The Norwegian Institute of Public Health recently identified a mutation in 2009 H1N1 influenza virus isolated from two patients who died and one with severe disease. It has been suggested that this mutation, which causes a change from the amino acid aspartic acid to glycine at position 225 of the viral HA protein (D225G), could make the virus more likely to infect deeper in the airways and cause more severe disease. What is the basis for this concern and does it have merit?
Attachment of all influenza A virus strains to cells requires sialic acids. There are a number of chemically different forms of sialic acids, and influenza virus strains vary in their affinity for them. Human influenza A strains bind preferentially to sialic acids linked to galactose by an alpha(2,6) bond, while avian and equine strains prefer alpha(2,3) linked sialic acids.
The type of sialic acid preferred by influenza viruses is controlled by amino acids in the HA protein. Amino acids 190 and 225 are important determinants of receptor binding specificity of the 1918 H1 hemagglutinin. The HA of the 1918 strain A/South Carolina/1/18 prefers alpha(2,3) linked sialic acids; the New York variant, isolated in September 1918, binds both alpha(2,3) and alpha(2,6) sialic acids. These two H1 hemagglutinins differ only by a single amino acid, position 225, which is aspartic acid (D) in the South Carolina strain and glycine (G) in the NY strain. When amino acid 190, which is D in both strains, is changed to E in the NY HA, the virus (AV18) preferentially binds alpha(2,3) sialic acids. These findings are summarized in the table.
Different isolates of the 2009 H1N1 influenza virus have D at HA at amino acid 190 and mostly D at amino acid 225. The virus prefers to bind to alpha(2,6) linked sialic acids. The amino acid change D225G would be expected to produce a virus with preference for both alpha(2,3) and alpha(2,6) linked sialic acids.
In the human respiratory tract, alpha(2,6) linked sialic acids are dominant on epithelial cells in the nasal mucosa, paranasal sinuses, pharynx, trachea, and bronchi. Alpha(2,3) linked sialic acids are found on nonciliated bronchiolar cells at the junction between the respiratory bronchiole and alveolus, and on type II cells lining the alveolar wall.
Based on these considerations, it could be hypothesized that the D225G change would allow the 2009 H1N1 virus to replicate deeper in the respiratory tract. But 2009 H1N1 virus without this amino acid change can already replicate deep in the respiratory tract of ferrets, and probably also in humans. Cells with alpha(2,6) linked sialic acids are present in the lower respiratory tract of humans. So it’s not clear if any effect on virulence would be conferred by the ability of the 2009 H1N1 strain to bind alpha(2,3) linked sialic acids.
An important consideration is that the D225G amino acid change has a negative impact on transmission. The change from D to G at amino acid 225 of the 1918 HA significantly impairs transmission among ferrets. When both D225G and D190E are present, transmission is abolished. These changes do not impair viral replication or virulence in the respiratory tract of inoculated animals.
Transmissibility is clearly a positive selection factor for viral evolution. There may be selection for increased virulence only if there is no negative impact on viral transmission. Given these considerations, the choice between an H1 HA amino acid at position 225 that allows efficient transmission (D225) or one that impairs transmission and might or might not allow multiplication deeper in the lung (D225G) seems obvious.
Tumpey, T., Maines, T., Van Hoeven, N., Glaser, L., Solorzano, A., Pappas, C., Cox, N., Swayne, D., Palese, P., Katz, J., & Garcia-Sastre, A. (2007). A Two-Amino Acid Change in the Hemagglutinin of the 1918 Influenza Virus Abolishes Transmission Science, 315 (5812), 655-659 DOI: 10.1126/science.1136212
Shen J, Ma J, & Wang Q (2009). Evolutionary Trends of A(H1N1) Influenza Virus Hemagglutinin Since 1918. PloS one, 4 (11) PMID: 19924230
Pingback: Tweets that mention The D225G change in 2009 H1N1 influenza virus is not a concern -- Topsy.com
But 2009 H1N1 virus without this amino acid change can already replicate deep in the respiratory tract of ferrets
How labile are flu viruses within an individual patient? Is the D225G polymorphism transmitting between patients, or is the same mutation recurring in many different patients? One could hypothesise a scenario where only virus particles with a D at position 225 are able to transmit, but the subsequent course of the infection depends on whether the the virus happens to pick up a G at position 225 within each given patient. Whether you detect a D or a G when you come to sequence a sample would depend on how long-established the infection was, how deep you were sampling within the respiratory tree, etc.
On a related note, how certain are we about what was going on in the ferrets? If I remember rightly, the study infected ferrets with H1N1 flu (i.e. with D at 225) and observed deep lung pathology. Did they then re-isolate the virus from the deep lung lesions and check whether it still had a D at position 225?
Perhaps the pathogenic factor is not whether a given strain has G at 225, but the ease (or otherwise) with which it can acquire a G as it replicates within the host.
Same goes for any other polymorphism that increases lethality but decreases transmission. If you need half a dozen nucleotide changes to get from the transmissible sequence to a lethal sequence, you're safer than if it's only a single hop away.
Good questions pjie2.
However, the question remains which selective advantages such sporadic “in-patient-mutated” virus could have over its non-mutated competitors during the course of the infection within the patient ?
As long as the “front of the infection wave” has not reached the deep lung tissue, the reproduction rate of the non-mutated version would be superior… and will thus stop before the deep lung tissue is reached.
Maybe it is only those rare cases where the “jump” happen to occur in a virus particle just at the peripheric border of the infection wave far enough down the respiratory tree to reach the first susceptible “2,3”-type cells. from which point on a “second wavefront” of the infection would be started running down the lower part of the respiratory tract.
just a thought …
Eh, that implies spread of flu in the respiratory tree is like some kind of slow-moving fluid, rather like the adverts for cough syrup. Does flu actually work like that? I mean, an air molecule can get from the outside world down to the deepest parts of the respiratory tree within a single breath – it's how breathing works. Virus particles won't be quite that mobile, but even so…
The picture I had in mind was a lot simpler – you have an active upper respiratory infection with D at 225. Periodically, random mutations happen which create a G at 225 – along with a host of other mutations, of course, since the sheer number of viral RNAs synthesised pretty much guarantees that any given mutation will arise at least once. These mutant particles get released and go wherever the air flow takes them, be that up or down.
If they get coughed out and land on someone else, there's no effect as they don't transmit well. If they get inhaled deeper into the lungs, then they may set up shop there. They won't get outcompeted by the parent virus because the parent virus isn't infecting the same cells.
I wonder if new deep sequencing technology will help answer these questions. Would it let you look at the whole population of viral sequences within an infection, and/or at different locations within the lung, rather than getting a consensus sequence?
Pingback: uberVU - social comments
I read, that there are 10 sequences from Ukraine at the secret database “GISAID”
and 4 of these are from lethal cases and exactly those 4 have 225G.
Is it true ? Why no comment on this , WHO,CDC,ECDC,journalists ?
Your table doesn't seem consistent with the text above it. Text: “The HA of the 1918 strain A/South Carolina/1/18 prefers alpha(2,3) linked sialic acids”. Table: cross in the (2,6) column and none in the (2,3) column. I don't know which is right?
Below the table, you write “Different isolates of the 2009 H1N1 influenza virus have D at HA at amino acid 190 and mostly D at amino acid 225. The virus prefers to bind to alpha(2,3) linked sialic acids.” and I think you mean (2,6) here too? If not I'm very confused.
From what I've read we don't seem to have a great track record at predicting the clinical features of a virus from changes to its genome. We don't really know enough about the genetics to predict the effect of a certain new mutation. So I'm not convinced by this line of reasoning.
The ferret model isn't terribly convincing either. An early study on the a Mexican strain of swine flu showed an astoundingly high death rate in ferrets although the same strain in people was relatively mild. Other experiments in ferrets concerning the transmissibility of the virus also seemed to contradict observations on human epidemiology.
Equally, of course, I'm not convinced that the mutation should concern us just from genetic speculation.
As far as I'm concerned the question, as to how concerned we should be, is best answered by observation of human clinical subjects.
Maybe after they read this post they will realize that the mutation
doesn't matter.
Thank you for picking up those errors; I've corrected them. For some
reason, when it comes to alpha(2,3) and alpha(2,6) preferences, I
always mix them up. I just did it last week when I lectured medical
students. One emailed me afterwards to let me know about it. At least
I know some people are listening!
“An important consideration is that the D225G amino acid change has a negative impact on transmission. The change from D to G at amino acid 225 of the 1918 HA significantly impairs transmission among ferrets. When both D225G and D190E are present, transmission is abolished”
Is there more data on this statement? It would be great to review it.
Thanks
The data for that statement are included in the Science paper
referenced at the bottom of the post.
Yes indeed 🙂 That's what happens when you go from microscope to computer screen 12 hours a day.
Thanks
Pingback: [Avian Flu Diary] Referral: Virology Blog On D225G Mutation | Influenza Virus Mashup
Very informative. If I understand this correctly D225G does not necessarily mean that a given individual will contract a severe case if they are infected with it, as compared to the wild strain. Is this because it all depends on the G placement at position 225? From the article it appears that that between the two we are looking lots of infection , little death (Wild), or little infection, lots of death (mutant). If that is the case it seems at present there is a fight for supremacy going on. Would it be a fair assumption to suggest that because of the ever growing number of cases of D225G becoming visible that we may be at a tipping point between the two? If D225G does not match up with 1918/1919 exactly, what would be a clearer indicator that we are going down that path? Additionally since many people still have not been vacinated both healthy and otherwise, would it be a prudent move to get a Pneumovax? Thanks…
47 comments 0 likes 2 points
but only 12 comments are shown. Is there so much spam or censorship ?
there are hundreds of HA-spikes, some may have 225D and some 225G,
so such a virus could bind to both types of cells. If that combined HA-production
sheme could be encoded genetically wouldn't that make a major advance in flu-genetics ?
no forum yet, sigh
Any comments on the D225G variants being 'low reactors'?
Pingback: The D225G change in 2009 H1N1 influenza virus is not a concern « FluTrackers
I don't think you have made any valid arguments to conclude that D225G doesn't matter. You have pretty much ignored that vast majority of actual data related to D225G in swine H1N1 and instead presented a theoretical arument that really doesn't fit the data.
The interest in D225G increased dramatically with the outbreak in Ukraine. Large numbers of previously healthy young adults began showing up at emergency rooms coughing up blood and dying within a few days. Both lungs were destroyed within a couple of days, and the large number of cases produced a significant amount of panic. Initially the infections were called a “mystery disease” but had all of the landmarks of a swine H1N1 infection.
Ukraine request help from WHO, who sent in a team of nine and sent samples off to Mill Hill in London. Most of the samples were from the hard hit regions of western Ukraine where hundreds had died.
I had looked at existing sequences and predicted the large number of hemorrhagic cases were due to a receptor binding domain and focused on position 225 which had already changed to D225E, D225N, and D225G in samples collected worldwide. I predicted the change would be D225G.
The sequences were withheld for 1-2 weeks, but were released at GISAID on Nov 18. Ten HA sequences from 10 individuals were included. One had been collected from an earlier cases in Kiev, while the other nine were from three locations in western Ukraine. Phylogenetic analysis showed that all 9 were from the same sub-clade and four had D225G. The next day more demographic data was added, and it was clear that 4 of the 10 patients had died, and all four had D225G. Three were from lung and one was from throat. The five survivors had the same sub-clade, but were wild type for position 225. The samples in the five survivors were nasopharygeal washes. The four positives were from two Oblasts and none were linked, providing STRONG evidence that D225G transmits.
Mill Hill has stated that the tested isolate from western Ukraine with D225G was California/7-like, low reactor. D225G was the only nonsynonymous HA change.
Another isolate from Kiev, which was wild type at position 225, was California/7-like.
I have many more examples of REAL data on D225G, including escape from immune recognition. Please repond to Ukraine data (D225G in four of four fatal cases). I have enough examples to keep your hands waving for DAYS.
Do you think the 4 fatal examples in Ukraine were SPONTANEOUS???????????????????
Reality check. In Ukraine four of four fatal cases had D225G (ZERO of five nasal washes from survivors had D225G, even though they were the same sub-clade). These cases were in two Oblast and NOT linked.
D225G TRANSMITS. The number of nonsynonymous HA changes was ONE (D225G).
Comments (hand waving allowed)?
I would say, the fact that they were not linked provides evidence against
tranmission, not for it.
it would be interesting to have several samples from one person, throat,nose lungs
and at different times.
In southern California, in March/April, 2009 the first two cases of swine H1N1 were in adjacent counties (San Diego and Imperial), in two people who had no link to swine or each other and were located over 100 miles apart. The finding of the same change in individuals who were NOT linked showed thet the swine H1N1 was widespread and transmitting EFFICIENTLY.
The same is true for the detection of the same polymorphism in four of four fatalities in western Ukraine, where cases were in the two most hard hit Oblasts and not linked, WHO handwaving notwithstanding.
Do you think that four of four was a coincidence of SPONTANEOUS “mutations”
There appear to be multiple samples from the first fatality in Norway, but so far only one has been made public (without D225G, but the same subclade as Ukraine and the first public sequence from Norway with D225G_.
> In southern California, the first two cases of swine H1N1 were in adjacent counties
> (San Diego and Imperial), in two people who had no link to swine or each other and
> were located over 100 miles apart. The finding of the same change in individuals who
> were NOT linked showed thet the swine H1N1 was widespread and transmitting EFFICIENTLY.
it could just have been occasional introductions from travelers without spreading.
We saw that often in Europe in May,June
> The same is true for the detection of the same polymorphism in four of four fatalities
> in western Ukraine, where cases were in the two most hard hit Oblasts and not linked
no proof for that. If it typically happens together with other polymorphisms, which are not present
in D225, that I would consider evidence for your view. Do we see this ?
> Do you think that four of four was a coincidence of SPONTANEOUS “mutations”
yes, ATM I consider that the most likely explanation. It seems to be linked to lung-samples,
but lung-viruses can't easily go into sputum and transmit.
Evidence for transmission would also be examples where 225G=239G(H1)
was found early in the disease. Do we have such examples ?
I think you need to take a step back and look at reality, instead of posting a number of ad hac arguments that are not reality based. H1N1 spreads and transmits efficiently in humans and has been doing so since the spring of 2009 (or earlier). This transmission doesn't require a lab test or a sequence from the database. In southern California, the lack of linkage demonstrated that the virus was spreading throught the community. The contacts of the two positives had symptoms (which were mild and they were not tested). Within days positives were appearing throughout the US, including classmates in San Antonio, Texas. The virus was spreading in US as well as Europe (although more in the US because of location). When European countries focused efferts on border checks the virus continued to spread within the European countries, but was undetected/unreported because of limited testing. The same was true throughout the world. Although lack of testing controls REPORTED cases, the lack of testing does not control the spread.
Hand waving about spontaneopus mutations and coincidences of 4 of 4 fatal cases are simply nonsensical exercises. Knock yourself out, If you think the 4 or 4 fatal cases in Ukraine is a coincidence along with a match of the subclade in Norway with D225G, talk to reporters. They might believe you and continue to distribute the “spontaneous mutation” nonsense to a guilible audience.
D225G was in California/7/2009.
So would I.
I don't think I have ever heard anyone say that a LACK of linkage between cases supports evidence of transmission. That's certainly an 'interesting' stance.
A sample of 4 is statistically insignificant as well.
What is the likelihood of transmission of HA wild type D225 infecting a human host and proliferating in the entire respiratory tree as both D225G and wild type. This would sustain transmissibility, but increase virulence. Without respiratory sampling from multiple sites how can we know this is not what is happening.
It does seem a bit far-fetched that the recent European deaths tightly associated with D225G are coincidental. Unfortunately those of us without a subscription can't get to the Science article to view the data behind the assertion that the D225G is no problemo because of limited transmission.
Clearly, D225G strains are not reacting with the vaccine that has been used to vaccinate tens of millions around the globe. That sounds like a great way to select for D225G. If Vincent is right, no worries. If niman is right, it'll be Katie-bar-the-door.
Bets, gentlemen?
It does seem a bit far-fetched that the recent European deaths tightly associated with D225G are coincidental. Unfortunately those of us without a subscription can't get to the Science article to view the data behind the assertion that the D225G is no problemo because of limited transmission.
Clearly, D225G strains are not reacting with the vaccine that has been used to vaccinate tens of millions around the globe. Of course, D225 does react. That sounds like a great way to select for D225G. If Vincent is right, no worries. If niman is right, it'll be Katie-bar-the-door.
Bets, gentlemen?
It does seem a bit far-fetched that the recent European deaths tightly associated with D225G are coincidental. Unfortunately those of us without a subscription can't get to the Science article to view the data behind the assertion that the D225G is no problemo because of limited transmission.
Clearly, D225G strains are not reacting with the vaccine that has been used to vaccinate tens of millions around the globe. That sounds like a great way to select for D225G. If Vincent is right, no worries. If niman is right, it'll be Katie-bar-the-door.
Bets, gentlemen?
It does seem a bit far-fetched that the recent European deaths tightly associated with D225G are coincidental. Unfortunately those of us without a subscription can't get to the Science article to view the data behind the assertion that the D225G is no problemo because of limited transmission.
Clearly, D225G strains are not reacting with the vaccine that has been used to vaccinate tens of millions around the globe. Of course, D225 does react. That sounds like a great way to select for D225G. If Vincent is right, no worries. If niman is right, it'll be Katie-bar-the-door.
Bets, gentlemen?
Pingback: H1N1 D225G mutation a reason of concern Part 2 « The LABVIRUS.COM Blog
Pingback: The D225G change in 2009 H1N1 influenza virus
Come on now gentlemen please play the game fairly and post this comment, you can’t stifle valid criticism and still call yourselves scientists – or has the Spanish Inquisition taken over and made virology sacrosanct and immune to scientific scrutiny ????????? So here’s the question again:
Precisely how did the Norwegian Institute of Public Health isolate the H1N1 influenza virus from which the alleged ‘mutation’ is supposed to cause a change from aspartic acid to glycine at position 225 of the alleged viral HA protein???? I bet one doesn’t get a really convincing scientific answer to that question, but until this question is effectively answered the physical existence of H1N1 will remain an unproven dogma.
My last comment didn’t make it to the blog – perhaps one struck a nerve somewhere. I thought that science is supposed to be an open book and is about posing and solving specific problems. The question of the effective isolation, biochemical characterization etc, of H1N1 is a problem, and that problem has not been solved to date despite all claims to the contrary. All we seem to obtain are speculative experimental conclusions without the premises and a waste of money.
Niman can only be right [and he’s far from being right] if and only if a complete virus was physically isolated and charcterized from the so-called ‘isolates’ allegedly ‘tested’ and there does not appear to be any evidence at all that could warrant such a conclusion. Its all smoke and mirrors and wishful thinking H1N1 and D225 strains are bunkum until proved otherwise.
I would like to add that any claim that a complete H1N1 particle has been effectively isolated [pure and free of contaminants], photographed and biochemically characterized without providing ALL the PROOF so that the methodologyand technique in question can be adequately scritinized is a scientific misrepresentation of fact and is tantamount to scientific fraud.
November J Virology, doi:10.1128/JVI.01136-10, shows D225G transmits via aerosol droplets in ferrets and guinea pigs and binds to alpha 2,3 AND alpha 2,6 as stated previously (and as is OBVIOUS from sequence analysis of pH1N1 fatal cases in Ukraine and Russia).
Discussion of paper at
http://fluboard.rhizalabs.com/forum/viewtopic.php?f=5&t=6297
The are thousands of public H1N1 sequences, from isolates, no pictures required.
Henry, you keep going on about gene sequences allegedly obtained from H1N1 isolates but you cant grow isolates without first isolating H1N1. You merely paotulate the existence of H1N1 like everyone else but you cant prove scientifically that it even exists can you other than a virtual realityon your computer.
Excuse the incomplete and somewhat garbled reply but my longer one didn’t make it.
Henry, electron micrographs (pictures) of the isolated virus in question are always required. If there is no virus then there is no picture to be required!
What sort of crock is this: ‘There APPEAR to be multiple samples from the first fatality in Norway’. Well folks, either there were samples taken from the fatality or there were not, if there were, who took them and from what, and precisely who and how was the complete virus isolated from which the D225G was allegedly derived??????????? I betcha can’t answer that question in full can you? If by some quirk of fate you can, then I’d be more than happy to subject it to the scientific scrutiny that it deserves, so feel free to post it.
The numbering of the isolate implied several isolates came from the same sample. D225G has been found in hundreds of samples, including those published in peer reviewed journals. You are just posting utter nonsense.
On a related note, the CDC just announced the SIXTH case of human triple reassortant H3N2 (trH3N2) from a hospitalized case in Minnesota (in November, 2010). This is the THIRD month in a row for an announced trH3N2 case in the US, and second case from MN this year. All six cases have been reported AFTER the start of the 2009 H1N1 pandemic.
Like all trH3N2 isolates it has a HUMAN H3 from the 1990’s and has D225G, which is true for all of the trH3N2 isolates, as well as SEASONAL H3N2 circulating in the mid to late 1990’s. The D225G in seasonal H3N2 TRANSMITS (as do recent trH3N2 cases base on sequence similarities.
You are posting utter nonsense. H1N1 is quite real as is D225G. Your posts are “internet babble”, at best (no picture required).