poland flank a massive duck farm outbreak in southern Hungary (and are not far from a growing duck farm outbreak in Bulgaria).  The other four sequences fall into two sets.  Both are more closely related to the Croatia sequences but are distinct.  One set (H5 sequences only) appear as precursors to the Croatia constellation, while the second set, which includes N8 have N8 sequences that are divergent from other  N8s in European constellations and may represent another constellation.

In addition to the demonstrated two constellations in Europe and two more in India, three countries  (Netherlands, Montenegro, Italy) have reported a new serotype (H5N5) which likely have additional reassortment in a subset of the internal genes.  The three locations for a straight line which crosses the location of a novel serotype in Germany (H5 positive and N8 negative), which may also be H5N5..

Thus, the diversity described above may only reflect a subset of the H5 constellations spreading in South Asia, Europe, Middle East, and Africa.  Release of additional full sequences would be useful.

[niman]

Today's audio on sequences

http://recombinomics.co/thedrnimanshow/2017/01/010417.mp3