Computational studies of H5N1 hemagglutinin binding with SA-α-2, 3-Gal and SA-α-2, 6-

[Sally Furniss]

Re: Computational studies of H5N1 hemagglutinin binding with SA-α-2, 3-Gal and SA-α-2

Link to the paper gsgs mentions http://www.flutrackers.com/forum/showthread.php?t=338

[AlaskaDenise]

Re: Computational studies of H5N1 hemagglutinin binding with SA-α-2, 3-Gal and SA-α-2, 6-

yes, that looks very informative and important but also
very long and hard to read. If I ever understood this well,
then I'd forgotten.
So, if someone could give a summary or conclusion chapter,
that would be useful.


I also just read this on another board: (JKT=Taubenberger)

significance of receptor binding: In this paper that JKT co-authored Glycan Microarray Analysis of the Hemagglutinins from Modern and Pandemic Influenza Viruses Reveals Different Receptor Specificities, Stevens et al they measured the receptor binding affinity of 2 different strains of the 1918 virus, one from South Carolina (18SC) and one from New York (18NY). What they found was that 18NY had one mutation from 18SC. Whereas 18SC showed strong affinity to alpha2,6 (human) and no affinity for alpha2,3 (avian) receptors, 18NY showed significant reduction in binding to alpha2,6 and a gain in affinity to alpha2,3 receptors. And yet, there was no difference in clinical presentation, pathogenesis, mortality, transmissibility between the 2 strains. What does this mean? It could be that receptor binding is only important in an on/off fashion, and not as a matter of degree. That is, as long as there is some affinity, it is sufficient for the virus to do the rest of the job, and the degree of affinity does not affect the outcome at all.

from same paper.....

.....Only one mutation of the NY variant (Asp190Glu) was sufficient to revert the HA receptor preference to that of classical avian strains. Thus, the species barrier, as defined by the receptor specificity preferences of 1918 human viruses compared to likely avian virus progenitors, can be circumvented by changes at only two positions in the HA receptor binding site.

taken from....http://lib.bioinfo.pl/auth:Palese,P
(there are several abstracts at this site on similar subject)

.

[JJackson]

Re: Computational studies of H5N1 hemagglutinin binding with SA-α-2, 3-Gal and SA-α-2, 6-

So, if someone could give a summary or conclusion chapter,
that would be useful.

I am happy to give it a go but of what?
SA binding? Steps to Endocytosis? Infection and the immune reponse? Life the universe and everything? (P.S. not one of the last two please!)

[gsgs]

Re: Computational studies of H5N1 hemagglutinin binding with SA-α-2, 3-Gal and SA-α-2, 6-

hey, your choice .
You decide, what you think is important and what you can do.
I'd like to see a better summary or abstract of the Lia-Wang
paper, (why compute?) but that's just me. Others ?
the long Mingus-lab-thread could also need a summary.

[Mellie]

Re: Computational studies of H5N1 hemagglutinin binding with SA-α-2, 3-Gal and SA-α-2

Mingus and posters to the Mingus lab thread...

http://www.flutrackers.com/forum/showthread.php?t=3901&page=2

Thanks very much. I have been going back over what I missed this summer and as usual, there is much food for thought and analysis. I look forward to that!

[JJackson]

Re: Computational studies of H5N1 hemagglutinin binding with SA-α-2, 3-Gal and SA-α-2, 6-

Why compute?

The three-dimensional configurations of the Hemagglutinin molecule and the way that the binding site interacts with the a2,3 or a2, 6 sialic acid terminal in the glycocalyx gives an indication of the strength and therefore the likelihood of successful bonding. As it is impossible to see this interaction the best that we can do is to visualise it with the aid of a computer model, this has the added advantage that it allows us to envisage the interactions of hypothetical binding site arrangements. As I noted in an earlier post this will allow us to calculate in advance whether changes to the existing binding site structure will increase would decrease the likelihood of affective binding to an alpha 2, 6 sugar. If we begin to see mutations in the known binding site domains which should be able to predict whether these a likely to give stronger alpha 2, 6 binding and therefore increase the chance of human infection. Computer modeling is the only way that we going to be able to do this.

The long thread in Mingus's lab covered such a lot of ground I think it's almost impossible for me condense this.
A few of the points which are worth taking from that long thread are
LP flu has a short cleavage site and is fussy about which proteases can cleave it, HP flu like 1918 and H5N1 can be cleave to buy a wide range of ubiquitous proteases.
The interactions between the different RNA strands is not simple, virulence is difficult to pin down to a sequence on any given strand, mutations in one area on a given strand can turn on, or off, virulence and yet the same mutation at the same point in another virus does not have the same effect. Another example of the interaction between the different proteins coded for by the RNA segments is the ability of neuraminidase to sequester proteases capable of cleaving Hemagglutinin.
Another point of discussion was the effect of passaging the virus in either chicken or MDCK (mouse) cells as the choice of cell will affect the evolution of the virus.
The discussion in the thread then began to look at the role of the proteases in cleavage, as mentioned earlier in seasonal flu Clara protease is normally responsible for cleaving the HA 0 into its two active parts HA1 & HA2 but there are some gram negative bacteria which can produce suitable proteases in the LRT and allow infection via this route.


I hope this helps, I am trying out a new bit of voice recognition software so if there are some odd words or spellings in here that is why.

[gsgs]

Re: Computational studies of H5N1 hemagglutinin binding with SA-α-2, 3-Gal and SA-α-2, 6-

what's proteases,glycocalyx ? MDCK=dog-cells

Can they compute the binding strength, if I apply some hypothetical mutation ?
Can they simulate the evolution of the virus then, to detect whether
2,6 - binding is in reach of random mutations ?

Is the mutation on the cleavage site in Indonesia important ?

[Mingus]

Re: Computational studies of H5N1 hemagglutinin binding with SA-α-2, 3-Gal and SA-α-2

what's proteases,glycocalyx ? MDCK=dog-cells

Can they compute the binding strength, if I apply some hypothetical mutation ?
Can they simulate the evolution of the virus then, to detect whether
2,6 - binding is in reach of random mutations ?

Is the mutation on the cleavage site in Indonesia important ?

I beleive the research team who did this computational model can answer all thoses questions.

The proplem with "mutation seeking" before this study was that it was mainly based on speculative extrapolation from the mutation who had effect on others flu sub-type like H3, H2 & H1.

With a model who seem reliable and who fit the now known experimental data about glycan affinity, they show basically thoses mutation to be the relevant ones for the H5 subtype glycan affinity:

Tyr 98, Trp 153, His 183, Glu 190, and the strongestGln 226

Also others relevant for a2,3:
Ser 136, Ile 155, Leu 194
Val 135, Ser 136, Ser 137

others relevant for a2,6:
Trp 153, Ile 155,
Leu 133, Val 135, Lys 193

The list is long but we are now sure that theses are THE mutations to check out given what we know from now.