Clade 2.3.2.1

Recombinomics Commentary 19:00
March 5, 2012

Detection of three new human infections with the deadly H5NI strain of bird flu in a week has set alarm bells ringing as scientists have found evidence of the virus in the live-bird market in crowded Dhaka.

The above comments describe three confirmed H5N1 cases in wet market workers in Dhaka, Bangladesh.  The first case has been WHO confirmed and has recovered, but the three 2012 cases has doubled the number of H5N1 confirmed cases in Bangladesh.  This significant up tick is almost certainly due to clade 2.3.2.1 which has now expanded its geographic spread and was confirmed in crows in India earlier this season.  New sub-clade migrate into India and Bangladesh each season, which are closely related to each other.

The first human clade 2.3.2 was described in media reports in the spring of 2008 when a culler (soldier) developed symptoms and was H5 PCR confirmed.  However, South Korea denied the culler was infected with clade 2.3.2 because prior human infections in China were clade 2.3.4.  Moreover, South Korea used the failure to isolate the H5 virus as an excuse for not filing a report.  However, the sequence had V223I and M230I, which were in the Gharbya cluster and fatal clade 2.3.2.1 cases were subsequently reported in China, including the recent case from Shenzhen who had no reported poultry contact (but had wild bird exposure).

The appearance of clade 2.3.2 in south Asia raised concerns of additional human cases, and the three confirmed cases in Bangladesh are likely confirmation of the realization of such concerns.

Recombinomics Commentary 16:45
January 23, 2012

"A man died in southern China on Sunday from the H5N1 bird flu virus, the Health Ministry reported. It was China’s second such death in less than a month.

On Sunday, Chinese censors generally blocked Internet users from reading reports of the latest death."

The above comments describe censorship of news on the H5N1 case (39M) from Guiyang, Guizhou who had no reported contact with poultry, suggesting that the case would be linked to clade 2.3.2.1 as seen in the recent case from Shenzhen, who also had no reported contact with poultry.  The sequences from the Shenzhen case, A/Guangdong-Shenzhen/1/2011, were released less than a week after collection, and the HA sequence had a number of receptor binding domain changes.  The two changes reported in the Gharbia cluster, V223I and M230I, were present, as expected.  However, the sequence also had S227R as well asQ196K, raising concerns that the combination of changes could be approaching the five changes reported in the paper censored by Science. 

That paper described five changes on two genes and four of the five changes have already been found in published sequences.  One of those changes is PB1 E627K, which was not in the Shenzhen clade 2.3.2.1, but was present in the most recent H5N1 sequence from cases in Cambodia, where all 8 of the cases in 2011 have died, as has the only case in 2012.  That case also had S227N, which is likely to be one of the HA changes in the aerosol version of H5N1 that transmits in ferrets.

The co-circulation of the above changes raises concerns that new combinations may be generated by recombination, leading to enhanced transmission, which may or may not match the changes described in the censored paper at Science.

It is unclear if the censorship of news on the most recent case will impact release of sequences from the Guizhou case.  Similarly, Indonesia quickly released sequences from the Bali cluster, which also had receptor binding domain changes.  That cluster was followed by a cluster from North Jakarta. And sequences from that cluster have not been released, raising concerns that the censorship of the papers and nature and Science will have a chilling effect on transparency and lead to more hidden sequences, as it becomes increasingly clear that H5N1 currently in circulation is very close to efficient transmission.

The natural versions pose a far greater hazard than some unnamed rogue state or terrorist who are said to benefit from the release of the five changes describe in the Science paper.  Since it is already known that 4 of 5 changes are already in circulation in a published sequence, the creation of a transmissible H5N1 is a trivial task, although the value of such bioweapon is far from clear, since influenza cannot be controlled, and effects on rogue nations would be significant.

Thus, the censorship by Nature and Science continues to endanger the world’s health, and the lack of understanding of influenza evolution by the NSABB is glaring.  Therefore replacement of the current board with one more knowledgeable about influenza evolution should be actively pursued.

Recombinomics Commentary 23:45
January 18, 2012

"It is for the first time that it has been found in the country, Mishra said. “We have compared it to the strains from Vietnam, Indonesia and other places in Asia and found similarities with the Vietnam strain,” added Mishra.

The NIV had received samples that included four crows from among those that had died in several areas of Jharkhand, including Jamshedpur, Bokaro and Hazaribagh. The process of characterisation of the virus is underway and it is a different strain of the avian influenza virus, NIV scientists said.

The strain belongs to Clade 2.3.2.1 while the H5N1 strain that was reported in the country in 2006 belonged to Clade 2.2."

The above comments confirm that clade 2.3.2.1 has migrated into south Asia, as expected.  In the past, clade 2.3.2.1 was largely confined to southern China and southeast Asia.  Hong Kong surveillance would identify the sub-clade in dead wild birds identified each year, generally between December and February.  However, in the spring of 2008 there was a major expansion of this sub-clade to wild birds in northern Japan as well as poultry in South Korea and poultry and wild birds in southeastern Russia.  These confirmed cases suggested that the Fujian clade 2.3.2 would be competing with Qinghai clade 2.2 due to infections in wild birds that share large flyways that overlap in Mongolia and Russia.

In 2008 one culler was infected and H5 PCR confirmed.  However, virus was not isolated and the human case was denied.  The denial included comments that human cases had involved the Fujian sub-clade 2.3.4 and not 2.3.2 even though the internal genes of the 2.3.2 were 2.3.4.  Fatal human cases involving 2.3.2 were subsequently confirmed in China and Hong Kong.  These cases were of concern because two receptor binding domain changes V223I and M230I were fixed in clade 2.3.2.  Many of the recent wild bird sequences had an additional receptor binding domain change S227R, and the recent fatal case in Shenzhen had the three changes above as well as Q196K raising concerns that many of the receptor binding domain changes use to create a more transmissible H5N1 in ferrets were recombining in wild birds and evolving in more transmissible H5N1 without the aid of scientists or terrorists.

It is likely that the recent H5N1 outbreaks in northeastern India as well as Bangladesh also involved clade 2.3.2.1 containing two or more of the above changes, increasing the likelihood of human infections.  Recently released sequences from 2011 isolates from Japan and South Korea share polymorphisms with clade 2.2.1 isolates in Egypt, including sub-clade 2.2.1 G, (see list here and here) which has PB2 E627K as well as sequences from seasonal and pandemic H1N1 (H1N1pdm09).  Moreover, these sequences include identities with H3N2v sequences in United States cases.

Full sequences from the H5N1 in India and Bangladesh would be useful, in addition to full sequences from human cases in Egypt.

Similarly, release of the receptor binding domain changes described in the censored papers at nature and Science is overdue.  The withholding of this information by Nature and Science continues to be hazardous to the world’s health.

Recombinomics Commentary 16:30
January 4, 2012

" In order to investigate the escape process and to enable predictions of escape, we serially passaged influenza A H5N1 virus in vitro 100 times under immune pressure. The generated escape viruses were characterized phenotypically and in detail by full genome deep sequencing. Mutations already found in natural isolates were detected, evidencing the in vivo relevance of the in vitro induced amino acid substitutions".
 
The above comments are from an upcoming publication, “Highly Pathogenic Avian Influenza Subtype H5N1 escaping neutralization: more than HA variation” which describe the identification of escape mutants which are present as a minor species in the original sample.  P198S was identified in a swan clade 2.2 (Qinghai strain) isolate from Germany, A/cygnus cygnus/Germany/R65/2006, using this approach, and it was also in a clade 2.2 swan isolate from Russia, A/Cygnus olor/Caspian Sea/2006, confirming it was in circulation in clade 2.2 in Europe in 2006.

P198S was in the recently released sequence from the fatal case in Shenzhen, China, A/Guangdong-Shenzhen/1/2011, which had a large number of HA changes, including the adjacent change Q196K, which is in clade 2.2 isolates in Egypt, Germany, and Kuwait (see list here).

The Shenzhen sequence is clade 2.3.2.1 (Fujian strain) also has another receptor binding domain change, S227R, but was generated via novel coding that was distinct from S227R in clade 2.3.2.1 in Hokkaido and Fukushima.  The coding found in the Shenzhen sequence is found in H5 sequences from wild birds.

H5N1 is frequently found in wild birds in Hong Kong at this time of year, due to migration patterns.  In 2008  clade 2.3.2 was found Japan, South Korea, and Primorsky (southeastern Russia) in poultry and wild birds, which create expansion concerns.  Clade 2.3.2 was subsequently identified in Mongolia, followed by Romania, matching the wild bird path that brought clade 2.2 to Europe, the Middle east, and Africa.

The movement of clade 2.3 into the clade 2.2 flyway created an environment for co-infections and recombination, leading to the acquisition of clade 2.2 polymorphisms in poultry and wild birds in Europe, the Middle East, and Africa.

These interactions lead to rapid evolution as seen in the large number of HA changes in the fatal case in Shenzhen, China.

Recombinomics Commentary 21:20
January 1, 2012 CDC experts found that the patients infected with the virus is 2.3.2.1 sublines highly pathogenic avian influenza virus to humans is highly pathogenic…..

the results show that virus is adamantane class of drugs are more sensitive but resistance to Tamiflu, the role of drugs still under study
 
The above translation indicates the H5N1 from the fatal case In Shenzhen, China is clade 2.3.2.1, which is common in wild birds.  Between December and February, Hong Kong typically finds clade 2.3.2.1 in wild birds.  A clade 2.3.2.1 case was identified in Hubei last year, and WHO has selected that isolate, A/Hubei/1/2010 as a pandemic vaccine target.  Last November an H5N1 case was also identified in Hong Kong, A/Hong Kong/6841/2010).  Both isolates are in the phylogenetic tree in figure 3 in the WHO vaccine update.  As seen in the top 100 matches for the Hubei isolate (here) or the Hong Kong isolate (here), wild bird isolate are closely related, although the Hong Kong isolate is closer to sequences from Laos.

Both of the human isolates from 2010 had receptor binding domain changes V223I, D225G, M230I, which are common in wild bird sequences in eastern Asia, and were also in the Gharbya cluster in Egypt in 2006.

The 2010 cases in China were wild type at positions 274 (275) and 294 in NA and therefore susceptible to Tamiflu (oseltamivir).

Although the translation suggests the latest case is wild type at HA positions 226 and 228, but changes at positions 223, 225, and 230 raise concerns that more efficient transmission can be achieved by changes at additional positions. 

Release of the sequences from the Shenzhen cases, as well as those that led to airborne transmission in ferrets would be useful.